Traditional CAPTCHA (Completely Automated Public Turing Test to Tell Computers and Humans Apart) schemes are increasingly vulnerable to automated attacks powered by deep neural networks (DNNs). Existing adversarial attack methods often rely on the original image characteristics, resulting in distortions that hinder human interpretation and limit their applicability in scenarios where no initial input images are available. To address these challenges, we propose the Unsourced Adversarial CAPTCHA (DAC), a novel framework that generates high-fidelity adversarial examples guided by attacker-specified semantics information. Leveraging a Large Language Model (LLM), DAC enhances CAPTCHA diversity and enriches the semantic information. To address various application scenarios, we examine the white-box targeted attack scenario and the black-box untargeted attack scenario. For target attacks, we introduce two latent noise variables that are alternately guided in the diffusion step to achieve robust inversion. The synergy between gradient guidance and latent variable optimization achieved in this way ensures that the generated adversarial examples not only accurately align with the target conditions but also achieve optimal performance in terms of distributional consistency and attack effectiveness. In untargeted attacks, especially for black-box scenarios, we introduce bi-path unsourced adversarial CAPTCHA (BP-DAC), a two-step optimization strategy employing multimodal gradients and bi-path optimization for efficient misclassification. Experiments show that the defensive adversarial CAPTCHA generated by BP-DAC is able to defend against most of the unknown models, and the generated CAPTCHA is indistinguishable to both humans and DNNs.

As a representative classification model, decision tree has been extensively applied in data mining. It generates a series of if-then rules based on the homogeneity of class distribution. In a society where data spreads everywhere for knowledge discovery, the privacy of the data respondents is likely to be leaked and abused. Based on this concern, we propose an overview of the rapidly evolving research results focusing on privacy preserving decision tree induction. The research results are summarised according to the characteristics of related privacy preservation techniques, which include data perturbation, cryptography, and data anonymisation. In addition, we demonstrate the comparison between the merits and demerits of these methods considering the specific property of decision tree induction. At last, we conclude the future trend of privacy preserving techniques.

The rapid development of pay-as-you-go cloud services motivates the increasing number of cloud resource demands. However, the volatile demands bring new challenges for current techniques to minimize the cost of cloud capacity planning and VM provisioning while satisfying the customer demands. The service vendors will incur enormous revenue loss within the long-term inappropriate planning, especially when the demands fluctuate abruptly and frequently. In this paper, we cast the cloud capacity planning as a classification problem and propose an integrated framework, which effectively predicts the abrupt changing demands, to reduce the cost of cloud resource provisioning. In this framework, we first apply Piecewise Linear Representation to segment the time series of cloud resource demands for labeling the changing trend of each period. Second, Weighted SVM is leveraged to fit the statistical information and the label of each period and predict the changing trend of the following period. Finally, an incremental learning strategy is utilized to ensure the low cost of updating the model using the upcoming requests. We evaluate our framework on the IBM Smart Cloud Enterprise (SCE) trace data and the experimental results show the effectiveness of our proposed framework.

Mobile crowd sensing emerges as a new paradigm which takes advantage of the pervasive sensor-embedded smartphones to collect data. Many incentive mechanisms for mobile crowd sensing have been proposed. However, none of them takes into consideration the cooperative compatibility of users for multiple cooperative tasks. In this paper, we design truthful incentive mechanisms to minimize the social cost such that each of the cooperative tasks can be completed by a group of compatible users. We study two bid models and formulate the Social Optimization Compatible User Selection (SOCUS) problem for each model. We also define three compatibility models and use real-life relationships from social networks to model the compatibility relationships. We design two incentive mechanisms, MCT - M and MCT - S, for the compatibility cases. Both of MCT - M and MCT - S consist of two steps: compatible user grouping and reverse auction. We further present a user grouping method through neural network model and clustering algorithm. Through both rigorous theoretical analysis and extensive simulations, we demonstrate that the proposed mechanisms achieve computational efficiency, individual rationality, and truthfulness. Moreover, MCT - M can output the optimal solution. By using neural network and clustering algorithm for user grouping, the proposed incentive mechanisms can reduce the social cost and overpayment ratio further with less grouping time.

The point-of-interest (POI) recommender exploits check-in information from location-based social networks (LBSNs) to recommend POIs that match user preferences. The preferences of users vary with region/locality, consumption type, and co-consumers. For instance, the places preferred for friends may be different from the places preferred for family, and the places preferred in one locality may be different from the places in another locality. These dynamic preferences can be crucial for an efficient recommender system. A locality consists of different preference trends, such as “known for food and recreational sites.” In real-world, different sets of users might be attracted toward different preference trends, and some users' preferences might overlap across multiple preference sets. Hence, an efficient aggregation of locality preferences is essential for the recommender systems. Many existing studies simply group items by their category and use simple collaborative filtering (CF) for recommendations, however, such techniques cannot efficiently handle the aggregated locality preferences. We propose a hierarchical recommendation model termed Hierarchical Contextual Location Recommendation System (HiRecS) that formulates users' preferences as a hierarchical structure and models the locality trend using aggregated hierarchy. For a locality, the root of hierarchy contains preferred k items from a set of visitors, and the subsequent levels contain preference wise subsets of those items. We also present a hierarchy aggregation technique to aggregate the hierarchical preferences from a similar set of users. The aggregated hierarchy is then contextually exploited for POI sequence recommendation. The core contributions of this article are: 1) it formulates the locality trends as hierarchical structures, presents a hierarchy aggregation technique, and models the personalized POI preferences with the aggregated hierarchy; 2) it exploits the aggregated popular trends to generate contextual POI sequence recommendation; and 3) it extensively evaluates the proposed model with two real-world data sets and demonstrates a significant performance gain of 0.006-5.91 on diversity metrics, 0.0349-17.51 on displacement metrics, and 0.114-0.289 on normalized discounted cumulative gain (NDCG) metrics, when compared to several baselines and relevant studies.

Online interactive recommender systems strive to promptly suggest users appropriate items (e.g., movies and news articles) according to the current context including both user and item content information. Such contextual information is often unavailable in practice, where only the users' interaction data on items can be utilized by recommender systems. The lack of interaction records, especially for new users and items, inflames the performance of recommendation further. To address these issues, both collaborative filtering, one of the most popular recommendation techniques relying on the interaction data only, and bandit mechanisms, capable of achieving the balance between exploitation and exploration, are adopted into an online interactive recommendation setting assuming independent items (i.e., arms). This assumption rarely holds in reality, since the real-world items tend to be correlated with each other. In this paper, we study online interactive collaborative filtering problems by considering the dependencies among items. We explicitly formulate item dependencies as the clusters of arms in the bandit setting, where the arms within a single cluster share the similar latent topics. In light of topic modeling techniques, we come up with a novel generative model to generate the items from their underlying topics. Furthermore, an efficient particle-learning based online algorithm is developed for inferring both latent parameters and states of our model by taking advantage of the fully adaptive inference strategy of particle learning techniques. Additionally, our inferred model can be naturally integrated with existing multi-armed selection strategies in an interactive collaborative filtering setting. Empirical studies on two real-world applications, online recommendations on movies and news, demonstrate both the effectiveness and efficiency of our proposed approach.

The importance of mining time lags of hidden temporal dependencies from sequential data is highlighted in many domains including system management, stock market analysis, climate monitoring, and more. Mining time lags of temporal dependencies provides useful insights into the understanding of sequential data and predicting its evolving trend. Traditional methods mainly utilize the predefined time window to analyze the sequential items, or employ statistical techniques to identify the temporal dependencies from a sequential data. However, it is a challenging task for existing methods to find the time lag of temporal dependencies in the real world, where time lags are fluctuating, noisy, and interleaved with each other. In order to identify temporal dependencies with time lags in this setting, this paper comes up with an integrated framework from both system and algorithm perspectives. Specifically, a novel parametric model is introduced to model the noisy time lags for temporal dependencies discovery between events. Based on the parametric model, an efficient expectation maximization approach is proposed for time lag discovery with maximum likelihood. Furthermore, this paper also contributes an approximation method for learning time lag to improve the scalability in terms of the number of events, without incurring significant loss of accuracy.

An efficient delivery of IT services for increasingly complex IT environments demands an intelligent automated solution for resolving existing and potential issues. An automation recommender system, promptly suggesting the most proper scripted resolution to an arriving IT incident ticket, would play a significant role in IT automation services. Hence, developing a comprehensive framework supporting becomes imperative for continuous improvement of automation recommendation.In this paper, we first identify the challenges of IT services followed by a discussion on AISTAR (an intelligent system for online IT ticket automation recommendation) designed and developed to provide them. Specifically, we define and formalize automation recommendation procedure as a multi-armed bandit problem with dependent arms, which is capable of achieving the optimal tradeoff between exploitation of the system for the best automation recommendation and exploration of automation execution information for future recommendation. Two novel multi-armed bandit models are proposed and integrated to handle the aforementioned challenges in IT automation services. Empirical studies on a large ticket dataset from IBM Global Services demonstrate both the effectiveness and efficiency of our intelligent integrated system. AISTAR is earmarked for Cognitive Event Automation for IBM Service delivery.

The Point-of-Interest (POI) preference of a user varies by locality, item type, and the co-visitors, e.g., user1 and user2 can have closest preference on food items but not on historic sites, etc. A locality can have different preference trends (e.g., popular for food, recreation, etc.) and a user's preference can span across multiple such trends. A good recommender should also exploit the aggregated locality preference trends. Most of the existing studies group items by category or global user preferences which might not be relevant for locality-based aggregated preferences. We propose HiCaPS (Hierarchical Contextual POI Sequence Recommender) that formulates user preferences as hierarchical structure and presents a hierarchy aggregation technique for POI recommendation. The top level of locality hierarchy contains preferred k items from a set of users and the subsequent levels contain preference wise subsets. The core contributions of this paper are: (i) it formulates user preferences as a preference hierarchy, presents a technique to aggregate preference hierarchies of a similar users, and models the target users' preference in terms of aggregated trend in a locality, (ii) it contextually exploits the aggregated trend to generate personalized POI sequences, and (iii) it extensively evaluates the proposed model with two real-world datasets and demonstrates performance gain (0.03 - 0.28 on pair F-score, 0.006 - 5.91 on diversity, 0.0349 - 17.51 on displacement, and 0.114 - 0.289 on NDCG) over baseline models.

Service providers are facing an increasingly intense competition and growing industry requirements, which dictates efficient and cost-effective service delivery. This is largely achieved by maximizing automation of IT maintenance procedures. Automation it largely depend on classification of the tickets. The automation of ticket classification requires labeled data, which is usually triaged by manually generated rules on word features. In this paper, we propose an unsupervised approach for facilitation of rule generation for labeling tickets created by event management. We first identify and remove noisy tickets with generic resolutions, and then use sparse classic canonical analysis (CCA) for feature selection to enable an efficient rule generation. Furthermore we discuss results of an extensive empirical study of ticket data that was conducted to validate the effectiveness and efficiency of our method.

The location-based social networks (LBSN) (e.g., Facebook, etc.) have been explored in the past decade for Point-of-Interest (POI) recommendation. Many of the existing systems focus on recommending a single location or a list which might not be contextually coherent. In this paper, we propose a model termed CLoSe (Contextualized Location Sequence Recommender) that generates contextually coherent POI sequences relevant to user preferences. The POI sequence recommenders are helpful in many day-to-day activities, for e.g., itinerary planning, etc. To the best of our knowledge, this paper is the first to formulate contextual POI sequence recommendation by exploiting Recurrent Neural Network (RNN). We incorporate check-in contexts to the hidden layer and global context to the hidden and output layers of RNN. We also demonstrate the efficiency of extended Long-short term memory (LSTM) in sequence generation. The main contributions of this paper are: (i) it exploits multi-context, personalized user preferences to formulate contextual POI sequence generation, (ii) it presents contextual extensions of RNN and LSTM that incorporate different contexts applicable to a POI and POI sequence, and (iii) it demonstrates significant performance gain of proposed model on pair-F1 and NDCG metrics when evaluated with two real-world datasets.

The increasing complexity of IT environments urgently requires the use of analytical approaches and automated problem resolution for more efficient delivery of IT services. In this paper, we model the automation recommendation procedure of IT automation services as a contextual bandit problem with dependent arms, where the arms are in the form of hierarchies. Intuitively, different automations in IT automation services, designed to automatically solve the corresponding ticket problems, can be organized into a hierarchy by domain experts according to the types of ticket problems. We introduce a novel hierarchical multi-armed bandit algorithms leveraging the hierarchies, which can match the coarse-to-fine feature space of arms. Empirical experiments on a real large-scale ticket dataset have demonstrated substantial improvements over the conventional bandit algorithms. In addition, a case study of dealing with the cold-start problem is conducted to clearly show the merits of our proposed algorithms.

Mobile crowdsourcing has become an efficient paradigm for performing large scale tasks. The incentive mechanism is important for the mobile crowdsourcing system to stimulate participants, and to achieve good service quality. In this paper, we focus on solving the insufficient participation problem in the budget constrained online crowdsourcing system. We present a two-tiered social crowdsourcing architecture, which can enable the selected registered users to recruit their social neighbors by diffusing the tasks to their social circles. In the two-tiered social crowdsourcing system, the tasks are associated with different end times, and both the registered users and their social neighbors have different arrival/departure times. An online incentive mechanism, MTSC, which consists of two steps: Agent Selection and Online Reverse Auction, is proposed for this novel mobile crowdsourcing system. Through both rigorous theoretical analysis and extensive simulations, we demonstrate that the proposed incentive mechanism achieves computational efficiency, individual rationality, budget feasibility, cost truthfulness, and time truthfulness.

Understanding contents in social networks by inferring high-quality latent topics from short texts is a significant task in social analysis, which is challenging because social network contents are usually extremely short, noisy and full of informal vocabularies.Due to the lack of sufficient word co-occurrence instances, well-known topic modeling methods such as LDA and LSA cannot uncover high-quality topic structures.Existing research works seek to pool short texts from social networks into pseudo documents or utilize the explicit relations among these short texts such as hashtags in tweets to make classic topic modeling methods work.In this paper, we explore this problem by proposing a topic model for noisy short texts with multiple relations called MRTM (Multiple Relational Topic Modeling).MRTM exploits both explicit and implicit relations by introducing a document-attribute distribution and a two-step random sampling strategy.Extensive experiments, compared with stateof-the-art topic modeling approaches, demonstrate that MRTM can alleviate the word co-occurrence sparsity and uncover highquality latent topics from noisy short texts.

The Location-Based Social Networks (LBSN) (e.g., Facebook, etc.) have many attributes (e.g., ratings, reviews, etc.) that play a crucial role for the Point-of-Interest (POI) recommendations. Unlike ratings, the reviews can help users to elaborate their consumption experience in terms of relevant factors of interest (aspects). Though some of the existing systems have exploited user reviews, most of them are less transparent and non-interpretable (as they conceal the reason behind recommendation). These reasons have motivated us towards explainable and interpretable recommendation. To the best of our knowledge, only a few of the researchers have exploited user reviews to incorporate the sentiment and opinions on different aspects for personalized and explainable POI recommendation. This paper proposes a model termed as ReEL (Review aware Explanation of Location Recommendation) which models the review-aspect correlation by exploiting deep neural network, formulates user-aspect bipartite relation as a bipartite graph, and models the explainable recommendation by using dense subgraph extraction and ranking-based techniques. The major contributions of this paper are: (i) it models users and POIs using the aspects posted on user reviews, and it provisions incorporation of multiple contexts (e.g., categorical, spatial, etc.) in POI recommendation, (ii) it formulates preference of users' on aspects as a bipartite relation, represents it as a location-aspect bipartite graph, and models the explainable recommendation with the notion of ordered dense subgraph extraction using bipartite cores, shingles, and ranking-based techniques, and (iii) it extensively evaluates the proposed models using three real-world datasets and demonstrates an improvement of 5.8% to 29.5% on F-score metric, when compared to the relevant studies.

In this study, a novel normal curvature‐induced variational model which involves a higher‐order regulariser based on the normal curvature prior information of image surface is proposed for image restoration. Furthermore, the authors derive a preferably equivalent formulation for the proposed normal curvature‐induced higher‐order regulariser. Then, they design an efficient algorithm to solve the proposed model by using the famous alternating direction method of multipliers technique. Finally, they assess the performance of the proposed method on both natural images and biomedical cell images by comparing it with the famous fast total variation (TV) method, fractional‐order TV method and Hessian‐nuclear‐norm regularisation method. Specifically, the proposed method can achieve better and more balanced results in terms of peak‐signal‐to‐noise ratio, convergence rate and restoration quality.

Understanding contents in social networks by inferring high-quality latent topics from short texts is a significant task in social analysis, which is challenging because social network contents are usually extremely short, noisy and full of informal vocabularies. Due to the lack of sufficient word co-occurrence instances, well-known topic modeling methods such as LDA and LSA cannot uncover high-quality topic structures. Existing research works seek to pool short texts from social networks into pseudo documents or utilize the explicit relations among these short texts such as hashtags in tweets to make classic topic modeling methods work. In this paper, we explore this problem by proposing a topic model for noisy short texts with multiple relations called MRTM (Multiple Relational Topic Modeling). MRTM exploits both explicit and implicit relations by introducing a document-attribute distribution and a two-step random sampling strategy. Extensive experiments, compared with the state-of-the-art topic modeling approaches, demonstrate that MRTM can alleviate the word co-occurrence sparsity and uncover high-quality latent topics from noisy short texts.

Pan-sharpening refers to the fusion of a low-resolution (LR) multispectral (MS) image and a high-resolution (HR) panchromatic (PAN) image to obtain an HR MS image (i.e., pan-sharpened MS image). From the point of view of variational complementary data fusion, it becomes an optimization problem with geometry and spectral preserving constraints. In this paper, a novel unified optimizing pan-sharpening model is proposed by integrating a data-generative fidelity term and a compound prior term, which incorporates both spatial fractional-order geometry and spectral-spatial low-rank priors. Specifically, the proposed model consists of three important ingredients: 1) data-generative fidelity term, which models the degradation relationship between the LR and HR MS images to enforce the geometry and spectral preserving constraints; 2) fractional-order total variation-based spatial fractional-order geometry prior term, which especially exploits the spatial fractional-order gradient feature consistence between the PAN and pan-sharpened MS images to transfer the spatial structure information of the PAN image into the pan-sharpened MS image; and 3) weighted nuclear norm-based spectral-spatial low-rank prior term, which exploits the nonlocal patches-based low-rank structural sparsity simultaneously in the pan-sharpened MS image and the LR MS image for further preserving image spatial structures and spectral information. Thus, the main novelty behind the proposed model is an optimizing mechanism by fully taking advantage of the spatial details and texture expressive power of the spatial fractional-order geometry prior as well as the spectral-spatial correlation preserving capacity of the low-rank prior. Finally, the proposed model can be implemented in an alternating direction method of multipliers framework, and thus, an efficient algorithm is presented. To verify the validity, the new proposed method is systematically compared with some state-of-the-art techniques using the Pleiades, GeoEye-1, QuickBird, and WorldView2 satellite data sets in the subjective, objective, and efficiency aspects. The results show that the proposed method performs better than the compared methods in terms of higher spatial and spectral qualities.

Recommender systems (RSs) provide an effective way of alleviating the information overload problem by selecting personalized items for different users. Latent factors-based collaborative filtering (CF) has become the popular approaches for RSs due to its accuracy and scalability. Recently, online social networks and user-generated content provide diverse sources for recommendation beyond ratings. Although social matrix factorization (Social MF) and topic matrix factorization (Topic MF) successfully exploit social relations and item reviews, respectively; both of them ignore some useful information. In this article, we investigate the effective data fusion by combining the aforementioned approaches. First, we propose a novel model MR3 to jointly model three sources of information (i.e., ratings, item reviews, and social relations) effectively for rating prediction by aligning the latent factors and hidden topics. Second, we incorporate the implicit feedback from ratings into the proposed model to enhance its capability and to demonstrate its flexibility. We achieve more accurate rating prediction on real-life datasets over various state-of-the-art methods. Furthermore, we measure the contribution from each of the three data sources and the impact of implicit feedback from ratings, followed by the sensitivity analysis of hyperparameters. Empirical studies demonstrate the effectiveness and efficacy of our proposed model and its extension.

The four papers in this special section examine the use of multimedia Big Data for leveraging big events such as natural disasters. Leveraging technological advances and using Big Data approaches for information management and processing in extreme events have attracted much research attention. Multimedia is increasingly becoming the most important and valuable source for insight and information on extreme events. It covers from everyone’s experiences to everything happening in the world. There will be lots of multimedia big data in extreme events—surveillance video, disaster images, social media, voice and video, to name a few, only if their volumes grow to the extent that the traditional multimedia processing and analysis systems may not handle effectively. As such, multimedia big data in extreme events are spurring on tremendous amounts of research and development of related technologies and applications. The papers in this issue is intended to provide a premier forum for researchers working on multimedia big data for extreme events to present their recent research results. It also provides an important opportunity for multidisciplinary studies connecting multimedia computing to disaster information management.

Crowdsourcing systems provide a cost effective and convenient way to collect labels, but they often fail to guarantee the quality of the labels. This paper proposes a novel framework that introduces noise correction techniques to further improve the quality of integrated labels that are inferred from the multiple noisy labels of objects. In the proposed general framework, information about the qualities of labelers estimated by a front-end ground truth inference algorithm is utilized to supervise subsequent label noise filtering and correction. The framework uses a novel algorithm termed adaptive voting noise correction (AVNC) to precisely identify and correct the potential noisy labels. After filtering out the instances with noisy labels, the remaining cleansed data set is used to create multiple weak classifiers, based on which a powerful ensemble classifier is induced to correct these noises. Experimental results on eight simulated data sets with different kinds of features and two real-world crowdsourcing data sets in different domains consistently show that: 1) the proposed framework can improve label quality regardless of inference algorithms, especially under the circumstance that each instance has a few repeated labels and 2) since the proposed AVNC algorithm considers both the number of and the probability of potential label noises, it outperforms the state-of-the-art noise correction algorithms.

A trouble ticket is an important information carrier in system maintenance, which records problem symptoms, the resolving process, and resolutions. A critical challenge for the ticket management system is how to quickly deal with trouble tickets and fix problems. Thousands of tickets, bouncing among multiple expert groups before being fixed, will consume limited system maintenance resources and may also violate the service level agreement. Thus, trouble tickets should be routed to the right expert group as quickly as possible in order to reduce the processing delay. In this paper, to address the challenge in ticket routing, we exploit three different routing models by mining the combination of problem descriptions and resolution sequences from the historical resolved tickets, and develop the corresponding routing recommendation algorithms to determine the next expert group to solve the problem. To evaluate the performance of routing recommendation algorithms, we conduct extensive experiments on a real ticket data set. The experimental results show that the proposed models and algorithm can effectively shorten the mean number of steps to resolve with a high ratio of the number of successfully resolved tickets to the total number of tickets, especially for the long routing sequences generated from manual assignments. These models and algorithms have the potential of being used in a ticket routing recommendation engine to greatly reduce human intervention in the routing process.

Many systems, such as distributed operating systems, complex networks, and high throughput web-based applications, are continuously generating large volume of event logs. These logs contain useful information to help system administrators to understand the system running status and to pinpoint the system failures. Generally, due to the scale and complexity of modern systems, the generated logs are beyond the analytic power of human beings. Therefore, it is imperative to develop a comprehensive log analysis system to support effective system management. Although a number of log mining techniques have been proposed to address specific log analysis use cases, few research and industrial efforts have been paid on providing integrated systems with an end-to-end solution to facilitate the log analysis routines.

In large scale and complex IT service environments, a problematic incident is logged as a ticket and contains the ticket summary (system status and problem description). The system administrators log the step-wise resolution description when such tickets are resolved. The repeating service events are most likely resolved by inferring similar historical tickets. With the availability of reasonably large ticket datasets, we can have an automated system to recommend the best matching resolution for a given ticket summary. In this paper, we first identify the challenges in real-world ticket analysis and develop an integrated framework to efficiently handle those challenges. The framework first quantifies the quality of ticket resolutions using a regression model built on carefully designed features. The tickets, along with their quality scores obtained from the resolution quality quantification, are then used to train a deep neural network ranking model that outputs the matching scores of ticket summary and resolution pairs. This ranking model allows us to leverage the resolution quality in historical tickets when recommending resolutions for an incoming incident ticket. In addition, the feature vectors derived from the deep neural ranking model can be effectively used in other ticket analysis tasks, such as ticket classification and clustering. The proposed framework is extensively evaluated with a large real-world dataset.

Mobile crowd sensing emerges as a new paradigm which takes advantage of the pervasive sensor-embedded smartphones to collect data efficiently. Many incentive mechanisms for mobile crowd sensing have been proposed. However, none of them takes into consideration the cooperative compatibility of users for multiple cooperative tasks. In this paper, we design truthful incentive mechanisms to minimize the social cost such that each of the cooperative tasks can be completed by a group of compatible users. We consider that the mobile crowd sensing is launched in an online community. We study two bid models and formulated the Social Optimization Compatible User Selection (SOCUS) problem for each model. We also define three compatibility models and use real-life relationships from social networks to model the compatibility relationships. We design two reverse auction based incentive mechanisms, MCT-M and MCT-S. Both of them consist of two steps: compatible user grouping and reverse auction. Through both rigorous theoretical analysis and extensive simulations, we demonstrate that the proposed mechanisms achieve computational efficiency, individual rationality and truthfulness. In addition, MCT-M can output the optimal solution.

Crowdsourcing has become an effective tool to utilize human intelligence to perform tasks that are challenging for machines. In the integrated crowdsourcing systems, the requesters are non- monopolistic and may show preferences over the workers. We are the first to design the incentive mechanisms, which consider the issue of stimulating the biased requesters in the competing crowdsourcing market. In this paper, we explore truthful task assignment mechanisms to maximize the total value of accomplished tasks for this new scenario. We present three models of crowdsourcing, which take the preferences of the requesters and the workload constraints of the workers into consideration. We design a task assignment mechanism, which follows the matching approach to solve the Valuation Maximizing Assignment (VMA) problem for each of the three models. Through both rigorous theoretical analyses and extensive simulations, we demonstrate that the proposed assignment mechanisms achieve computational efficiency, workload feasibility, preference (universal) truthfulness and constant approximation.

This paper proposes a novel general label aggregation method for both binary and multi-class labeling in crowdsourcing, namely Bi-Layer Clustering (BLC), which clusters two layers of features - the conceptual-level and the physical-level features - to infer true labels of instances. BLC first clusters the instances using the conceptual-level features extracted from their multiple noisy labels and then performs clustering again using the physical-level features. It can facilitate tracking the uncertainty changes of the instances, so that the integrated labels that are likely to be falsely inferred on the conceptual layer can be easily corrected using the estimated labels on the physical layer. Experimental results on two real-world crowdsourcing data sets show that BLC outperforms seven state-of-the-art methods.

Maximal automation of routine IT maintenance procedures is an ultimate goal of IT service management. System monitoring, an effective and reliable means for IT problem detection, generates monitoring ticket. In light of the ticket description, the underlying categories of the IT problem are determined, and the ticket is assigned to the corresponding processing teams for problem resolving. Automatic IT problem category determination acts as a critical part during the routine IT maintenance procedures. In practice, IT problem categories are naturally organized in a hierarchy by specialization. Utilizing the category hierarchy, this paper comes up with a hierarchical multi-label classification method to classify the monitoring tickets. In order to find the most effective classification, a novel contextual hierarchy (CH) loss is introduced in accordance with the problem hierarchy. Consequently, an arising optimization problem is solved by a new greedy algorithm named GLabel. Furthermore, as well as the ticket instance itself, the knowledge from the domain experts, which partially indicates some categories the given ticket may or may not belong to, can also be leveraged to guide the hierarchical multi-label classification. Accordingly, a multi-label inference with the domain expert knowledge is conducted on the basis of the given label hierarchy. The experiment demonstrates the great performance improvement by incorporating the domain knowledge during the hierarchical multi-label classification over the ticket data.

The frequency and intensity of natural disasters has increased significantly in recent decades, and this trend is expected to continue. Hence, understanding and predicting human evacuation behavior and mobility will play a vital role in planning effective humanitarian relief, disaster management, and long-term societal reconstruction. However, existing models are shallow models, and it is difficult to apply them for understanding the “deep knowledge” of human mobility. Therefore, in this study, we collect big and heterogeneous data (e.g., GPS records of 1.6 million users over 3 years, data on earthquakes that have occurred in Japan over 4 years, news report data, and transportation network data), and we build an intelligent system, namely, DeepMob, for understanding and predicting human evacuation behavior and mobility following different types of natural disasters. The key component of DeepMob is based on a deep learning architecture that aims to understand the basic laws that govern human behavior and mobility following natural disasters, from big and heterogeneous data. Furthermore, based on the deep learning model, DeepMob can accurately predict or simulate a person’s future evacuation behaviors or evacuation routes under different disaster conditions. Experimental results and validations demonstrate the efficiency and superior performance of our system, and suggest that human mobility following disasters may be predicted and simulated more easily than previously thought.

on Web Information Systems Engineering (WISE 2015) was held in Miami, Florida, USA, November 1-3, 2015.Building on the success of its predecessors, WISE 2015 continues to be a major international forum for researchers, professionals, and industrial practitioners to share their knowledge in the rapidly growing area of Web technologies, methodologies, and applications.The first WISE event took place in Hong Kong, China (2000).Then the trip continued to Kyoto, Japan (2001); Singapore (2002);

Large-scale time series data are prevalent across diverse application domains including system management, biomedical informatics, social networks, finance, etc. Temporal dependency discovery performs an essential part to identify the hidden interactions among the observed time series and helps to gain more insight into the behavior of the applications. However, the time-varying sparsity of the interactions among time series often poses a big challenge to temporal dependency discovery in practice. This paper formulates the temporal dependency problem with a novel Bayesian model allowing for both the sparsity and evolution of the hidden interactions among the observed time series. Taking advantage of the Bayesian modeling, an online inference method is proposed for time-varying temporal dependency discovery. Extensive empirical studies on both the synthetic and real application time series data are conducted to demonstrate the effectiveness and the efficiency of the proposed method.

In disaster management, people are interested in the development and the evolution of the disasters. If they intend to track the information of the disaster, they will be overwhelmed by the large number of disaster-related documents, microblogs, and news, etc. To support disaster management and minimize the loss during the disaster, it is necessary to efficiently and effectively collect, deliver, summarize, and analyze the disaster information, letting people in affected area quickly gain an overview of the disaster situation and improve their situational awareness.

There are two broad categories of citywide human mobility, routine, composed of daily or periodic travel, and rare, which occurs during events such as the Olympic Games or natural disasters. State-of-the-art studies have shown that routine mobility patterns can be modeled stochastically, while rare human mobility modeling, essential to a variety of urban computing scenarios, such as emergency management and traffic regulation, is a much more challenging and understudied problem. Instead of training a rare-event-specific human mobility model, which suffers from the particularity of the rare events, in this paper we provide a new insight into rare events and propose a novel algorithm, CityCoupling, which establishes an intercity spatial mapping that uses human mobility in one city as input and reproduces human mobility in another city. More intuitively, we attempt to answer the question "What if this rare event happened in another city?". To find the optimal intercity spatial mapping, we utilize an expectation-maximization algorithm to estimate a probabilistic geographical correspondence matrix by regarding intercity trajectory matching as latent variables. Thereafter, a Gibbs sampling-based multiple hidden Markov model generates simulated trajectories. We apply our approach to a large mobile phone GPS dataset in Japan and determine the spatial mapping between Tokyo and Osaka to transfer the human mobility at the Great Eastern Japan Earthquake in Tokyo, which was heavily affected, to simulate what might have occurred if Osaka had been struck by the earthquake. We conduct the evaluation by assuming that New Year's Countdown is a rare event that occurs simultaneously in both Tokyo and Osaka, and thus we quantitatively compare our simulation with the ground truth in Osaka.

Contextual multi-armed bandit problems have gained increasing popularity and attention in recent years due to their capability of leveraging contextual information to deliver online personalized recommendation services (e.g., online advertising and news article selection). To predict the reward of each arm given a particular context, existing relevant research studies for contextual multi-armed bandit problems often assume the existence of a fixed yet unknown reward mapping function. However, this assumption rarely holds in practice, since real-world problems often involve underlying processes that are dynamically evolving over time.

The evolution of the World Wide Web (WWW) and the smart-phone technologies have played a key role in the revolution of our daily life. The location-based social networks (LBSN) have emerged and facilitated the users to share the check-in information and multimedia contents. The Point of Interest (POI) recommendation system uses the check-in information to predict the most potential check-in locations. The different aspects of the check-in information, for instance, the geographical distance, the category, and the temporal popularity of a POI; and the temporal check-in trends, and the social (friendship) information of a user play a crucial role in an efficient recommendation. In this paper, we propose a fused recommendation model termed MAPS (Multi Aspect Personalized POI Recommender System) which will be the first in our knowledge to fuse the categorical, the temporal, the social and the spatial aspects in a single model. The major contribution of this paper are: (i) it realizes the problem as a graph of location nodes with constraints on the category and the distance aspects (i.e. the edge between two locations is constrained by a threshold distance and the category of the locations), (ii) it proposes a multi-aspect fused POI recommendation model, and (iii) it extensively evaluates the model with two real-world data sets.

Co-clustering aims to simultaneously cluster the objects and features to explore intercorrelated patterns. However, it is usually difficult to obtain good co-clustering results by just analyzing the object-feature correlation data due to the sparsity of the data and the noise. Meanwhile, most co-clustering algorithms cannot take the prior information into consideration and may produce unmeaningful results. Semi-supervised co-clustering aims to incorporate the known prior knowledge into the co-clustering algorithm. In this paper, a new technique named constraint co-projections for semi-supervised co-clustering (CPSSCC) is presented. Constraint co-projections can not only make use of two popular techniques including pairwise constraints and constraint projections, but also simultaneously perform the object constraint projections and feature constraint projections. The two popular techniques are illustrated for semi-supervised co-clustering when some objects and features are believed to be in the same cluster a priori. Furthermore, we also prove that the co-clustering problem can be formulated as a typical eigen-problem and can be efficiently solved with the selected eigenvectors. To the best of our knowledge, constraint co-projections is first stated in this paper and this is the first work on using CPSSCC. Extensive experiments on benchmark data sets demonstrate the effectiveness of the proposed method. This paper also shows that CPSSCC has some favorable features compared with previous related co-clustering algorithms.

The cloud service marketplace (CSM) is an exploratory project aiming to provide “an AppStore for Services.” It is an intelligent online marketplace that facilitates service discovery and acquisition for enterprise customers. Traditional service discovery and acquisition are time-consuming. In the era of OneClick Checkout and pay-as-you-go service plans, users expect services to be purchased online efficiently and conveniently. However, as services are complex and different from software apps, the currently prevailing App Store based on keyword search is inadequate for services. In CSM, exploring and configuring services are an iterative process. Customers provide their requirements in natural language and interact with the system through questioning and answering. Learning from the input, the system can incrementally clarify users’ intention, narrow down the candidate services, and profile the configuration information for the candidates at the same time. CSM’s back end is built around the Services Knowledge Graph (SKG) and leverages data mining technologies to enable the semantic understanding of customers’ requirements. To quantitatively assess the value of CSM, empirical evaluation on real and synthetic datasets and case studies are given to demonstrate the efficacy and effectiveness of the proposed system.

In recent years, IT service providers have rapidly achieved an automated service delivery model. Software monitoring systems are designed to actively collect and signal event occurrences and, when necessary, automatically generate incident tickets. Repeating events generate similar tickets, which in turn have a vast number of repeated problem resolutions likely to be found in earlier tickets. In this paper, we develop techniques to recommend appropriate resolution for incoming events by making use of similarities between the events and historical resolutions of similar events. Built on the traditional k-nearest neighbor algorithm (KNN), our proposed algorithms take into account false positives often generated by monitoring systems. An additional penalty is incorporated into the algorithms to control the number of misleading resolutions in the recommendation results. Moreover, as the effectiveness of the KNN heavily relies on the underlying similarity measurement, we proposed two other approaches to significantly improve our recommendation with respect to resolution relevance. One approach uses topic-level features to incorporate resolution information into the similarity measurement; the other uses metric learning to learn a more effective similarity measure. Extensive empirical evaluations on three ticket data sets demonstrate the effectiveness and efficiency of our proposed methods.

Rui Xia, Cheng Wang, Xin-Yu Dai, Tao Li. Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015.

Update summarization aims to generate brief summaries of recent documents to capture new information different from earlier documents. In this paper, we propose a new method to generate the sentence similarity graph using a novel similarity measure based on Helliger distance and apply semi-supervised learning on the sentence graph to select the sentences with maximum consistency and minimum redundancy to form the summaries. We use TAC 2011 data to evaluate our proposed method and compare it with existing baselines. The experimental results show the effectiveness of our proposed method.

In recent years, IT Service Providers have been rapidly introducing automation to their service delivery model. Driven by market pressure to reduce cost and maintain quality of services, they are looking for technologies that will allow rapid progress towards attainment of truly automated service delivery. Software monitoring systems are designed to actively collect and signal event occurrences and, when necessary, automatically generate incident tickets. Repeating events generate similar tickets, which in turn have a vast number of repeated problem resolutions likely to be found in earlier tickets. In our work, we develop techniques to recommend an appropriate resolution for incoming events by making use of similarities between the events and historical resolutions of similar events. The traditional KNN (K Nearest Neighbor) algorithm has been first applied to recommend resolutions for incoming tickets. Massive heterogeneous applications as well as various monitoring software are running on clients' servers to accomplish required tasks and to monitor system health via different metrics. It leads to generation of correlated tickets that have different symptom descriptions but similar resolutions. Furthermore, change of servers' environments can also induce similar situations in which ticket descriptions differ before and after change but could have similar resolutions. These correlated tickets cause performance degradation in ticket resolution recommendation. Therefore, we propose using SCL (structural corresponding learning) based feature adaptation to uncover feature mapping in different time intervals. Moreover, to put more insights into the periodic regularities existing in our ticket datasets, we apply our algorithm on tickets grouped by different time interval granularities. Extensive empirical evaluations on real-world ticket data sets demonstrate the effectiveness and efficiency of our proposed methods.

Popular online publishers produce huge amount of news articles every day, so it is important to summarize the most up-to-the-minute information to help users quickly know the progresses of their interested news events. In this paper, we develop NewsCubeSum, a novel personalized news summarization system utilizing OLAP and supervised sentence selection techniques to generate brief summaries delivering news updates in multiple dimensions (such as time, entity, and topic). An illustrative case study and experimental results on summarization performance comparisons are provided to show the effectiveness of NewsCubeSum.

Inventory forecasting aims to predict the demand of a specific item in the future and reserve the amount of item based on the forecasting results. An accurate and reliable inventory prediction can avoid product overstock and greatly reduce the maintenance cost. Inventory data is a kind of time series data, which has its own characteristics of large volume, long time span, wide covering range and poor regularity. The existing inventory forecasting methods usually only consider the contemporary data or similar goods historical data and achieve the prediction by calculating past average, which cannot capture the complex characteristics, such as long term trend, periodic, and special events. In this work, we treat inventory management as a data mining problem and propose a two-step dynamic prediction model, which first adopts six machine learning techniques and combines them with time series analysis methods to obtain a forecasting basis, then takes into account multiple factors of inventory to fulfill a dynamic inventory forecasting. Moreover, our proposed dynamic forecasting model, as one of core algorithms, is incorporated into an intelligent inventory management system. The experimental results and practical application demonstrate the effectiveness and efficiency of our proposed method.

In recent years, IT Service Providers have been rapidly transforming to an automated service delivery model. This is due to advances in technology and driven by the unrelenting market pressure to reduce cost and maintain quality. Tremendous progress has been made to date towards attainment of truly automated service delivery; that is, the ability to deliver the same service automatically using the same process with the same quality. However, automating Incident and Problem Management continuous to be a difficult problem, particularly due to the growing complexity of IT environments. Software monitoring systems are designed to actively collect and signal event occurrances and, when necessary, automatically generate incident tickets. Repeating events generate similar tickets, which in turn have a vast number of repeated problem resolutions likely to be found in earlier tickets. In this paper we find an appropriate resolution by making use of similarities between the events and previous resolutions of similar events. Traditional KNN (K Nearest Neighbor) algorithm has been used to recommend resolutions for incoming tickets. However, the effectiveness of recommendation heavily relies on the underlying similarity measure in KNN. In this paper, we significantly improve the similarity measure used in KNN by utilizing both the event and resolution information in historical tickets via a topic-level feature extraction using the LDA (Latent Dirichlet Allocation) model. In addition, when resolution categories are available, we propose to learn a more effective similarity measure using metric learning. Extensive empirical evaluations on three ticket data sets demonstrate the effectiveness and efficiency of our proposed methods.

Patent document retrieval, as a recall-orientated search task, does not allow missing relevant patent documents due to the great commercial value of patents and significant costs of processing a patent application or patent infringement case. Thus, it is important to retrieve all possible relevant documents rather than only a small subset of patents from the top ranked results. However, patents are often lengthy and rich in technical terms, and it often requires enormous human efforts to compare a given document with retrieved results. In this paper, we formulate the problem of comparing patent documents as a comparative summarization problem, and explore automatic strategies that generate comparative summaries to assist patent analysts in quickly reviewing any given patent document pairs. To this end, we present a novel approach, named PatentCom, which first extracts discriminative terms from each patent document, and then connects the dots on a term co-occurrence graph. In this way, we are able to comprehensively extract the gists of the two patent documents being compared, and meanwhile highlight their relationship in terms of commonalities and differences. Extensive quantitative analysis and case studies on real world patent documents demonstrate the effectiveness of our proposed approach.

Personalized recommendation services have gained increasing popularity and attention in recent years as most useful information can be accessed online in real-time. Most online recommender systems try to address the information needs of users by virtue of both user and content information. Despite extensive recent advances, the problem of personalized recommendation remains challenging for at least two reasons. First, the user and item repositories undergo frequent changes, which makes traditional recommendation algorithms ineffective. Second, the so-called cold-start problem is difficult to address, as the information for learning a recommendation model is limited for new items or new users. Both challenges are formed by the dilemma of exploration and exploitation. In this paper, we formulate personalized recommendation as a contextual bandit problem to solve the exploration/exploitation dilemma. Specifically in our work, we propose a parameter-free bandit strategy, which employs a principled resampling approach called online bootstrap, to derive the distribution of estimated models in an online manner. Under the paradigm of probability matching, the proposed algorithm randomly samples a model from the derived distribution for every recommendation. Extensive empirical experiments on two real-world collections of web data (including online advertising and news recommendation) demonstrate the effectiveness of the proposed algorithm in terms of the click-through rate. The experimental results also show that this proposed algorithm is robust in the cold-start situation, in which there is no sufficient data or knowledge to tune the hyper-parameters.

In recent years, crimes against children and cases of missing children have increased at a high rate. Therefore, there is an urgent need for safety support systems to prevent crimes against children or for antiloss, especially when parents are not with their children, such as to and from school. However, existing children’s tracking systems are not smart enough to provide the safety supports, as they simply locate the children’s positions without offering any notification to parents that their children may be in danger. In addition, there is limited research on children’s tracking and their antiloss. In this article, based on location histories, we introduce novel notions of children’s life patterns that capture their general lifestyles and regularities, and develop an intelligent data mining framework to learn the safe regions and safe routes of children on the cloud side. When the children may be in danger, their parents will receive automatic notifications from the cloud. We also propose an effective energy-efficient positioning scheme that leverages the location tracking accuracy of the children while keeping energy overhead low by using a hybrid global positioning system and a global system for mobile communications. To the best of our knowledge, this is the first attempt in applying data mining techniques to applications designed for children’s safety. Our proposed techniques have been incorporated into Soter , a children’s safeguard system that is used to provide cloud service for smart bracelets produced by Qihoo. The case studies on real smart bracelet users of Qihoo demonstrate the effectiveness of our proposed methods and Soter for children’s safety.

With the exponential growth of the usage of web map services, geo-data analysis has become more and more popular. This article develops an online spatial data analysis and visualization system, TerraFly GeoCloud, which helps end-users visualize and analyze spatial data and share the analysis results. Built on the TerraFly Geo spatial database, TerraFly GeoCloud is an extra layer running upon the TerraFly map and can efficiently support many different visualization functions and spatial data analysis models. Furthermore, users can create unique URLs to visualize and share the analysis results. TerraFly GeoCloud also enables the MapQL technology to customize map visualization using SQL-like statements. The system is available at http://terrafly.fiu.edu/GeoCloud/.

Social media has become increasingly prevalent in the last few years, not only enabling people to connect with each other by social links, but also providing platforms for people to share information and interact over diverse topics. Rich user-generated information, for example, users’ relationships and daily posts, are often available in most social media service websites. Given such information, a challenging problem is to provide reasonable user and community recommendation for a target user, and consequently, help the target user engage in the daily discussions and activities with his/her friends or like-minded people. In this article, we propose a unified framework of recommending users and communities that utilizes the information in social media. Given a user’s profile or a set of keywords as input, our framework is capable of recommending influential users and topic-cohesive interactive communities that are most relevant to the given user or keywords. With the proposed framework, users can find other individuals or communities sharing similar interests, and then have more interaction with these users or within the communities. We present a generative topic model to discover user-oriented and community-oriented topics simultaneously, which enables us to capture the exact topical interests of users, as well as the focuses of communities. Extensive experimental evaluation and case studies on a dataset collected from Twitter demonstrate the effectiveness of our proposed framework compared with other probabilistic-topic-model-based recommendation methods.

Bag-of-words (BOW) is now the most popular way to model text in statistical machine learning approaches in sentiment analysis. However, the performance of BOW sometimes remains limited due to some fundamental deficiencies in handling the polarity shift problem. We propose a model called dual sentiment analysis (DSA), to address this problem for sentiment classification. We first propose a novel data expansion technique by creating a sentiment-reversed review for each training and test review. On this basis, we propose a dual training algorithm to make use of original and reversed training reviews in pairs for learning a sentiment classifier, and a dual prediction algorithm to classify the test reviews by considering two sides of one review. We also extend the DSA framework from polarity (positive-negative) classification to 3-class (positive-negative-neutral) classification, by taking the neutral reviews into consideration. Finally, we develop a corpus-based method to construct a pseudo-antonym dictionary, which removes DSA's dependency on an external antonym dictionary for review reversion. We conduct a wide range of experiments including two tasks, nine datasets, two antonym dictionaries, three classification algorithms, and two types of features. The results demonstrate the effectiveness of DSA in supervised sentiment classification.

Inventory management refers to tracing inventory levels, orders and sales of a retailing business. In the current retailing market, a tremendous amount of data regarding stocked goods (items) in an inventory will be generated everyday. Due to the increasing volume of transaction data and the correlated relations of items, it is often a non-trivial task to efficiently and effectively manage stocked goods. In this demo, we present an intelligent system, called iMiner, to ease the management of enormous inventory data. We utilize distributed computing resources to process the huge volume of inventory data, and incorporate the latest advances of data mining technologies into the system to perform the tasks of inventory management, e.g., forecasting inventory, detecting abnormal items, and analyzing inventory aging. Since 2014, iMiner has been deployed as the major inventory management platform of ChangHong Electric Co., Ltd, one of the world's largest TV selling companies in China.

The fast growth of technologies has driven the advancement of our society. It is often necessary to quickly grasp the linkage between different technologies in order to better understand the technical trend. The availability of huge volumes of granted patent documents provides a reasonable basis for analyzing the relationships between technologies. In this paper, we propose a unified framework, named PatentDom, to identify important patents related to key techniques from a large number of patent documents. The framework integrates different types of patent information, including patent content, citations of patents, and temporal relations, and provides a concise yet comprehensive technology summary. The identified key patents enable a variety of patent-related analytical applications, e.g., outlining the technology evolution of a particular domain, tracing a given technique to prior technologies, and mining the technical connection of two given patent documents. Empirical analysis and extensive case studies on a collection of US patent documents demonstrate the efficacy of our proposed framework.

Given a short query audio clip, the goal of audio retrieval is to automatically fetch all similar clips from a given audio database. Different from traditional audio similarity which is mainly based on priori knowledge of objective reality, this paper proposes to use a more subjective method to m

Advanced manufacturing such as aerospace, semi-conductor, and flat display device often involves complex production processes, and generates large volume of production data. In general, the production data comes from products with different levels of quality, assembly line with complex flows and equipments, and processing craft with massive controlling parameters. The scale and complexity of data is beyond the analytic power of traditional IT infrastructures. To achieve better manufacturing performance, it is imperative to explore the underlying dependencies of the production data and exploit analytic insights to improve the production process. However, few research and industrial efforts have been reported on providing manufacturers with integrated data analytical solutions to reveal potentials and optimize the production process from data-driven perspectives.

Maximal automation of routine IT maintenance procedures is an ultimate goal of IT service management. System monitoring, an effective and reliable means for IT problem detection, generates monitoring tickets to be processed by system administrators. IT problems are naturally organized in a hierarchy by specialization. The problem hierarchy is used to help triage tickets to the processing team for problem resolving. In this paper, a hierarchical multi-label classification method is proposed to classify the monitoring tickets by utilizing the problem hierarchy. In order to find the most effective classification, a novel contextual hierarchy (CH) loss is introduced in accordance with the problem hierarchy. Consequently, an arising optimization problem is solved by a new greedy algorithm. An extensive empirical study over ticket data was conducted to validate the effectiveness and efficiency of our method.

The cold-start problem has attracted extensive attention among various online services that provide personalized recommendation. Many online vendors employ contextual bandit strategies to tackle the so-called exploration/exploitation dilemma rooted from the cold-start problem. However, due to high-dimensional user/item features and the underlying characteristics of bandit policies, it is often difficult for service providers to obtain and deploy an appropriate algorithm to achieve acceptable and robust economic profit.

Event summarization is an effective process that mines and organizes event patterns to represent the original events. It allows the analysts to quickly gain the general idea of the events. In recent years, several event summarization algorithms have been proposed, but they all focus on how to find out the optimal summarization results, and are designed for one-time analysis. As event summarization is a comprehensive analysis work, merely handling this problem with a single optimal algorithm is not enough. In the absence of an integrated summarization solution, we propose an extensible framework – META – to enable analysts to easily and selectively extract and summarize events from different views with different resolutions. In this framework, we store the original events in a carefully-designed data structure that enables an efficient storage and multiresolution analysis. On top of the data model, we define a summarization language that includes a set of atomic operators to manipulate the meta-data. Furthermore, we present 5 commonly used summarization tasks, and show that all these tasks can be easily expressed by the language. Experimental evaluation on both real and synthetic datasets demonstrates the efficiency and effectiveness of our framework.

The fast growth of technologies has driven the advancement of our society. It is often necessary to quickly grab the evolution of technologies in order to better understand the technology trend. The availability of huge volumes of granted patent documents provides a reasonable basis for analyzing technology evolution. In this paper, we propose a unified framework, named PatentLine, to generate a technology evolution tree for a given topic or a classification code related to granted patents. The framework integrates different types of patent information, including patent content, citations of patents, temporal relations, etc., and provides a concise yet comprehensive evolution summary. The generated summary enables a variety of patent-related analyses such as identifying relevant prior art and detecting technology gap. A case study on a collection of US patents demonstrates the efficacy of our proposed framework.

Semi-supervised clustering aims to incorporate the known prior knowledge into the clustering algorithm. Pairwise constraints and constraint projections are two popular techniques in semi-supervised clustering. However, both of them only consider the given constraints and do not consider the neighbors around the data points constrained by the constraints. This paper presents a new technique by utilizing the constrained pairwise data points and their neighbors, denoted as constraint neighborhood projections that requires fewer labeled data points (constraints) and can naturally deal with constraint conflicts. It includes two steps: 1) the constraint neighbors are chosen according to the pairwise constraints and a given radius so that the pairwise constraint relationships can be extended to their neighbors, and 2) the original data points are projected into a new low-dimensional space learned from the pairwise constraints and their neighbors. A CNP-Kmeans algorithm is developed based on the constraint neighborhood projections. Extensive experiments on University of California Irvine (UCI) datasets demonstrate the effectiveness of the proposed method. Our study also shows that constraint neighborhood projections (CNP) has some favorable features compared with the previous techniques.

Ensemble clustering, as an important extension of the clustering problem, refers to the problem of combining different (input) clusterings of a given dataset to generate a final (consensus) clustering that is a better fit in some sense than existing clusterings. Over the past few years, many ensemble clustering approaches have been developed. However, most of them are designed for partitional clustering methods, and few research efforts have been reported for ensemble hierarchical clustering methods. In this article, a hierarchical ensemble clustering framework that can naturally combine both partitional clustering and hierarchical clustering results is proposed. In addition, a novel method for learning the ultra-metric distance from the aggregated distance matrices and generating final hierarchical clustering with enhanced cluster separation is developed based on the ultra-metric distance for hierarchical clustering. We study three important problems: dendrogram description, dendrogram combination, and dendrogram selection. We develop two approaches for dendrogram selection based on tree distances, and we investigate various dendrogram distances for representing dendrograms. We provide a systematic empirical study of the ensemble hierarchical clustering problem. Experimental results demonstrate the effectiveness of our proposed approaches.

Modern scientific databases and web databases maintain large and heterogeneous data. These real-world databases contain hundreds or even thousands of relations and attributes. Traditional predefined query forms are not able to satisfy various ad-hoc queries from users on those databases. This paper proposes DQF, a novel database query form interface, which is able to dynamically generate query forms. The essence of DQF is to capture a user's preference and rank query form components, assisting him/her in making decisions. The generation of a query form is an iterative process and is guided by the user. At each iteration, the system automatically generates ranking lists of form components and the user then adds the desired form components into the query form. The ranking of form components is based on the captured user preference. A user can also fill the query form and submit queries to view the query result at each iteration. In this way, a query form could be dynamically refined until the user is satisfied with the query results. We utilize the expected F-measure for measuring the goodness of a query form. A probabilistic model is developed for estimating the goodness of a query form in DQF. Our experimental evaluation and user study demonstrate the effectiveness and efficiency of the system.

The five articles in this special section focus on data mining techniques and applications in the music industry. Music has been an important application area for data mining and machine learning techniques for many years. Music data mining is an interdisciplinary area that studies computational methods for understanding and delivering music data and is a topic of growing importance with large commercial relevance and substantial potential. The research area of music data mining has gradually evolved during this time period in order to address the challenge of effectively accessing and interacting with these increasing large collections of music and associated data such as styles, artists, lyrics and music reviews. The algorithms and systems developed frequently employ sophisticated and advanced data mining and machine learning techniques in their attempt to better capture the frequently elusive relevant music information.

Domain ontology, as a conceptual model, provides a meaningful framework for semantic representation of textual information. In this paper, we explore the feasibility of using the ontology in solving multi-document summarization problems in the domain of disaster management. We provide an empirical study of different approaches in which the ontology has been used for summarization tasks. Extensive experiments on a collection of press releases relevant to Hurricane Wilma in 2005 demonstrate that ontology-based multi-document summarization methods outperform other baselines in terms of the summary quality.

In this paper, we propose a framework of recommending users and communities in social media. Given a user's profile, our framework is capable of recommending influential users and topic-cohesive interactive communities that are most relevant to the given user. In our framework, we present a generative topic model to discover user-oriented and community-oriented topics simultaneously, which enables us to capture the exact topic interests of users, as well as the focuses of communities. Extensive evaluation on a data set obtained from Twitter has demonstrated the effectiveness of our proposed framework compared with other probabilistic topic model based recommendation methods.

With the exponential growth of the usage of web map services, the geo data analysis has become more and more popular. This paper develops an online Spatial Data Analysis System, TerraFly GeoCloud, which facilitates the end user to visualize and analyze spatial data, and to share the analysis results. Built on the TerraFly Geo spatial database, TerraFly GeoCloud is an extra layer running upon TerraFly map supporting many different visualization functions and spatial data analysis models. TerraFly GeoCloud also enables the MapQL technology to create maps using SQL-like statements. The TerraFly GeoCloud system is available at http://terrafly.fiu.edu/GeoCloud/.

Sequential data is prevalent in many scientific and commercial applications such as bioinformatics, system security and networking. Similarity search has been widely studied for symbolic and time series data in which each data object is a symbol or numeric value. Textual event sequences are sequences of events, where each object is a message describing an event. For example, system logs are typical textual event sequences and each event is a textual message recording internal system operations, statuses, configuration modifications or execution errors. Similar segments of an event sequence reveals similar system behaviors in the past which are helpful for system administrators to diagnose system problems. Existing search indexing for textual data only focus on unordered data. Substring matching methods are able to efficiently find matched segments over a sequence, however, their sequences are single values rather than texts. In this paper, we propose a method, suffix matrix, for efficiently searching similar segments over textual event sequences. It provides an integration of two disparate techniques: locality-sensitive hashing and suffix arrays. This method also supports the k-dissimilar segment search. A k-dissimilar segment is a segment that has at most k dissimilar events to the query sequence. By using random sequence mask proposed in this paper, this method can have a high probability to reach all k-dissimilar segments without increasing much search cost. We conduct experiments on real system log data and the experimental results show that our proposed method outperforms alternative methods using existing techniques.

Automatic system monitoring is an efficient and reliable mean for problem detection in enterprise IT infrastructures. The performance of monitoring systems depends on their configurations specified by the system administrators. In dynamic and large IT environments, the IT infrastructures are frequently changed to meet various business requirements, so the configurations may not be always consistent with the updated status. Misconfigurations can lead to false positive (false alarms) and false negative (missing alerts) for the system administrators. The false negatives can cause serious system faults. This paper presents an automatic approach for discovering the false negatives from incident tickets that are created by humans. The discovered results help the system administrators correct the misconfigurations and minimize the false negatives in future. This approach applies a text classification model for analyzing the descriptions of incident tickets and identifying the corresponding system issues. The domain knowledge for describing those issues can be incorporated to assist with this model. Experiments are conducted on real system incident tickets from a large enterprise IT infrastructure. The experimental results demonstrate the effectiveness of the proposed approach.

Online recruiting systems have gained immense attention in the wake of more and more job seekers searching jobs and enterprises finding candidates on the Internet. A critical problem in a recruiting system is how to maximally satisfy the desires of both job seekers and enterprises with reasonable recommendations or search results. In this paper, we investigate and compare various online recruiting systems from a product perspective. We then point out several key functions that help achieve a win-win situation between job seekers and enterprises for a successful recruiting system. Based on the observations and key functions, we design, implement and deploy a web-based application of recruiting system, named iHR, for Xiamen Talent Service Center. The system utilizes the latest advances in data mining and recommendation technologies to create a user-oriented service for a myriad of audience in job marketing community. Empirical evaluation and online user studies demonstrate the efficacy and effectiveness of our proposed system. Currently, iHR has been deployed at http://i.xmrc.com.cn/XMRCIntel.

The competitive business climate and the complexity of IT environments dictate efficient and cost-effective service delivery and support of IT services. These are largely achieved by automating routine maintenance procedures, including problem detection, determination and resolution. System monitoring provides an effective and reliable means for problem detection. Coupled with automated ticket creation, it ensures that a degradation of the vital signs, defined by acceptable thresholds or monitoring conditions, is flagged as a problem candidate and sent to supporting personnel as an incident ticket. This paper describes an integrated framework for minimizing false positive tickets and maximizing the monitoring coverage for system faults.

The advent of Big Data era drives data analysts from different domains to use data mining techniques for data analysis. However, performing data analysis in a specific domain is not trivial; it often requires complex task configuration, onerous integration of algorithms, and efficient execution in distributed environments.Few efforts have been paid on developing effective tools to facilitate data analysts in conducting complex data analysis tasks.

Personalized news recommender systems have gained increasing attention in recent years. Within a news reading community, the implicit correlations among news readers, news articles, topics and named entities, e.g., what types of named entities in articles are preferred by users, and why users like the articles, could be valuable for building an effective news recommender. In this paper, we propose a novel news personalization framework by mining such correlations. We use hypergraph to model various high-order relations among different objects in news data, and formulate news recommendation as a ranking problem on fine-grained hypergraphs. In addition, by transductive inference, our proposed algorithm is capable of effectively handling the so-called cold-start problem. Extensive experiments on a data set collected from various news websites have demonstrated the effectiveness of our proposed algorithm.

Techniques to efficiently discover, collect, organize, search, and disseminate real-time disaster information have become national priorities for efficient crisis management and disaster recovery tasks. We have developed techniques to facilitate information sharing and collaboration between both private and public sector participants for major disaster recovery planning and management. We have designed and implemented two parallel systems: a web-based prototype of a Business Continuity Information Network system and an All-Hazard Disaster Situation Browser system that run on mobile devices. Data mining and information retrieval techniques help impacted communities better understand the current disaster situation and how the community is recovering. Specifically, information extraction integrates the input data from different sources; report summarization techniques generate brief reviews from a large collection of reports at different granularities; probabilistic models support dynamically generating query forms and information dashboard based on user feedback; and community generation and user recommendation techniques are adapted to help users identify potential contacts for report sharing and community organization. User studies with more than 200 participants from EOC personnel and companies demonstrate that our systems are very useful to gain insights about the disaster situation and for making decisions.

Given a collection of document groups, a natural question is to identify the differences among these groups. Although traditional document summarization techniques can summarize the content of the document groups one by one, there exists a great necessity to generate a summary of the differences among the document groups. In this article, we study a novel problem of summarizing the differences between document groups. A discriminative sentence selection method is proposed to extract the most discriminative sentences that represent the specific characteristics of each document group. Experiments and case studies on real-world data sets demonstrate the effectiveness of our proposed method.

Users often find that their queries against a database return too many answers, many of them irrelevant. A common solution is to rank the query results. The effectiveness of a ranking function depends on how well it captures users' preferences. However, database systems often do not have the complete information about users' preferences and users' preferences are often heterogeneous (i.e., some preferences are static and common to all users while some are dynamic and diverse). Existing solutions do not address these two issues. In this paper, we propose a novel approach to address these shortcomings: 1) it addresses the heterogeneous issue by using skyline to capture users' static and common preferences and using users' current navigational behavior to capture users' dynamic and diverse preferences; 2) it addresses the incompleteness issue by using a machine learning technique to learn a ranking function based on training examples constructed from the above two types of information. Experimental results demonstrate the benefits of our approach.

Given a collection of web images with the corresponding textual descriptions, in this paper, we propose a novel cross-domain learning method to classify these web multimedia objects by transferring the correlation knowledge among different information sources. Here, the knowledge is extracted from unlabeled objects through unsupervised learning and applied to perform supervised classification tasks. To mine more meaningful correlation knowledge, instead of using commonly used visual words in the traditional bag-of-visual-words (BoW) model, we discover higher level visual components (words and phrases) to incorporate the spatial and semantic information into our image representation model, i.e., bag-of-visual-phrases (BoP). By combining the enriched visual components with the textual words, we calculate the frequently co-occurring pairs among them to construct a cross-domain correlated graph in which the correlation knowledge is mined. After that, we investigate two different strategies to apply such knowledge to enrich the feature space where the supervised classification is performed. By transferring such knowledge, our cross-domain transfer learning method can not only handle large scale web multimedia objects, but also deal with the situation that the textual descriptions of a small portion of web images are missing. Empirical experiments on two different datasets of web multimedia objects are conducted to demonstrate the efficacy and effectiveness of our proposed cross-domain transfer learning method.

The popularity of cloud service spurs the increasing demands of virtual resources to the service vendors. Along with the promising business opportunities, it also brings new technique challenges such as effective capacity planning and instant cloud resource provisioning. In this paper, we describe our research efforts on improving the service quality for the capacity planning and instant cloud resource provisioning problem. We first formulate both of the two problems as a generic cost-sensitive prediction problem. Then, considering the highly dynamic environment of cloud, we propose an asymmetric and heterogeneous measure to quantify the prediction error. Finally, we design an ensemble prediction mechanism by combining the prediction power of a set of prediction techniques based on the proposed measure. To evaluate the effectiveness of our proposed solution, we design and implement an integrated prototype system to help improve the service quality of the cloud. Our system considers many practical situations of the cloud system, and is able to dynamically adapt to the changing environment. A series of experiments on the IBM Smart Cloud Enterprise (SCE) trace data demonstrate that our method can significantly improve the service quality by reducing the resource provisioning time while maintaining a low cloud overhead.

This paper introduces a novel framework for generating pictorial storylines for given topics from text and image data on the Internet. Unlike traditional text summarization and timeline generation systems, the proposed framework combines text and image analysis and delivers a storyline containing textual, pictorial, and structural information to provide a sketch of the topic evolution. A key idea in the framework is the use of an approximate solution for the dominating set problem. Given a collection of topic-related objects consisting of images and their text descriptions, a weighted multi-view graph is first constructed to capture the contextual and temporal relationships among these objects. Then the objects are selected by solving the minimum-weighted connected dominating set problem defined on this graph. Comprehensive experiments on real-world data sets demonstrate the effectiveness of the proposed framework.

Reciprocal recommender systems refer to systems from which users can obtain recommendations of other individuals by satisfying preferences of both parties being involved. Different from the traditional user-item recommendation, reciprocal recommenders focus on the preferences of both parties simultaneously, as well as some special properties in terms of "reciprocal". In this paper, we propose MEET -- a generalized framework for reciprocal recommendation, in which we model the correlations of users as a bipartite graph that maintains both local and global "reciprocal" utilities. The local utility captures users' mutual preferences, whereas the global utility manages the overall quality of the entire reciprocal network. Extensive empirical evaluation on two real-world data sets (online dating and online recruiting) demonstrates the effectiveness of our proposed framework compared with existing recommendation algorithms. Our analysis also provides deep insights into the special aspects of reciprocal recommenders that differentiate them from user-item recommender systems.

Microblogging service has emerged to be a dominant web medium for billions of individuals sharing and spreading instant news and information, therefore monitoring the event evolution on microblog sphere is crucial for providing both better user experience and deeper understanding on real-time events. In this paper we explore the problem of generating storylines from microblogs for user input queries. This problem is challenging due to the sparse, dynamic and social nature of microblogs. Given a query of an ongoing event, we propose to sketch the real-time storyline of the event by a two-level solution. We first propose a language model with dynamic pseudo relevance feedback to obtain relevant tweets, and then generate storylines via graph optimization. Comprehensive experiments on Twitter data sets demonstrate the effectiveness of the proposed methods in each level and the overall framework.

A variety of news recommender systems based on different strategies have been proposed to provide news personalization services for online news readers. However, little research work has been reported on utilizing the implicit "social" factors (i.e., the potential influential experts in news reading community) among news readers to facilitate news personalization. In this paper, we investigate the feasibility of integrating content-based methods, collaborative filtering and information diffusion models by employing probabilistic matrix factorization techniques. We propose PRemiSE, a novel Personalized news Recommendation framework via implicit Social Experts, in which the opinions of potential influencers on virtual social networks extracted from implicit feedbacks are treated as auxiliary resources for recommendation. Empirical results demonstrate the efficacy and effectiveness of our method, particularly, on handling the so-called cold-start problem.

In this work, we have designed and developed a Multimedia-Aided Disaster Information Integration System (MADIS), which utilizes advanced data mining techniques to analyze situation reports and pictures as well as text captured in the field and automatically link the reports directly to relevant mult

Identifying functional modules from protein-protein interaction networks is an important and challenging task. This paper presents a new approach called PPIBM which is designed to integrate gene expression data analysis and clustering of protein-protein interactions. The proposed approach relies on a Bayesian model which uses as its base protein-protein interactions given as part of input. The proposed method is evaluated with standard measures and its performance is compared with the state-of-the-art network analysis methods. Experimental results on both real-world data and synthetic data demonstrate the effectiveness of the proposed approach.

Time lag is a key feature of hidden temporal dependencies within sequential data. In many real-world applications, time lag plays an essential role in interpreting the cause of discovered temporal dependencies. Traditional temporal mining methods either use a predefined time window to analyze the item sequence, or employ statistical techniques to simply derive the time dependencies among items. Such paradigms cannot effectively handle varied data with special properties, e.g., the interleaved temporal dependencies.

Cloud computing as a service promises many business benefits. The cost to pay is that it also faces many technique challenges. One of the challenges is to effectively manage cloud capacity in response to the increased demand changes in clouds, as computing customers now can provision and de-provision virtual machines more frequently. This paper studies cloud capacity prediction as a response to the challenge. We propose an integrated solution for intelligent cloud capacity estimation. In this solution, a novel measure is introduced to quantify and guide the prediction process. Then an ensemble method is utilized to predict the future provisioning/de-provisioning demands respectively. The cloud capacity is estimated using the active virtual machines and the future provisioning/de-provisioning demands altogether. Our proposed solution is simple and with low computational cost. The experiments on the IBM Smart Cloud Enterprise trace data shows our solution is effective.

Today's competitive business climate and the complexity of IT environments dictate efficient and cost effective service delivery and support of IT services. This is largely achieved through automating of routine maintenance procedures including problem detection, determination and resolution. System monitoring provides effective and reliable means for problem detection. Coupled with automated ticket creation, it ensures that a degradation of the vital signs, defined by acceptable thresholds or monitoring conditions, is flagged as a problem candidate and sent to supporting personnel as an incident ticket. This paper describes a novel methodology and a system for minimizing non-actionable tickets while preserving all tickets which require corrective action. Our proposed method defines monitoring conditions and the optimal corresponding delay times based on an off-line analysis of historical alerts and the matching incident tickets. Potential monitoring conditions are built on a set of predictive rules which are automatically generated by a rule-based learning algorithm with coverage, confidence and rule complexity criteria. These conditions and delay times are propagated as configurations into run-time monitoring systems.

With the rapid growth of social media websites, microblogging has become a popular way to spread instant news and events. Due to the dynamic and social nature of microblogs, extracting useful information from microblogs is more challenging than from the traditional news articles. In this paper we study the problem of summarizing the differences from microblogs. Given a collection of microblogs discussing an event/topic, we propose to generate a short summary delivering the differences among these microblogs, such as the different points of view for a news topic and the changes and evolution of an ongoing event.

To gain competitiveness and sustained growth in the 21st century, most businesses are on a mission to become more customer-centric. In order to succeed in this endeavor, it is crucial not only to synthesize and analyze the VOC (the VO ice of the C ustomer) data (i.e., the feedbacks or requirements raised by customers), but also to quickly turn these data into actionable knowledge. Although there are many technologies being developed in this complex problem space, most existing approaches in analyzing customer requests are ad hoc, time-consuming, error-prone, people-based processes which hardly scale well as the quantity of customer information explodes. This often results in the slow response to customer requests. In this article, in order to mine VOC to extract useful knowledge for the best product or service quality, we develop a hybrid framework that integrates domain knowledge with data-driven approaches to analyze the semi-structured customer requests. The framework consists of capturing functional features, discovering the overlap or correlation among the features, and identifying the evolving feature trend by using the knowledge transformation model. In addition, since understanding the relative importance of the individual customer request is very critical and has a direct impact on the effective prioritization in the development process, we develop a novel semantic enhanced link-based ranking (SELRank) algorithm for relatively rating/ranking both customer requests and products. The framework has been successfully applied on Xerox Office Group Feature Enhancement Requirements (XOG FER) datasets to analyze customer requests.

Over the last decade, document clustering, as one of the key tasks in information organization and navigation, has been widely studied. Many algorithms have been developed for addressing various challenges in document clustering and for improving clustering performance. However, relatively few research efforts have been reported on evaluating and understanding document clustering results. In this article, we present DClusterE , a comprehensive and effective framework for document clustering evaluation and understanding using information visualization. DClusterE integrates cluster validation with user interactions and offers rich visualization tools for users to examine document clustering results from multiple perspectives. In particular, through informative views including force-directed layout view, matrix view, and cluster view, DClusterE provides not only different aspects of document inter/intra-clustering structures, but also the corresponding relationship between clustering results and the ground truth. Additionally, DClusterE supports general user interactions such as zoom in/out, browsing, and interactive access of the documents at different levels. Two new techniques are proposed to implement DClusterE : (1) A novel multiplicative update algorithm (MUA) for matrix reordering to generate narrow-banded (or clustered) nonzero patterns from documents. Combined with coarse seriation, MUA is able to provide better visualization of the cluster structures. (2) A Mallows-distance-based algorithm for establishing the relationship between the clustering results and the ground truth, which serves as the basis for coloring schemes. Experiments and user studies are conducted to demonstrate the effectiveness and efficiency of DClusterE .

Given a collection of document groups, a natural question is to identify the differences among them. Although traditional document summarization techniques can summarize the content of the document groups one by one, there exists a great necessity to generate a summary of the differences among the document groups. In this article, we study a novel problem, that of summarizing the differences between document groups. A discriminative sentence selection method is proposed to extract the most discriminative sentences which represent the specific characteristics of each document group. Experiments and case studies on real-world data sets demonstrate the effectiveness of our proposed method.

In music information retrieval (MIR) an important research topic, which has attracted much attention recently, is the utilization of user-assigned tags, artist-related style, and mood labels, which can be extracted from music listening web sites, such as Last.fm (http://www.last.fm/) and All Music Guide (http://www.allmusic.com/). A fundamental research problem in the area is how to understand the relationships among artists/songs and these different pieces of information. Co-clustering is the problem of simultaneously clustering two types of data (e.g., documents and words, and webpages and urls). We can naturally bring this idea to the situation at hand and consider clustering artists and tags together, artists and styles together, or artists and mood labels together. Once such co-clustering has been successfully completed, one can identify co-existing clusters of artists and tags, styles, or mood labels (T/S/M). For simplicity, we use the acronym T/S/M to refer to tag(s), style(s), or mood(s) for the rest of the paper. When dealing with tags it is worth noticing that some tags are more specific versions of others. This naturally suggests that the tags could be organized in hierarchical clusters. Such hierarchical organizations exist for styles and mood labels, so we will consider hierarchical co-clustering of artists and T/S/M. In this paper, we systematically study the application of hierarchical co-clustering (HCC) methods for organizing the music data. There are two standard strategies for hierarchical clustering. One is the divisive strategy, in which we attempt to divide the input data set into smaller groups recursively, and the other is the agglomerative strategy, in which we attempt to combine initially individually separated data points into larger groups by finding the most closely related pair at each iteration. We will compare these two strategies against each other. We apply a previously known divisive hierarchical co-clustering method and a novel agglomerative hierarchical co-clustering. In addition, we demonstrate that these two methods have the capability of incorporating instance-level constraints to achieve better performance. We perform experiments to show that these two hierarchical co-clustering methods can be effectively deployed for organizing the music data and they present reasonable clustering performance comparing with the other clustering methods. A case study is also conducted to show that HCC provides us a new method to quantify the artist similarity.

Due to their damage to Internet security, malware and phishing website detection has been the Internet security topics that are of great interests. Compared with malware attacks, phishing website fraud is a relatively new Internet crime. However, they share some common properties: 1) both malware samples and phishing websites are created at a rate of thousands per day driven by economic benefits; and 2) phishing websites represented by the term frequencies of the webpage content share similar characteristics with malware samples represented by the instruction frequencies of the program. Over the past few years, many clustering techniques have been employed for automatic malware and phishing website detection. In these techniques, the detection process is generally divided into two steps: 1) feature extraction, where representative features are extracted to capture the characteristics of the file samples or the websites; and 2) categorization, where intelligent techniques are used to automatically group the file samples or websites into different classes based on computational analysis of the feature representations. However, few have been applied in real industry products. In this paper, we develop an automatic categorization system to automatically group phishing websites or malware samples using a cluster ensemble by aggregating the clustering solutions that are generated by different base clustering algorithms. We propose a principled cluster ensemble framework to combine individual clustering solutions that are based on the consensus partition, which can not only be applied for malware categorization, but also for phishing website clustering. In addition, the domain knowledge in the form of sample-level/website-level constraints can be naturally incorporated into the ensemble framework. The case studies on large and real daily phishing websites and malware collection from the Kingsoft Internet Security Laboratory demonstrate the effectiveness and efficiency of our proposed method.

Probabilistic Latent Semantic Analysis (PLSA) has been popularly used in document analysis. However, as it is currently formulated, PLSA strictly requires the number of word latent classes to be equal to the number of document latent classes. In this paper, we propose Bi-mixture PLSA, a new formulation of PLSA that allows the number of latent word classes to be different from the number of latent document classes. We further extend Bi-mixture PLSA to incorporate the sentence information, and propose Bi-mixture PLSA with sentence bases (Bi-PLSAS) to simultaneously cluster and summarize the documents utilizing the mutual influence of the document clustering and summarization procedures. Experiments on real-world datasets demonstrate the effectiveness of our proposed methods.

Event mining is a useful way to understand computer system behaviors. The focus of recent works on event mining has been shifted to event summarization from discovering frequent patterns. Event summarization seeks to provide a comprehensible explanation of the event sequence on certain aspects. Previous methods have several limitations such as ignoring temporal information, generating the same set of boundaries for all event patterns, and providing a summary which is difficult for human to understand. In this paper, we propose a novel framework called natural event summarization that summarizes an event sequence using inter-arrival histograms to capture the temporal relationship among events. Our framework uses the minimum description length principle to guide the process in order to balance between accuracy and brevity. Also, we use multi-resolution analysis for pruning the problem space. We demonstrate how the principles can be applied to generate summaries with periodic patterns and correlation patterns in the framework. Experimental results on synthetic and real data show our method is capable of producing usable event summary, robust to noises, and scalable.

Modern computing systems generate large amounts of log data. System administrators or domain experts utilize the log data to understand and optimize system behaviors. Most system logs are raw textual and unstructured. One main fundamental challenge in automated log analysis is the generation of system events from raw textual logs. Log messages are relatively short text messages but may have a large vocabulary, which often result in poor performance when applying traditional text clustering techniques to the log data. Other related methods have various limitations and only work well for some particular system logs. In this paper, we propose a message signature based algorithm logSig to generate system events from textual log messages. By searching the most representative message signatures, logSig categorizes log messages into a set of event types. logSig can handle various types of log data, and is able to incorporate human's domain knowledge to achieve a high performance. We conduct experiments on five real system log data. Experiments show that logSig outperforms other alternative algorithms in terms of the overall performance.

Multimedia information plays an increasingly important role in humans daily activities. Given a set of web multimedia objects (images with corresponding texts), a challenging problem is how to group these images into several clusters using the available information. Previous researches focus on either adopting individual information, or simply combining image and text information together for clustering. In this paper, we propose a novel approach (Dynamic Weighted Clustering) to separate images under the "supervision" of text descriptions, Also, we provide a comparative experimental investigation on utilizing text and image information to tackle web image clustering. Empirical experiments on a manually collected web multimedia object (related to the events after disasters) dataset are conducted to demonstrate the efficacy of our proposed method.

The promise of cloud computing is to provide computing resources instantly whenever they are needed. The state-of-art virtual machine (VM) provisioning technology can provision a VM in tens of minutes. This latency is unacceptable for jobs that need to scale out during computation. To truly enable on-the-fly scaling, new VM needs to be ready in seconds upon request. In this paper, We present an online temporal data mining system called ASAP, to model and predict the cloud VM demands. ASAP aims to extract high level characteristics from VM provisioning request stream and notify the provisioning system to prepare VMs in advance. For quantification issue, we propose Cloud Prediction Cost to encodes the cost and constraints of the cloud and guide the training of prediction algorithms. Moreover, we utilize a two-level ensemble method to capture the characteristics of the high transient demands time series. Experimental results using historical data from an IBM cloud in operation demonstrate that ASAP significantly improves the cloud service quality and provides possibility for on-the-fly provisioning.

In this paper, we explore how to use ranking SVM to train the feature weights for query-focused multi-document summarization. To apply a supervised learning method to sentence extraction in multi-document summarization, we need to derive the sentence labels for training corpus from the existing human labeling data in form of. However, this process is not trivial, because the human summaries are abstractive, and do not necessarily well match the sentences in the documents. In this paper, we try to address the above problem from the following two aspects. First, we make use of sentence-to-sentence relationships to better estimate the probability of a sentence in the document set to be a summary sentence. Second, to make the derived training data less sensitive, we adopt a cost sensitive loss in the ranking SVM's objective function. The experimental results demonstrate the effectiveness of our proposed method.

Semi-supervised clustering (i.e., clustering with knowledge-based constraints) has emerged as an important variant of the traditional clustering paradigms. However, most existing semi-supervised clustering algorithms are designed for partitional clustering methods and few research efforts have been reported on semi-supervised hierarchical clustering methods. In addition, current semi-supervised clustering methods have been focused on the use of background information in the form of instance level must-link and cannot-link constraints, which are not suitable for hierarchical clustering where data objects are linked over different hierarchy levels. In this paper, we propose a novel semi-supervised hierarchical clustering framework based on ultra-metric dendrogram distance. The proposed framework is able to incorporate triple-wise relative constraints. We establish the connection between hierarchical clustering and ultra-metric transformation of dissimilarity matrix and propose two techniques (the constrained optimization technique and the transitive dissimilarity based technique) for semi-supervised hierarchical clustering. Experimental results demonstrate the effectiveness and the efficiency of our proposed methods.

Due to their damages to Internet security, malware (such as virus, worms, trojans, spyware, backdoors, and rootkits) detection has caught the attention not only of anti-malware industry but also of researchers for decades. Resting on the analysis of file contents extracted from the file samples, like Application Programming Interface (API) calls, instruction sequences, and binary strings, data mining methods such as Naive Bayes and Support Vector Machines have been used for malware detection. However, besides file contents, relations among file samples, such as a "Downloader" is always associated with many Trojans, can provide invaluable information about the properties of file samples. In this paper, we study how file relations can be used to improve malware detection results and develop a file verdict system (named "Valkyrie") building on a semi-parametric classifier model to combine file content and file relations together for malware detection. To the best of our knowledge, this is the first work of using both file content and file relations for malware detection. A comprehensive experimental study on a large collection of PE files obtained from the clients of anti-malware products of Comodo Security Solutions Incorporation is performed to compare various malware detection approaches. Promising experimental results demonstrate that the accuracy and efficiency of our Valkyrie system outperform other popular anti-malware software tools such as Kaspersky AntiVirus and McAfee VirusScan, as well as other alternative data mining based detection systems.

The improvement of Crisis Management and Disaster Recovery techniques are national priorities in the wake of man-made and nature inflicted calamities of the last decade. Our prior work has demonstrated that the efficiency of sharing and managing information plays an important role in business recovery efforts after disaster event. With the proliferation of smart phones and wireless tablets, professionals who have an operational responsibility in disaster situations are relying on such devices to maintain communication. Further, with the rise of social media, technology savvy consumers are also using these devices extensively for situational updates. In this paper, we address several critical tasks which can facilitate information sharing and collaboration between both private and public sector participants for major disaster recovery planning and management. We design and implement an All-Hazard Disaster Situation Browser (ADSB) system that runs on Apple's mobile operating system (iOS) and iPhone and iPad mobile devices. Our proposed techniques create a collaborative solution on a mobile platform using advanced data mining and information retrieval techniques for disaster preparedness and recovery that helps impacted communities better understand the current disaster situation and how the community is recovering. Specifically, hierarchical summarization techniques are used to generate brief reviews from a large collection of reports at different granularities; probabilistic models are proposed to dynamically generate query forms based on user's feedback; and recommendation techniques are adapted to help users identify potential contacts for report sharing and community organization. Furthermore, the developed techniques are designed to be all-hazard capable so that they can be used in earthquake, terrorism, or other unanticipated disaster situations.

In this paper, we initially provide an experimental study on the evolution of user interests in real-world news recommender systems, and then propose a novel recommendation approach, in which the long-term and short-term reading preferences of users are seamlessly integrated when recommending news items. Given a hierarchy of newly-published news articles, news groups that the user might prefer are differentiated using the long-term profile, and then in each selected news group, a list of news items are chosen based on the short-term user profile. Extensive empirical experiments on a collection of news articles obtained from various popular news websites demonstrate the efficacy of our method.

Consensus clustering has emerged as an important extension of the classical clustering problem. Given a set of input clusterings of a given dataset, consensus clustering aims to find a single final clustering which is a better fit in some sense than the existing clusterings. There is a significant drawback in generating a single consensus clustering since different input clusterings could differ significantly. In this paper, we develop a new framework, called Multiple Consensus Clustering (MCC), to explore multiple clustering views of a given dataset from a set of input clusterings. Instead of generating a single consensus, MCC organizes the different input clusterings into a hierarchical tree structure and allows for interactive exploration of multiple clustering solutions. A dynamic programming algorithm is proposed to obtain a flat partition from the hierarchical tree using the modularity measure. Multiple consensuses are finally obtained by applying consensus clustering algorithms to each cluster of the partition. Extensive experimental results on 11 real world data sets and a case study on a Protein-Protein Interaction (PPI) data set demonstrate the effectiveness of our proposed method.

Multi-document summarization aims to distill the most representative information from a set of documents to generate a summary. Given a set of documents as input, most of existing multi-document summarization approaches utilize different sentence selection techniques to extract a set of sentences from the document set as the summary. The submodularity hidden in textual-unit similarity motivates us to incorporate this property into our solution to multi-document summarization tasks. In this poster, we propose a new principled and versatile framework for different multi-document summarization tasks using the submodular function [8].

Recommending news articles has become a promising research direction as the Internet provides fast access to real-time information from multiple sources around the world. Traditional news recommendation systems strive to adapt their services to individual users by virtue of both user and news content information. However, the latent relationships among different news items, and the special properties of new articles, such as short shelf lives and value of immediacy, render the previous approaches inefficient.

Multi-way data or tensors are generalizations of matrices. Clustering multi-way data is a very important research topic due to the intrinsic rich structures in real-world datasets. Despite significant progress made on subspace clustering for two-way

Document understanding techniques such as document clustering and multidocument summarization have been receiving much attention recently. Current document clustering methods usually represent the given collection of documents as a document-term matrix and then conduct the clustering process. Although many of these clustering methods can group the documents effectively, it is still hard for people to capture the meaning of the documents since there is no satisfactory interpretation for each document cluster. A straightforward solution is to first cluster the documents and then summarize each document cluster using summarization methods. However, most of the current summarization methods are solely based on the sentence-term matrix and ignore the context dependence of the sentences. As a result, the generated summaries lack guidance from the document clusters. In this article, we propose a new language model to simultaneously cluster and summarize documents by making use of both the document-term and sentence-term matrices. By utilizing the mutual influence of document clustering and summarization, our method makes; (1) a better document clustering method with more meaningful interpretation; and (2) an effective document summarization method with guidance from document clustering. Experimental results on various document datasets show the effectiveness of our proposed method and the high interpretability of the generated summaries.

In multiple instance learning problems, patterns are often given as bags and each bag consists of some instances. Most of existing research in the area focuses on multiple instance classification and multiple instance regression, while very limited work has been conducted for multiple instance clustering (MIC). This paper formulates a novel framework, maximum margin multiple instance clustering (M(3)IC), for MIC. However, it is impractical to directly solve the optimization problem of M(3)IC. Therefore, M(3)IC is relaxed in this paper to enable an efficient optimization solution with a combination of the constrained concave-convex procedure and the cutting plane method. Furthermore, this paper presents some important properties of the proposed method and discusses the relationship between the proposed method and some other related ones. An extensive set of empirical results are shown to demonstrate the advantages of the proposed method against existing research for both effectiveness and efficiency.

Due to the importance of high-quality customer service, many companies use intelligent helpdesk systems (e.g., case-based systems) to improve customer service quality. However, these systems face two challenges: 1) Case retrieval measures: most case-based systems use traditional keyword-matching-based ranking schemes for case retrieval and have difficulty to capture the semantic meanings of cases and 2) result representation: most case-based systems return a list of past cases ranked by their relevance to a new request, and customers have to go through the list and examine the cases one by one to identify their desired cases. To address these challenges, we develop iHelp, an intelligent online helpdesk system, to automatically find problem-solution patterns from the past customer-representative interactions. When a new customer request arrives, iHelp searches and ranks the past cases based on their semantic relevance to the request, groups the relevant cases into different clusters using a mixture language model and symmetric matrix factorization, and summarizes each case cluster to generate recommended solutions. Case and user studies have been conducted to show the full functionality and the effectiveness of iHelp.

Query-based multi-document summarization aims to create a short summary given a collection of documents and a query. Most of the existing methods treat the query as one single sentence and rank the sentences in the documents based on their similarities with the query sentence. However, these methods lack of intensive analysis on the given query which typically consist of several topic aspects. In this paper, we propose a topic aspect extraction method to discover the aspect words and sentences contained in the query narrative texts and the input documents, and then incorporate these aspect words and sentences into a cross propagation model based on the sentence-term bipartite graph for document summarization. Experiments on DUC benchmark data show the effectiveness of our proposed approach on the topic-driven document summarization task.

Document summarization has become a hot topic in recent years. However, most of existing summarization methods work on a batch of documents and do not consider that documents may arrive in a sequence and the corresponding summaries need to be updated in real time. In this paper, we propose a new summarization method based on an incremental hierarchical clustering framework to update summaries as soon as a new document arrives. Extensive experimental results demonstrate the effectiveness and efficiency of our proposed method.

Modern computing systems are instrumented to generate huge amounts of system logs and these data can be utilized for understanding and complex system behaviors. One main fundamental challenge in automated log analysis is the generation of system events from raw textual logs. Recent works apply clustering techniques to translate the raw log messages into system events using only the word/term information. In this paper, we first illustrate the drawbacks of existing techniques for event generation from system logs. We then propose Log Tree, a novel and algorithm-independent framework for events generation from raw system log messages. Log Tree utilizes the format and structural information of the raw logs in the clustering process to generate system events with better accuracy. In addition, an indexing data structure, Message Segment Table, is proposed in Log Tree to significantly improve the efficiency of events creation. Extensive experiments on real system logs demonstrate the effectiveness and efficiency of Log Tree.

Keyword (Feature) selection enhances and improves many Information Retrieval (IR) tasks such as document categorization, automatic topic discovery, etc. The problem of keyword selection is usually solved using supervised algorithms. In this paper, we propose an unsupervised approach that combines keyword selection and document clustering (topic discovery) together. The proposed approach extends non-negative matrix factorization (NMF) by incorporating a weight matrix to indicate the importance of the keywords. The proposed approach is further extended to a weighted version in which each document is also assigned a weight to assess its importance in the cluster. This work considers both theoretical and empirical weighted feature subset selection for NMF and draws the connection between unsupervised feature selection and data clustering. We apply our proposed approaches to various document understanding tasks including document clustering, summarization, and visualization. Experimental results demonstrate the effectiveness of our approach for these tasks.

Ensemble clustering has emerged as an important elaboration of the classical clustering problems. Ensemble clustering refers to the situation in which a number of different (input) clusterings have been obtained for a particular dataset and it is desired to find a single (consensus) clustering which is a better fit in some sense than the existing clusterings. Many approaches have been developed to solve ensemble clustering problems over the last few years. However, most of these ensemble techniques are designed for partitional clustering methods. Few research efforts have been reported for ensemble hierarchical clustering methods. In this paper, we propose a hierarchical ensemble clustering framework which can naturally combine both partitional clustering and hierarchical clustering results. We notice the importance of ultra-metric distance for hierarchical clustering and propose a novel method for learning the ultra-metric distance from the aggregated distance matrices and generating final hierarchical clustering with enhanced cluster separation. Experimental results demonstrate the effectiveness of our proposed approaches.

In this paper, resting on the analysis of instruction frequency and function-based instruction sequences, we develop an Automatic Malware Categorization System (AMCS) for automatically grouping malware samples into families that share some common characteristics using a cluster ensemble by aggregating the clustering solutions generated by different base clustering algorithms. We propose a principled cluster ensemble framework for combining individual clustering solutions based on the consensus partition. The domain knowledge in the form of sample-level constraints can be naturally incorporated in the ensemble framework. In addition, to account for the characteristics of feature representations, we propose a hybrid hierarchical clustering algorithm which combines the merits of hierarchical clustering and k-medoids algorithms and a weighted subspace K-medoids algorithm to generate base clusterings. The categorization results of our AMCS system can be used to generate signatures for malware families that are useful for malware detection. The case studies on large and real daily malware collection from Kingsoft Anti-Virus Lab demonstrate the effectiveness and efficiency of our AMCS system.

Crisis Management and Disaster Recovery have gained immense importance in the wake of recent man and nature inflicted calamities. A critical problem in a crisis situation is how to efficiently discover, collect, organize, search and disseminate real-time disaster information. In this paper, we address several key problems which inhibit better information sharing and collaboration between both private and public sector participants for disaster management and recovery. We design and implement a web based prototype implementation of a Business Continuity Information Network (BCIN) system utilizing the latest advances in data mining technologies to create a user-friendly, Internet-based, information-rich service and acting as a vital part of a company's business continuity process. Specifically, information extraction is used to integrate the input data from different sources; the content recommendation engine and the report summarization module provide users personalized and brief views of the disaster information; the community generation module develops spatial clustering techniques to help users build dynamic community in disasters. Currently, BCIN has been exercised at Miami-Dade County Emergency Management.

With the explosion of user-generated web2.0 content in the form of blogs, wikis and discussion forums, the Internet has rapidly become a massive dynamic repository of public opinion on an unbounded range of topics. A key enabler of opinion extraction and summarization is sentiment classification: the task of automatically identifying whether a given piece of text expresses positive or negative opinion towards a topic of interest. Building high-quality sentiment classifiers using standard text categorization methods is challenging due to the lack of labeled data in a target domain. In this paper, we consider the problem of cross-domain sentiment analysis: can one, for instance, download rated movie reviews from rottentomatoes.com or IMBD discussion forums, learn linguistic expressions and sentiment-laden terms that generally characterize opinionated commentary and then successfully transfer this knowledge to the target domain, thereby building high-quality sentiment models without manual effort? We outline a novel sentiment transfer mechanism based on constrained non-negative matrix tri-factorizations of term-document matrices in the source and target domains. The constrained matrix factorization framework naturally incorporates document labels via a least squares penalty incurred by a certain linear model and enables direct and explicit knowledge transfer across different domains. We obtain promising empirical results with this approach.

In this poster, we develop a novel method, called HCC, for hierarchical co-clustering. HCC brings together two interrelated but distinct themes from clustering: hierarchical clustering and co-clustering.

In this poster, we propose a novel document summarization approach named Ontology-enriched Multi-Document Summarization(OMS) for utilizing background knowledge to improve summarization results. OMS first maps the sentences of input documents onto an ontology, then links the given query to a specific node in the ontology, and finally extracts the summary from the sentences in the subtree rooted at the query node. By using the domain-related ontology, OMS can better capture the semantic relevance between the query and the sentences, and thus lead to better summarization results. As a byproduct, the final summary generated by OMS can be represented as a tree showing the hierarchical relationships of the extracted sentences. Evaluation results on the collection of press releases by Miami-Dade County Department of Emergency Management during Hurricane Wilma in 2005 demonstrate the efficacy of OMS.

In this paper, we propose feature subset non-negative matrix factorization (NMF), which is an unsupervised approach to simultaneously cluster data points and select important features. We apply our proposed approach to various document understanding tasks including document clustering, summarization, and visualization. Experimental results demonstrate the effectiveness of our approach for these tasks.

Given a collection of documents, various multi-document summarization methods have been proposed to generate a short summary. However, few studies have been reported on aggregating different summarization methods to possibly generate better summarization results. We propose a weighted consensus summarization method to combine the results from single summarization systems. Experimental results on DUC2004 data sets demonstrate the performance improvement by aggregating multiple summarization systems, and our proposed weighted consensus summarization method outperforms other combination methods.

Given a collection of documents, most of existing multidocument summarization methods automatically generate a static summary for all the users. However, different users may have different opinions on the documents, thus there is a necessity for improving users' interactions in the summarization process. In this paper, we propose an interactive document summarization system using information visualization techniques.

We present several new variations on the theme of nonnegative matrix factorization (NMF). Considering factorizations of the form X=FG(T), we focus on algorithms in which G is restricted to containing nonnegative entries, but allowing the data matrix X to have mixed signs, thus extending the applicable range of NMF methods. We also consider algorithms in which the basis vectors of F are constrained to be convex combinations of the data points. This is used for a kernel extension of NMF. We provide algorithms for computing these new factorizations and we provide supporting theoretical analysis. We also analyze the relationships between our algorithms and clustering algorithms, and consider the implications for sparseness of solutions. Finally, we present experimental results that explore the properties of these new methods.

Gene expression data usually contain a large number of genes but a small number of samples. Feature selection for gene expression data aims at finding a set of genes that best discriminate biological samples of different types. Using machine learning techniques, traditional gene selection based on empirical mutual information suffers the data sparseness issue due to the small number of samples. To overcome the sparseness issue, we propose a model-based approach to estimate the entropy of class variables on the model, instead of on the data themselves. Here, we use multivariate normal distributions to fit the data, because multivariate normal distributions have maximum entropy among all real-valued distributions with a specified mean and standard deviation and are widely used to approximate various distributions. Given that the data follow a multivariate normal distribution, since the conditional distribution of class variables given the selected features is a normal distribution, its entropy can be computed with the log-determinant of its covariance matrix. Because of the large number of genes, the computation of all possible log-determinants is not efficient. We propose several algorithms to largely reduce the computational cost. The experiments on seven gene data sets and the comparison with other five approaches show the accuracy of the multivariate Gaussian generative model for feature selection, and the efficiency of our algorithms.

With advancement in science and technology, computing systems are becoming increasingly more complex with a growing number of heterogeneous software and hardware components. They are thus becoming more difficult to monitor, manage, and maintain. Traditional approaches to system management have been largely based on domain experts through a knowledge acquisition solution that translates domain knowledge into operating rules and policies. This process has been well known as cumbersome, labor intensive, and error prone. In addition, traditional approaches for system management are difficult to keep up with the rapidly changing environments. There is a pressing need for automatic and efficient approaches to monitor and manage complex computing systems. In this paper, we propose an integrated data-driven framework for computing system management by acquiring the needed knowledge automatically from a large amount of historical log data. Specifically, we apply text mining techniques to automatically categorize the log messages into a set of canonical categories, incorporate temporal information to improve categorization performance, develop temporal mining techniques to discover the relationships between different events, and take a novel approach called event summarization to provide a concise interpretation of the temporal patterns.

Malware is software designed to infiltrate or damage a computer system without the owner's informed consent (e.g., viruses, backdoors, spyware, trojans, and worms). Nowadays, numerous attacks made by the malware pose a major security threat to computer users. Unfortunately, along with the development of the malware writing techniques, the number of file samples that need to be analyzed, named "gray list," on a daily basis is constantly increasing. In order to help our virus analysts, quickly and efficiently pick out the malicious executables from the "gray list," an automatic and robust tool to analyze and classify the file samples is needed. In our previous work, we have developed an intelligent malware detection system (IMDS) by adopting associative classification method based on the analysis of application programming interface (API) execution calls. Despite its good performance in malware detection, IMDS still faces the following two challenges: (1) handling the large set of the generated rules to build the classifier; and (2) finding effective rules for classifying new file samples. In this paper, we first systematically evaluate the effects of the postprocessing techniques (e.g., rule pruning, rule ranking, and rule selection) of associative classification in malware detection, and then, propose an effective way, i.e., CIDCPF, to detect the malware from the "gray list." To the best of our knowledge, this is the first effort on using postprocessing techniques of associative classification in malware detection. CIDCPF adapts the postprocessing techniques as follows: first applying Chi-square testing and Insignificant rule pruning followed by using Database coverage based on the Chi-square measure rule ranking mechanism and Pessimistic error estimation, and finally performing prediction by selecting the best First rule. We have incorporated the CIDCPF method into our existing IMDS system, and we call the new system as CIMDS system. Case studies are performed on the large collection of file samples obtained from the Antivirus Laboratory at Kingsoft Corporation and promising experimental results demonstrate that the efficiency and ability of detecting malware from the "gray list" of our CIMDS system outperform popular antivirus software tools, such as McAfee VirusScan and Norton Antivirus, as well as previous data-mining-based detection systems, which employed Naive Bayes, support vector machine, and decision tree techniques. In particular, our CIMDS system can greatly reduce the number of generated rules, which makes it easy for our virus analysts to identify the useful ones.

Given a collection of document groups, a quick question is what are the differences in these groups. In this paper, we study a novel problem of summarizing the differences between document groups. A discriminative sentence selection method is proposed to extract the most discriminative sentences which represent the specific characteristics of each document group. Experiments on real world data sets demonstrate the effectiveness of our proposed method.

Linear discriminant analysis (LDA) is a popular feature extraction method that has aroused considerable interests in computer vision and pattern recognition fields. The projection vectors of LDA is usually achieved by maximizing the between-class scatter and simultaneously minimizing the within-class scatter of the data set. However, in practice, there is usually a lack of sufficient labeled data, which makes the estimated projection direction inaccurate. To address the above limitations, in this paper, we propose a novel semi-supervised discriminant analysis approach. Unlike traditional graph based methods, our algorithm incorporates the geometric information revealed by both labeled and unlabeled data points in a semi-parametric way. Specifically, the final projections of the data points will contain two parts: a discriminant part learned by traditional LDA (or KDA) on the labeled points and a geometrical part learned by kernel PCA on the whole data set. Therefore we call our algorithm semi-parametric semi-supervised discriminant analysis (SSDA). Experimental results on face recognition and image retrieval tasks are presented to show the effectiveness of our method.

Laplacian embedding provides a low dimensional representation for a matrix of pairwise similarity data using the eigenvectors of the Laplacian matrix. The true power of Laplacian embedding is that it provides an approximation of the ratio cut clustering. However, ratio cut clustering requires the solution to be nonnegative. In this paper, we propose a new approach, nonnegative Laplacian embedding, which approximates ratio cut clustering in a more direct way than traditional approaches. From the solution of our approach, clustering structures can be read off directly. We also propose an efficient algorithm to optimize the objective function utilized in our approach. Empirical studies on many real world datasets show that our approach leads to more accurate ratio cut solution and improves clustering accuracy at the same time.

Large IT service providers track service requests and their execution through problem/change tickets. It is important to classify the tickets based on the problem/change description in order to understand service quality and to optimize service processes. However, two challenges exist in solving this classification problem: 1) ticket descriptions from different classes are of highly diverse characteristics, which invalidates most standard distance metrics; 2) it is very expensive to obtain high-quality labeled data. To address these challenges, we develop two seemingly independent methods 1) discriminative neighborhood metric learning (DNML) and 2) active learning with median selection (ALMS), both of which are, however, based on the same core technique: iterated representative selection. A case study on real IT service classification application is presented to demonstrate the effectiveness and efficiency of our proposed methods.

Multi-task learning refers to the learning problem of performing inference by jointly considering multiple related tasks. There have already been many research efforts on supervised multi-task learning. However, collecting sufficient labeled data for each task is usually time consuming and expensive. In this paper, we consider the semi-supervised multitask learning (SSMTL) problem, where we are given a small portion of labeled points together with a large pool of unlabeled data within each task. We assume that the different tasks can form some task clusters and the task in the same cluster share similar classifier parameters. The final learning problem is relaxed to a convex one and an efficient gradient descent strategy is proposed. Finally the experimental results on both synthetic and real world data sets are presented to show the effectiveness of our method.

Clustering is one of the most fundamental and important problems in computer vision and pattern recognition communities. Maximum margin clustering (MMC) is a recently proposed clustering technique which has shown promising experimental results. The main theme behind MMC is to extend the standard maximum margin principle in support vector machine (SVM) to the unsupervised scenario. This paper will consider the problem of maximum margin clustering on data manifolds. Specifically, we propose an approach called manifold regularized maximum margin clustering (MRMMC) which combines both the maximum margin data discrimination and data manifold information in a unified clustering objective and propose an efficient algorithm to solve it. Finally the experimental results on several real world data sets are presented to show the effectiveness of our method.

Label propagation is an approach to assign class labels to unlabeled data given some partially labeled data. In this paper, we systematically generalize the Laplacian matrix based label propagation method from pairwise graph data to data objects described by bipartite and general K-partite graphs. By deriving explicit label propagation formula, we show how information on one type of variables can be transformed to other types of variables. For example, in a word-document-author multi-relational dataset, information on words and on authors can effectively enhance the document labeling. Motivating examples are presented to illustrate these new concepts. Extensive experiments are performed on real-life datasets to show the effectiveness of our label propagation.

In this article we present a new approach - weighted consensus clustering to identify the clusters in Protein-protein interaction (PPI) networks where each cluster corresponds to a group of functionally similar proteins. In weighed consensus clustering, different input clustering results weigh differently, i.e., a weight for each input clustering is introduced and the weights are automatically determined by an optimization process. We evaluate our proposed method with standard measures such as modularity, normalized mutual information (NMI) and the Gene Ontology (GO) consortium database and compare the performance of our approach with other consensus clustering methods. Experimental results demonstrate the effectiveness of our proposed approach.

Currently, the most significant line of defense against malware is anti-virus products which focus on authenticating valid software from a white list, blocking invalid software from a black list, and running any unknown software (i.e., the gray list) in a controlled manner. The gray list, containing unknown software programs which could be either normal or malicious, is usually authenticated or rejected manually by virus analysts. Unfortunately, along with the development of the malware writing techniques, the number of file samples in the gray list that need to be analyzed by virus analysts on a daily basis is constantly increasing. In this paper, we develop an intelligent file scoring system (IFSS for short) for malware detection from the gray list by an ensemble of heterogeneous base-level classifiers derived by different learning methods, using different feature representations on dynamic training sets. To the best of our knowledge, this is the first work of applying such ensemble methods for malware detection. IFSS makes it practical for virus analysts to identify malware samples from the huge gray list and improves the detection ability of anti-virus software. It has already been incorporated into the scanning tool of Kingsoft's Anti-Virus software. The case studies on large and real daily collection of the gray list illustrate that the detection ability and efficiency of our IFSS system outperforms other popular scanning tools such as NOD32 and Kaspersky.

Most datasets in real applications come in from multiple sources. As a result, we often have attributes information about data objects and various pairwise relations (similarity) between data objects. Traditional clustering algorithms use either data attributes only or pairwise similarity only. We propose to combine K-means clustering on data attributes and normalized cut spectral clustering on pairwise relations. We show that these two methods can be coherently integrated together to make use of different data sources to obtain good clustering results. We also show that our integrated KL (K-means – Laplacian) clustering method can be naturally extended to semi-supervised clustering, data embedding and metric learning. Finally the experimental results on benchmark data sets are presented to show the effectiveness of our method.

With the explosion of user-generated web2.0 content in the form of blogs, wikis and discussion forums, the Internet has rapidly become a massive dynamic repository of public opinion on an unbounded range of topics. A key enabler of opinion extraction and summarization is sentiment classification: the task of automatically identifying whether a given piece of text expresses positive or negative opinion towards a topic of interest. Building high-quality sentiment classifiers using standard text categorization methods is challenging due to the lack of labeled data in a target domain. In this paper, we consider the problem of cross-domain sentiment analysis: can one, for instance, download rated movie reviews from rottentomatoes.com or IMBD discussion forums, learn linguistic expressions and sentiment-laden terms that generally characterize opinionated reviews and then successfully transfer this knowledge to the target domain, thereby building high-quality sentiment models without manual effort? We outline a novel sentiment transfer mechanism based on constrained non-negative matrix tri-factorizations of term-document matrices in the source and target domains. We report some preliminary results with this approach.

In this poster, we develop an evolutionary document summarization system for discovering the changes and differences in each phase of a disaster evolution. Given a collection of document streams describing an event, our system generates a short summary delivering the main development theme of the event by extracting the most representative and discriminative sentences at each phase. Experimental results on the collection of press releases for Hurricane Wilma in 2005 demonstrate the efficacy of our proposal.

This paper presents a novel method for mining product reviews, where it mines reviews by identifying product features, expressions of opinions and relations between them. By taking advantage of the fact that most of product features are phrases, a concept of shallow dependency parsing is introduced, which extends traditional dependency parsing to phrase level. This concept is then implemented for extracting relation between product features and expressions of opinions. Experimental evaluations show that the mining task can benefit from shallow dependency parsing.

Music recommendation is receiving increasing attention as the music industry develops venues to deliver music over the Internet. The goal of music recommendation is to present users lists of songs that they are likely to enjoy. Collaborative-filtering and content-based recommendations are two widely used approaches that have been proposed for music recommendation. However, both approaches have their own disadvantages: collaborative-filtering methods need a large collection of user history data and content-based methods lack the ability of understanding the interests and preferences of users. To overcome these limitations, this paper presents a novel dynamic music similarity measurement strategy that utilizes both content features and user access patterns. The seamless integration of them significantly improves the music similarity measurement accuracy and performance. Based on this strategy, recommended songs are obtained by a means of label propagation over a graph representing music similarity. Experimental results on a real data set collected from http://www.newwisdom.net demonstrate the effectiveness of the proposed approach.

Clustering is an old research topic in data mining and machine learning. Most of the traditional clustering methods can be categorized as local or global ones. In this paper, a novel clustering method that can explore both the local and global information in the data set is proposed. The method, Clustering with Local and Global Regularization (CLGR), aims to minimize a cost function that properly trades off the local and global costs. We show that such an optimization problem can be solved by the eigenvalue decomposition of a sparse symmetric matrix, which can be done efficiently using iterative methods. Finally, the experimental results on several data sets are presented to show the effectiveness of our method.

Efficient and intelligent music information retrieval is a very important topic of the 21st century. With the ultimate goal of building personal music information retrieval systems, this paper studies the problem of identifying ldquosimilarrdquo artists using features from diverse information sources. In this paper, we first present a clustering algorithm that integrates features from both sources to perform bimodal learning. We then present an approach based on the generalized constraint clustering algorithm by incorporating the instance-level constraints. The algorithms are tested on a data set consisting of 570 songs from 53 albums of 41 artists using artist similarity provided by All Music Guide. Experimental results show that the accuracy of artist similarity identification can be significantly improved.

Clustering multi-way data is a very important research topic due to the intrinsic rich structures in real-world datasets. In this paper, we propose the subspace clustering algorithm on multi-way data, called ASI-T (Adaptive Subspace Iteration on Tensor). ASI-T is a special version of High Order SVD (HOSVD), and it simultaneously performs subspace identification using 2DSVD and data clustering using K-Means. The experimental results on synthetic data and real-world data demonstrate the effectiveness of ASI-T.

Distance metric learning is an old problem that has been researched in the supervised learning field for a very long time. In this paper, we consider the problem of learning a proper distance metric under the guidance of some weak supervisory information. Specifically, those information are in the form of pairwise constraints which specify whether a pair of data points are in the same class (must link constraints) or in the different classes (cannot link constraints). Given those constraints, our algorithm aims to learn a distance metric under which the points with must link constraints are pushed as close as possible, while simultaneously the points with cannot link constraints are pulled away as far as possible. Finally the experimental results are presented to show the effectiveness of our method.

Document understanding techniques such as document clustering and multi-document summarization have been receiving much attention in recent years. Current document clustering methods usually represent documents as a term-document matrix and perform clustering algorithms on it. Although these clustering methods can group the documents satisfactorily, it is still hard for people to capture the meanings of the documents since there is no satisfactory interpretation for each document cluster. In this paper, we propose a new language model to simultaneously cluster and summarize the documents. By utilizing the mutual influence of the document clustering and summarization, our method makes (1) a better document clustering method with more meaningful interpretation and (2) a better document summarization method taking the document context information into consideration.

Nonnegative matrix factorization (NMF) is a versatile model for data clustering. In this paper, we propose several NMF inspired algorithms to solve different data mining problems. They include (1) multi-way normalized cut spectral clustering, (2) graph matching of both undirected and directed graphs, and (3) maximal clique finding on both graphs and bipartite graphs. Key features of these algorithms are (a) they are extremely simple to implement; and (b) they are provably convergent. We conduct experiments to demonstrate the effectiveness of these new algorithms. We also derive a new spectral bound for the size of maximal edge bicliques as a byproduct of our approach.

Singular Value Decomposition (SVD)/Principal Component Analysis (PCA) have played a vital role in finding patterns from many datasets. Recently tensor factorization has been used for data mining and pattern recognition in high index/order data. High Order SVD (HOSVD) is a commonly used tensor factorization method and has recently been used in numerous applications like graphs, videos, social networks, etc.

Consensus clustering has emerged as an important extension of the classical clustering problem. We propose weighted consensus clustering, where each input clustering is weighted and the weights are determined in such a way that the final consensus clustering provides a better quality solution, in which clusters are better separated comparing to standard consensus clustering. Theoretically, we show that a reformulation of the well-known L1 regularization LASSO problem is equivalent to the weight optimization of our weighted consensus clustering, and thus our approach provides sparse solutions which may resolve the difficult situation when the input clusterings diverge significantly. We also show that the weighted consensus clustering resolves the redundancy problem when many input clusterings correlate highly. Detailed algorithms are given. Experiments are carried out to demonstrate the effectiveness of the weighted consensus clustering.

The recent years have witnessed a surge of interests of semi-supervised clustering methods, which aim to cluster the data set under the guidance of some supervisory information. Usually those supervisory information takes the form of pairwise constraints that indicate the similarity/dissimilarity between the two points. In this paper, we propose a novel matrix factorization based approach for semi-supervised clustering. In addition, we extend our algorithm to co-cluster the data sets of different types with constraints. Finally the experiments on UCI data sets and real world Bulletin Board Systems (BBS) data sets show the superiority of our proposed method.

We introduce the posterior probabilistic clustering (PPC), which provides a rigorous posterior probability interpretation for Nonnegative Matrix Factorization (NMF) and removes the uncertainty in clustering assignment. Furthermore, PPC is closely related to probabilistic latent semantic indexing (PLSI).

In most IR clustering problems, we directly cluster the documents, working in the document space, using cosine similarity between documents as the similarity measure. In many real-world applica-tions, however, we usually have knowledge on the word side and wish to transform this knowledge to the document (concept) side. In this paper, we provide a mechanism for this knowledge trans-formation. To the best of our knowledge, this is the rst model for such type of knowledge transformation. This model uses a nonneg-ative matrix factorization model X = FSGT, where X is the word-document semantic matrix, F is the posterior probability of a word belonging to a word cluster and represents knowledge in the word space, G is the posterior probability of a document belonging to a document cluster and represents knowledge in the document space, and S is a scaled matrix factor which provides a condensed view of X. We show how knowledge on words can improve document clustering, i.e, knowledge in the word space is transformed into the document space. We perform extensive experiments to validate our approach.

Analyzing three-way data has attracted a lot of attention recently due to the intrinsic rich structures in real-world datasets. The PARATUCKER model has been proposed to combine the axis capabilities of the Parafac model and the structural generality of the Tucker model. However, no algorithms have been developed for fitting the PARATUCKER model. In this paper, we propose TANPT algorithm to solve the PARATUCKER model. We apply the algorithm for temporal relation co-clustering on author-topic evolution. Experiments on DBLP datasets demonstrate its effectiveness.

Multi-document summarization aims to create a compressed summary while retaining the main characteristics of the original set of documents. Many approaches use statistics and machine learning techniques to extract sentences from documents. In this paper, we propose a new multi-document summarization framework based on sentence-level semantic analysis and symmetric non-negative matrix factorization. We first calculate sentence-sentence similarities using semantic analysis and construct the similarity matrix. Then symmetric matrix factorization, which has been shown to be equivalent to normalized spectral clustering, is used to group sentences into clusters. Finally, the most informative sentences are selected from each group to form the summary. Experimental results on DUC2005 and DUC2006 data sets demonstrate the improvement of our proposed framework over the implemented existing summarization systems. A further study on the factors that benefit the high performance is also conducted.

Helpdesk databases are used to store past interactions between customers and companies to improve customer service quality. One common scenario of using helpdesk database is to find whether recommendations exist given a new problem from a customer. However, customers often provide incomplete or even inaccurate information. Manually preparing a list of clarification questions does not work for large databases. This paper investigates the problem of automatic generation of a minimal number of questions to reach an appropriate recommendation. This paper proposes a novel dynamic active probing method. Compared to other alternatives such as decision tree and case-based reasoning, this method has two distinctive features. First, it actively probe the customer to get useful information to reach the recommendation, and the information provided by customer will be immediately used by the method to dynamically generate the next questions to probe. This feature ensures that all available information from the customer is used. Second, this method is based on a probabilistic model, and uses a data augmentation method which avoids overfitting when estimating the probabilities in the model. This feature ensures that the method is robust to databases that are incomplete or contain errors. Experimental results verify the effectiveness of our approach.

Given a user keyword query, current Web search engines return a list of individual Web pages ranked by their "goodness" with respect to the query. Thus, the basic unit for search and retrieval is an individual page, even though information on a topic is often spread across multiple pages. This degrades the quality of search results, especially for long or uncorrelated (multitopic) queries (in which individual keywords rarely occur together in the same document), where a single page is unlikely to satisfy the user's information need. We propose a technique that, given a keyword query, on the fly generates new pages, called composed pages, which contain all query keywords. The composed pages are generated by extracting and stitching together relevant pieces from hyperlinked Web pages and retaining links to the original Web pages. To rank the composed pages, we consider both the hyperlink structure of the original pages and the associations between the keywords within each page. Furthermore, we present and experimentally evaluate heuristic algorithms to efficiently generate the top composed pages. The quality of our method is compared to current approaches by using user surveys. Finally, we also show how our techniques can be used to perform query-specific summarization of Web pages.

The recent development of microarray gene expression techniques have made it possible to offer phenotype classification of many diseases. However, in gene expression data analysis, each sample is represented by quite a large number of genes, and many of them are redundant or insignificant to clarify the disease problem. Therefore, how to efficiently select the most useful genes has been becoming one of the most hot research topics in the gene expression data analysis. In this paper, a novel unsupervised gene selection method is proposed based on matrix factorization, such that the original gene matrix can be optimally reconstructed using those selected genes. To make our algorithm more efficient, we derive a kmeans preclustering approach to accelerate our algorithm. Finally the experimental results on several data sets are presented to show the effectiveness of our method.

In this paper, we have modified a constrained clustering algorithm to perform exploratory analysis on gene expression data using prior knowledge presented in the form of constraints. We have also studied the effectiveness of various constraints sets. To address the problem of automatically generating constraints from biological text literature, we considered two methods (cluster-based and similarity-based). We concluded that incomplete information in the form of constraints set should be generated carefully, in order to outperform the standard clustering algorithm, which works on the data source without any constraints. For sufficiently large constraints sets, the constrained clustering algorithm outperformed the MSC algorithm. The novelty of research presented here is the study of effectiveness of constraints sets and robustness of the constrained clustering algorithm using multiple sources of biological data, and incorporating biomedical text literature into constrained clustering algorithm in form of constraints sets.

Gene expression data usually contains a large number of genes, but a small number of samples. Feature selection for gene expression data aims at finding a set of genes that best discriminate biological samples of different types. In this paper, we present a two-stage selection algorithm by combining ReliefF and mRMR: In the first stage, ReliefF is applied to find a candidate gene set; In the second stage, mRMR method is applied to directly and explicitly reduce redundancy for selecting a compact yet effective gene subset from the candidate set. We also perform comprehensive experiments to compare the mRMR-ReliefF selection algorithm with ReliefF, mRMR and other feature selection methods using two classifiers as SVM and Naive Bayes, on seven different datasets. The experimental results show that the mRMR-ReliefF gene selection algorithm is very effective.

Consensus clustering and semi-supervised clustering are important extensions of the standard clustering paradigm. Consensus clustering (also known as aggregation of clustering) can improve clustering robustness, deal with distributed and heterogeneous data sources and make use of multiple clustering criteria. Semi-supervised clustering can integrate various forms of background knowledge into clustering. In this paper, we show how consensus and semi-supervised clustering can be formulated within the framework of nonnegative matrix factorization (NMF). We show that this framework yields NMF-based algorithms that are: (1) extremely simple to implement; (2) provably correct and provably convergent. We conduct a wide range of comparative experiments that demonstrate the effectiveness of this NMF-based approach.

An interesting problem in nonnegative matrix factorization (NMF) is to factorize the matrix X which is of some specific class, for example, binary matrix. In this paper, we extend the standard NMF to binary matrix factorization (BMF for short): given a binary matrix X, we want to factorize X into two binary matrices W, H (thus conserving the most important integer property of the objective matrix X) satisfying X ap WH. Two algorithms are studied and compared. These methods rely on a fundamental boundedness property of NMF which we propose and prove. This new property also provides a natural normalization scheme that eliminates the bias of factor matrices. Experiments on both synthetic and real world datasets are conducted to show the competency and effectiveness of BMF.

We combine linear discriminant analysis (LDA) and K-means clustering into a coherent framework to adaptively select the most discriminative subspace. We use K-means clustering to generate class labels and use LDA to do subspace selection. The clustering process is thus integrated with the subspace selection process and the data are then simultaneously clustered while the feature subspaces are selected. We show the rich structure of the general LDA-Km framework by examining its variants and their relationships to earlier approaches. Relations among PCA, LDA, K-means are clarified. Extensive experimental results on real-world datasets show the effectiveness of our approach.

A novel algorithm for interactive multilabel image/video segmentation is proposed in this paper. Given a small number of pixels with user-defined (or pre-defined) labels, our method can automatically propagate those labels to the remaining unlabeled pixels through an iterative procedure. Theoretical analysis of the convergence property of this algorithm is developed along with the corresponding connections with energy minimization of the hidden Markov random field models. To make the algorithm more efficient, we also derive a multi-level way for propagating the labels. Finally the segmentation results on natural images are presented to show the effectiveness of our method.

Given a document collection, it is often desirable to find the core subset of documents focusing on a specific topic. We propose a new algorithm for this task. Document clustering aims at par- titioning the document-term datasets into differ- ent groups by optimizing certain objective func- tions. However, they are not suitable for finding hotspots that are described by a small set of doc- uments with few tightly coupled terms. In this pa- per we propose a novel hotspot finding algorithm, DCC (Dense Concept Clustering) in document collections. DCC can extract distinct small top- ics with most representative documents and words simultaneously. The hotspots are dense bicliques in binary document-word matrices and they can be discovered sequentially one at a time using the generalized Motzkin-Straus formalism. The rep- resentative documents and words are tightly cor- related for concept descriptions. Experiments on real document datasets show the effectiveness of the proposed algorithm.

Green's function for the Laplace operator represents the propagation of influence of point sources and is the foundation for solving many physics problems. On a graph of pairwise similarities, the Green's function is the inverse of the combinatorial Laplacian; we resolve the zero-mode difficulty by showing its physical origin as the consequence of the Von Neumann boundary condition. We propose to use Green's function to propagate label information for both semi-supervised and unsupervised learning. We also derive this learning framework from the kernel regularization using Reproducing Kernel Hilbert Space theory at strong regularization limit. Green's function provides a well-defined distance metric on a generic weighted graph, either as the effective distance on the network of electric resistors, or the average commute time in random walks. We show that for unsupervised learning this approach is identical to Ratio Cut and Normalized Cut spectral clustering algorithms. Experiments on newsgroups and six UCI datasets illustrate the effectiveness of this approach. Finally, we propose a novel item-based recommender system using Green's function and show its effectiveness.

In system management applications, an overwhelming amount of data are generated and collected in the form of temporal events. While mining temporal event data to discover interesting and frequent patterns has obtained rapidly increasing research efforts, users of the applications are overwhelmed by the mining results. The extracted patterns are generally of large volume and hard to interpret, they may be of no emphasis, intricate and meaningless to non-experts, even to domain experts. While traditional research efforts focus on finding interesting patterns, in this paper, we take a novel approach called event summarization towards the understanding of the seemingly chaotic temporal data. Event summarization aims at providing a concise interpretation of the seemingly chaotic data, so that domain experts may take actions upon the summarized models. Event summarization decomposes the temporal information into many independent subsets and finds well fitted models to describe each subset.

The proliferation of malware has presented a serious threat to the security of computer systems. Traditional signature-based anti-virus systems fail to detect polymorphic and new, previously unseen malicious executables. In this paper, resting on the analysis of Windows API execution sequences called by PE files, we develop the Intelligent Malware Detection System (IMDS) using Objective-Oriented Association (OOA) mining based classification. IMDS is an integrated system consisting of three major modules: PE parser, OOA rule generator, and rule based classifier. An OOA_Fast_FP-Growth algorithm is adapted to efficiently generate OOA rules for classification. A comprehensive experimental study on a large collection of PE files obtained from the anti-virus laboratory of King-Soft Corporation is performed to compare various malware detection approaches. Promising experimental results demonstrate that the accuracy and efficiency of our IMDS system out perform popular anti-virus software such as Norton AntiVirus and McAfee VirusScan, as well as previous data mining based detection systems which employed Naive Bayes, Support Vector Machine (SVM) and Decision Tree techniques.

In recent years, document clustering has been receiving more and more attentions as an important and fundamental technique for unsupervised document organization, automatictopic extraction, and fast information retrieval or filtering. In this paper, we propose a novel method for clustering documents using regularization. Unlike traditional globally regularized clustering methods, our method first construct a local regularized linear label predictor for each document vector, and then combine all those local regularizers with a global smoothness regularizer. So we call our algorithm Clustering with Local and Global Regularization (CLGR). We will show that the cluster memberships of the documents can be achieved by eigenvalue decomposition of a sparse symmetric matrix, which can be efficiently solved by iterative methods. Finally our experimental evaluations on several datasets are presented to show the superiorities of CLGR over traditional document clustering methods.

Database queries are often exploratory and users often find their queries return too many answers, many of them irrelevant. Existing work either categorizes or ranks the results to help users locate interesting results. The success of both approaches depends on the utilization of user preferences. However, most existing work assumes that all users have the same user preferences, but in real life different users often have different preferences. This paper proposes a two-step solution to address the diversity issue of user preferences for the categorization approach. The proposed solution does not require explicit user involvement. The first step analyzes query history of all users in the system offline and generates a set of clusters over the data, each corresponding to one type of user preferences. When user asks a query, the second step presents to the user a navigational tree over clusters generated in the first step such that the user can easily select the subset of clusters matching his needs. The user then can browse, rank, or categorize the results in selected clusters. The navigational tree is automatically constructed using a cost-based algorithm which considers the cost of visiting both intermediate nodes and leaf nodes in the tree. An empirical study demonstrates the benefits of our approach. Categories and Subject Descriptors

IntClust is a software package for clustering gene-expression data with repeated measurements based on interval data analysis. By utilizing interval data for representing replicated microarray data, IntClust is able to take into account the scopes where replicate microarray data are distributed instead of simple data points. The soft ware package offers several transformation models for interval data representations, supports different extended distance similarity/distance measures for interval data analysis, provides some variations of modified K-means clustering, and presents three popular clustering quality evaluation measures. Our experiments show that IntClust improves the clustering performance of gene-expression microarray data over traditional approaches. The software package is available at http://cadse24.cs.fiu.edu/IntClust

In this paper, we study the problems of motif discovery and gene regulation. First, although the sliding window technique based on profiles or consensus sequences is a standard method for discovering motifs in the genomes with prior knowledge of transcription binding sites in orthologous genes from related organisms, it usually has high computational costs. In this paper, we propose an efficient approximation method employing randomized algorithms to identify motifs. The approximation method can be easily combined with the sliding-window technique for efficient and accurate motif discovery. Second, we mine frequent motif combinations and sequential motif patterns to investigate the regulatory relationships between motifs and provide a better understanding of gene expression, regulation, and transcription

Biclustering has many applications in text mining, Web clickstream mining, and bioinformatics. When data entries are binary, the tightest biclusters become bicliques. We propose a flexible and highly efficient algorithm to compute bicliques. We first generalize the Motzkin-Straus formalism for computing the maximal clique from L1constraint to Lpconstraint, which enables us to provide a generalized Motzkin-Straus formalism for computing maximal-edge bicliques. By adjusting parameters, the algorithm can favor biclusters with more rows less columns, or vice verse, thus increasing the flexibility of the targeted biclusters. We then propose an algorithm to solve the generalized Motzkin-Straus optimization problem. The algorithm is provably convergent and has a computational complexity of O(/E/) where /E/ is the number of edges. Using this algorithm, we bicluster the yeast protein complex interaction network. We find that biclustering protein complexes at the protein level does not clearly reflect the functional linkage among protein complexes in many cases, while biclustering at protein domain level can reveal many underlying linkages. We show several new biologically significant results.

The nonnegative matrix factorization (NMF) has been shown recently to be useful for clustering and various extensions and variations of NMF have been proposed recently. Despite significant research progress in this area, few attempts have been made to establish the connections between various factorization methods while highlighting their differences. In this paper we aim to provide a comprehensive study on matrix factorization for clustering. In particular, we present an overview and summary on various matrix factorization algorithms and theoretically analyze the relationships among them. Experiments are also conducted to empirically evaluate and compare various factorization methods. In addition, our study also answers several previously unaddressed yet important questions for matrix factorizations including the interpretation and normalization of cluster posterior and the benefits and evaluation of simultaneous clustering. We expect our study would provide good insights on matrix factorization research for clustering.

Efficient and intelligent music information retrieval is a very important topic of the 21st century. With the ultimate goal of building personal music information retrieval systems, this paper studies the problem of identifying "similar" artists using both lyrics and acoustic data. In this paper, we present a clustering algorithm that integrates features from both sources to perform bimodal learning. The algorithm is tested on a data set consisting of 570 songs from 53 albums of 41 artists using artist similarity provided by All Music Guide. Experimental results show that the accuracy of artist similarity classifiers can be significantly improved and that artist similarity can be efficiently identified.

A novel scheme for item-based recommendation is proposed in this paper. In our framework, the items are described by an undirected weighted graph Q = (V,epsiv). V is the node set which is identical to the item set, and epsiv is the edge set. Associate with each edge eij isin epsiv is a weight omegaij ges 0, which represents similarity between items i and j. Without the loss of generality, we assume that any user's ratings to the items should be sufficiently smooth with respect to the intrinsic structure of the items, i.e., a user should give similar ratings to similar items. A simple algorithm is presented to achieve such a smooth solution. Encouraging experimental results are provided to show the effectiveness of our method.

In classic pattern recognition problems, classes are mutually exclusive by definition. However, in many applications, it is quite natural that some instances belong to multiple classes at the same time. In other words, these applications are multi-labeled, classes are overlapped by definition and each instance may be associated to multiple classes. In this paper, we present a comparative study on various multi-label approaches using both gene and scene data sets. We expect our research efforts provide useful insights on the relationships among various classifiers as well as various evaluation measures and shed lights on future research. Although there is no clear winner across various performance measures, SVM binary and multi-label ADTree perform better than the others on most counts. We then propose a meta-learning approach by combining SVM binary and ADTree. Our experiments demonstrate that the combined method can take the advantages of the single approaches

Interval data is described by a group of variables, each of which contains a range of continuous values instead of the traditional single continuous or discrete value. Traditional data analysis simply replaces each interval by its representative (e.g., center or mean) and ignores the structure information of intervals. In this paper, we study the problem of clustering interval data using the modified or extended interval data dissimilarity measures. Our contributions are two-fold. First, we discuss various approaches for measuring the dissimilarities/distances between interval data, investigate the relations among them, and present a comprehensive experimental study on clustering interval data. We show that the extended interval data clustering achieves better performance than traditional ones and produces more meaningful and explanatory results. Second, we propose a two-stage approach for clustering interval data by exploiting the relations between the traditional distances and the modified distances. Experimental results show the effectiveness of our approach

Currently, most research on nonnegative matrix factorization (NMF)focus on 2-factor $X=FG^T$ factorization. We provide a systematicanalysis of 3-factor $X=FSG^T$ NMF. While it unconstrained 3-factor NMF is equivalent to it unconstrained 2-factor NMF, itconstrained 3-factor NMF brings new features to it constrained 2-factor NMF. We study the orthogonality constraint because it leadsto rigorous clustering interpretation. We provide new rules for updating $F,S, G$ and prove the convergenceof these algorithms. Experiments on 5 datasets and a real world casestudy are performed to show the capability of bi-orthogonal 3-factorNMF on simultaneously clustering rows and columns of the input datamatrix. We provide a new approach of evaluating the quality ofclustering on words using class aggregate distribution andmulti-peak distribution. We also provide an overview of various NMF extensions andexamine their relationships.

In this paper, we show that PLSI and NMF optimize the same objective function, although PLSI and NMF are different algorithms as verified by experiments. In addition, we also propose a new hybrid method that runs PLSI and NMF alternatively to achieve better solutions.

No abstract available.

For today's business organizations, workflow models play important roles in analyzing the productivity, evaluating the performances and costs, optimizing the business operations, and supporting evolving services and products. Workflow mining, the process of empirically extracting structured process descriptions from a set of real executions, thus has attracted a lot of attention recently. However, there are several challenges that have not been fully addressed in the previous research: (i) How can we mine process models with optional tasks? (ii) How can we efficiently use new available workflow log data to incrementally update pre-existing workflow models or to complete previous partial process models? (iii) How can we compare two different workflow models of similar organizations? In this paper, we present our research efforts to address the above challenges. We present a workflow mining algorithm that is able to mine process models with optional tasks and propose an incremental workflow mining algorithm based on intermediate relationships such as ordering and independence. The intermediate relationships can also be used to facilitate the comparison of two process models. We illustrate our algorithms on example data derived from real world applications.

Efficient and intelligent music information retrieval is a very important topic of the 21st century. With the ultimate goal of building personal music information retrieval systems, this paper studies the problem of intelligent music information retrieval. Huron points out that since the preeminent functions of music are social and psychological, the most useful characterization would be based on four types of information: genre, emotion, style,and similarity. This paper introduces Daubechies Wavelet Coefficient Histograms (DWCH)for music feature extraction for music information retrieval. The histograms are computed from the coefficients of the db/sub 8/ Daubechies wavelet filter applied to 3 s of music. A comparative study of sound features and classification algorithms on a dataset compiled by Tzanetakis shows that combining DWCH with timbral features (MFCC and FFT), with the use of multiclass extensions of support vector machine,achieves approximately 80% of accuracy, which is a significant improvement over the previously known result on this dataset. On another dataset the combination achieves 75% of accuracy. The paper also studies the issue of detecting emotion in music. Rating of two subjects in the three bipolar adjective pairs are used. The accuracy of around 70% was achieved in predicting emotional labeling in these adjective pairs. The paper also studies the problem of identifying groups of artists based on their lyrics and sound using a semi-supervised classification algorithm. Identification of artist groups based on the Similar Artist lists at All Music Guide is attempted. The semi-supervised learning algorithm resulted in nontrivial increases in the accuracy to more than 70%. Finally, the paper conducts a proof-of-concept experiment on similarity search using the feature set.

Automatic music genre classification is a fundamental component of music information retrieval systems and has been gaining importance and enjoying a growing amount of attention with the emergence of digital music on the Internet. Although considerable research has been conducted in automatic music genre classification, little has been done on hierarchical classification with taxonomies. The underlying hierarchical taxonomy identifies the relationships of dependence between different genres and provides valuable sources of information for genre classification. This paper investigates the use of taxonomy for music genre classification. Our empirical experiments on two datasets show that using taxonomy improves the classification performance. We also propose an approach for automatically generating genre taxonomies based on the confusion matrix via linear discriminant projection. Our work also provides some insights for future research.

Traditional approaches to system management have been largely based on domain experts through a knowledge acquisition process that translates domain knowledge into operating rules and policies. This has been well known and experienced as a cumbersome, labor intensive, and error prone process. In addition, this process is difficult to keep up with the rapidly changing environments. In this paper, we will describe our research efforts on establishing an integrated framework for mining system log files for automatic management. In particular, we apply text mining techniques to categorize messages in log files into common situations, improve categorization accuracy by considering the temporal characteristics of log messages, develop temporal mining techniques to discover the relationships between different events, and utilize visualization tools to evaluate and validate the interesting temporal patterns for system management.

No abstract available.

Many problems can be reduced to the problem of combining multiple clusterings. In this paper, we first summarize different application scenarios of combining multiple clusterings and provide a new perspective of viewing the problem as a categorical clustering problem. We then show the connections between various consensus and clustering criteria and discuss the complexity results of the problem. Finally we propose a new method to determine the final clustering. Experiments on kinship terms and clustering popular music from heterogeneous feature sets show the effectiveness of combining multiple clusterings.

The paper investigates the use of acoustic based features for music information retrieval. Two specific problems are studied: similarity search (searching for music sound files similar to a given music sound file) and emotion detection (detection of emotion in music sounds). The Daubechies wavelet coefficient histograms (Li, T. et al., SIGIR'03, p.282-9, 2003), which consist of moments of the coefficients calculated by applying the Db8 wavelet filter, are combined with the timbral features extracted using the MARSYAS system of G. Tzanctakis and P. Cook (see IEEE Trans. on Speech and Audio Process., vol.10, no.5, p.293-8, 2002) to generate compact music features. For the similarity search, the distance between two sound files is defined to be the Euclidean distance of their normalized representations. Based on the distance measure, the closest sound files to an input sound file are obtained. Experiments on jazz vocal and classical sound files achieve a very high level of accuracy. Emotion detection is cast as a multiclass classification problem, decomposed as a multiple binary classification problem, and is resolved with the use of support vector machines trained on the extracted features. Our experiments on emotion detection achieved reasonably accurate performance and provided some insights on future work.

Entropy-type measures for the heterogeneity of clusters have been used for a long time. This paper studies the entropy-based criterion in clustering categorical data. It first shows that the entropy-based criterion can be derived in the formal framework of probabilistic clustering models and establishes the connection between the criterion and the approach based on dissimilarity co-efficients. An iterative Monte-Carlo procedure is then presented to search for the partitions minimizing the criterion. Experiments are conducted to show the effectiveness of the proposed procedure.

Efficient and intelligent music information retrieval is a very important topic of the 21st century. With the ultimate goal of building personal music information retrieval systems, this paper studies the problem of identifying "similar" artists using both lyrics and acoustic data. The approach for using a small set of labeled samples for the seed labeling to build classifiers that improve themselves using unlabeled data is presented. This approach is tested on a data set consisting of 43 artists and 56 albums using artist similarity provided by All Music Guide. Experimental results show that using such an approach the accuracy of artist similarity classifiers can be significantly improved and that artist similarity can be efficiently identified.

Document clustering has long been an important problem in information retrieval. In this paper, we present a new clustering algorithm ASI1 , which uses explicitly modeling of the subspace structure associated with each cluster. ASI simultaneously performs data reduction and subspace identification via an iterative alternating optimization procedure. Motivated from the optimization procedure, we then provide a novel method to determine the number of clusters. We also discuss the connections of ASI with various existential clustering approaches. Finally, extensive experimental results on real data sets show the effectiveness of ASI algorithm.

This paper studies the problem of building multiclass classifiers for tissue classification based on gene expression. The recent development of microarray technologies has enabled biologists to quantify gene expression of tens of thousands of genes in a single experiment. Biologists have begun collecting gene expression for a large number of samples. One of the urgent issues in the use of microarray data is to develop methods for characterizing samples based on their gene expression. The most basic step in the research direction is binary sample classification, which has been studied extensively over the past few years. This paper investigates the next step-multiclass classification of samples based on gene expression. The characteristics of expression data (e.g. large number of genes with small sample size) makes the classification problem more challenging. The process of building multiclass classifiers is divided into two components: (i) selection of the features (i.e. genes) to be used for training and testing and (ii) selection of the classification method. This paper compares various feature selection methods as well as various state-of-the-art classification methods on various multiclass gene expression datasets. Our study indicates that multiclass classification problem is much more difficult than the binary one for the gene expression datasets. The difficulty lies in the fact that the data are of high dimensionality and that the sample size is small. The classification accuracy appears to degrade very rapidly as the number of classes increases. In particular, the accuracy was very low regardless of the choices of the methods for large-class datasets (e.g. NCI60 and GCM). While increasing the number of samples is a plausible solution to the problem of accuracy degradation, it is important to develop algorithms that are able to analyze effectively multiple-class expression data for these special datasets.

Text categorization is an important research area and has been receiving much attention due to the growth of the on-line information and of Internet. Automated text categorization is generally cast as a multi-class classification problem. Much of previous work focused on binary document classification problems. Support vector machines (SVMs) excel in binary classification, but the elegant theory behind large-margin hyperplane cannot be easily extended to multi-class text classification. In addition, the training time and scaling are also important concerns. On the other hand, other techniques naturally extensible to handle multi-class classification are generally not as accurate as SVM. This paper presents a simple and efficient solution to multi-class text categorization. Classification problems are first formulated as optimization via discriminant analysis. Text categorization is then cast as the problem of finding coordinate transformations that reflects the inherent similarity from the data. While most of the previous approaches decompose a multiclass classification problem into multiple independent binary classification tasks, the proposed approach enables direct multi-class classification. By using Generalized Singular Value Decomposition (GSVD), a coordinate transformation that reflects the inherent class structure indicated by the generalized singular values is identified. Extensive experiments demonstrate the efficiency and effectiveness of the proposed approach.

Discriminant analysis is known to learn discriminative feature transformations. We study its use in multiclass classification problems. The performance is tested on a large collection of benchmark datasets.

Content-based music genre classification is a fundamental component of music information retrieval systems and has been gaining importance and enjoying a growing amount of attention with the emergence of digital music on the Internet. Currently little work has been done on automatic music genre classification, and in addition, the reported classification accuracies are relatively low. This paper proposes a new feature extraction method for music genre classification, DWCHs. DWCHs stands for Daubechies Wavelet Coefficient Histograms. DWCHs capture the local and global information of music signals simultaneously by computing histograms on their Daubechies wavelet coefficients. Effectiveness of this new feature and of previously studied features are compared using various machine learning classification algorithms, including Support Vector Machines and Linear Discriminant Analysis. It is demonstrated that the use of DWCHs significantly improves the accuracy of music genre classification.

No abstract available.

Many real-world datasets contain noise which could degrade the performances of learning algorithms. Motivated from the success of wavelet denoising techniques in image data, we explore a general solution to alleviate the effect of noisy data by wavelet preprocessing for medical/biological data classification. Our experiments are divided into two categories: one is of different classification algorithms on a specific database, and the other is of a specific classification algorithm (decision tree) on different databases. The experiment results show that the wavelet denoising of noisy data is able to improve the accuracies of those classification methods, if the localities of the attributes are strong enough.