Knowledge graphs are vital for many tasks, including recommendation systems and node search. Learning to hash knowledge graph is to infer binary-vector representations of the graph. Compared with traditional knowledge graph embedding that learns continuous-vector representations, knowledge graph hashing could significantly reduce storage and computational time due to its binary nature. Despite the potential advantage, the problem of knowledge graph hashing is challenging due to the large-scale binary decision variables. In this article, we propose a novel discrete optimization framework for knowledge graph hashing. We treat the relations between heads and tails in the knowledge graph as element-wise rotation to learn binary codes. An alternating optimization algorithm is then proposed to produce high-quality code that captures knowledge graph information well. Furthermore, to obtain superior binary representations, we employ a dynamic range method during the alternating optimization process to adjust the approximations of the ReLU function \([x]_{+}\) . This ensures that valuable measures of dissimilarity are not overlooked, leading to more accurate computations. The evaluation results on five publicly available datasets demonstrate the superiority of the proposed algorithm against several state-of-the-art baseline methods.

Recommender systems play a pivotal role in curating high-quality content for users, predominantly leveraging data-driven algorithms and machine learning methodologies. However, the intrinsic data-centric nature of these systems raises critical concerns; biased datasets and algorithms can inadvertently propagate biases to end-users. Furthermore, machine learning techniques, while powerful, can overfit a user’s preference, leading to a monotonous stream of content suggestions. Both the CS and IS community have well-recognized the need of fairness and diversity in recommender systems and many studies are proposed to mitigate these challenges. Yet, a tangible solution that holistically addressed all three components—accuracy, fairness, and diversity—in unison remains elusive. This article aims to bridge this gap, introducing a novel Influence-Function-Guided, Fair, and Diverse Data Enhancement (InFoDance) approach that enhances all three perspectives simultaneously. It consists of four interconnected modules: model training, candidate data generation, influence function-based candidate evaluation, and virtual data selection. It iteratively generates virtual data to update the trained recommender system. The empirical evaluation has shown that our approach can improve accuracy, fairness, and diversity by up to 24.27%, 55.29%, and 1.85% simultaneously and significantly outperform the state-of-the-art baselines on multiple evaluation metrics.

In today's competitive and fast-evolving business environment, it is a critical time for organizations to rethink how to deal with talent and management-related tasks in a quantitative manner. Indeed, thanks to the era of big data, the availability of large-scale talent data provides unparalleled opportunities for business leaders to understand the rules of talent and management, which in turn deliver intelligence for effective decision-making and management for their organizations. In the past few years, talent and management computing have increasingly attracted attention from KDD communities, and a number of research/applied data science efforts have been devoted. To this end, the purpose of this workshop, i.e., the 5th International Workshop on Talent and Management Computing (TMC'2024), is to bring together researchers and practitioners to discuss both the critical problems faced by talent and management-related domains and potential data-driven solutions by leveraging state-of-the-art data mining technologies.

Since social media such as Facebook and X (formerly known as Twitter) have permeated various aspects of daily life, people have strong incentives to influence information dissemination on these platforms and differentiate their content from the fierce competition. Existing dissemination strategies typically employ marketing techniques, such as seeking publicity through renowned actors or targeted advertising placements. Despite their various forms, most simply spread information to strengthen user impressions without conducting formal analyses of specific influence enhancement. And coupled with high costs, most fall short of expectations. To this end, we ingeniously formulate the task of social media dissemination as poisoning attacks, which influence specified content’s dissemination among target users by intervening in some users’ social media behaviors (including retweeting, following, and profile modifying). Correspondingly, we propose a novel poisoning attack, Influence-based Social Media Attack (ISMA) to generate discrete poisoning behaviors, which is difficult to achieve with existing attacks. In ISMA, we first contribute an efficient influence evaluator to quantify the spread influence of poisoning behaviors. Based on the estimated influence, we then present an imperceptible hierarchical selector and a profile modification method ProMix to select influential behaviors to poison. Notably, our attack is driven by custom attack objectives, which allows one to flexibly design different optimization goals to change the information flow, which could solve the blindness of existing influence maximization methods. Besides, behaviors such as retweeting are gentle and simple to implement. These properties make our attack more cost-effective and practical. Extensive experiments on two large-scale real-world datasets demonstrate the superiority of our method as it significantly outperforms baselines, and additionally, the proposed evaluator’s analysis of user influence provides new insights for influence maximization on social media.

Knowledge graphs have important applications for many computational tasks, such as personalized recommendations, information search, and natural language processing. Knowledge graph embedding, which is to learn representations of nodes and relations, is very critical to facilitate these applications and thus has been studied extensively in the literature. Most existing knowledge graphs (e.g., Freebase) have a data scarcity issue, i.e., the number of observed triplets is much less than that of all possible pairs of nodes. While the data augmentation technique has been widely applied to addressing data scarcity in other domains (e.g., image data), there are few prior studies exploring it for knowledge graph embedding probably because the discrete data structure of knowledge graphs prohibits the employment of most data augmentation methods. To fill this research gap, this paper introduces a novel data augmentation framework, namely knowledge graph mixup (KG Mixup), to enhance knowledge graph embedding. Based on the proposed framework, we develop two specific methods: vanilla mixup and influence mixup. Both approaches generate virtual mixup triplets and incorporate them into the learning process through a new mixup loss function. While vanilla mixup generates virtual triplets based on a uniform distribution, the influence mixup approach employs the influence function to guide the generation of mixup samples. Experiments with multiple datasets have shown that both approaches significantly outperform knowledge graph embedding models trained by the ordinary training framework.

Reinforcement learning (RL) has shown great promise in optimizing long-term user interest in recommender systems. However, existing RL-based recommendation methods need a large number of interactions for each user to learn the recommendation policy. The challenge becomes more critical when recommending to new users who have a limited number of interactions. To that end, in this article, we address the cold-start challenge in the RL-based recommender systems by proposing a novel context-aware offline meta-level model-based RL approach for user adaptation. Our proposed approach learns to infer each user's preference with a user context variable that enables recommendation systems to better adapt to new users with limited contextual information. To improve adaptation efficiency, our approach learns to recover the user choice function and reward from limited contextual information through an inverse RL method, which is used to assist the training of a meta-level recommendation agent. To avoid the need for online interaction, the proposed method is trained using historically collected offline data. Moreover, to tackle the challenge of offline policy training, we introduce a mutual information constraint between the user model and recommendation agent. Evaluation results show the superiority of our developed offline policy learning method when adapting to new users with limited contextual information. In addition, we provide a theoretical analysis of the recommendation performance bound.

Vector retrieval focuses on finding the k-nearest neighbors from a bunch of data points, and is widely used in a diverse set of areas such as information retrieval and recommender system. The current state-of-the-art methods represented by HNSW usually generate indexes with a big memory footprint, restricting the scale of data they can handle, except resorting to a hybrid index with external storage. The space-partitioning learned indexes, which only occupy a small memory, have made great breakthroughs in recent years. However, these methods rely on a large amount of labeled data for supervised learning, so model complexity affects the generalization.

In today's competitive and fast-evolving business environment, it is a critical time for organizations to rethink how to deal with the talent and management related tasks in a quantitative manner. Indeed, thanks to the era of big data, the availability of large-scale talent data provides unparalleled opportunities for business leaders to understand the rules of talent and management, which in turn deliver intelligence for effective decision making and management for their organizations. In the past few years, talent and management computing have increasingly attracted attentions from KDD communities, and a number of research/applied data science efforts have been devoted. To this end, the purpose of this workshop, i.e., the 4th International Workshop on Talent and Management Computing (TMC'2023), is to bring together researchers and practitioners to discuss both the critical problems faced by talent and management related domains, and potential data-driven solutions by leveraging state-of-the-art data mining technologies.

We study the problem of balancing effectiveness and efficiency in automated feature selection. Feature selection is to find an optimal feature subset from large feature space. After exploring many feature selection methods, we observe a computational dilemma: 1) traditional feature selection (e.g., mRMR) is mostly efficient, but difficult to identify the best subset; 2) the emerging reinforced feature selection automatically navigates feature space to search the best subset, but is usually inefficient. Are automation and efficiency always apart from each other? Can we bridge the gap between effectiveness and efficiency under automation? Motivated by this dilemma, we aim to develop a novel feature space navigation method. In our preliminary work, we leveraged interactive reinforcement learning to accelerate feature selection by external trainer-agent interaction. Our preliminary work can be significantly improved by modeling the structured knowledge of its downstream task (e.g., decision tree) as learning feedback. In this journal version, we propose a novel interactive and closed-loop architecture to simultaneously model interactive reinforcement learning (IRL) and decision tree feedback (DTF). Specifically, IRL is to create an interactive feature selection loop and DTF is to feed structured feature knowledge back to the loop. The DTF improves IRL from two aspects. First, the tree-structured feature hierarchy generated by decision tree is leveraged to improve state representation. In particular, we represent the selected feature subset as an undirected graph of feature-feature correlations and a directed tree of decision features. We propose a new embedding method capable of empowering Graph Convolutional Network (GCN) to jointly learn state representation from both the graph and the tree. Second, the tree-structured feature hierarchy is exploited to develop a new reward scheme. In particular, we personalize reward assignment of agents based on decision tree feature importance. In addition, observing agents’ actions can also be a feedback, we devise another new reward scheme, to weigh and assign reward based on the selected frequency ratio of each agent in historical action records. Finally, we present extensive experiments with real-world datasets to demonstrate the improved performances of our method.

Information network embedding is an important way to enable efficient graph analytics. However, it still faces with computational challenges in problems such as link prediction and node recommendation, particularly with the increasing scale of networks. Both hashing and quantization are promising approaches for accelerating these problems by orders of magnitude. In the preliminary work, we have proposed to learn binary codes for information networks, but graph analytics may suffer from large accuracy degradation. To reduce information loss while achieving memory and search efficiency, we further propose to learn quantized codes for information networks. In particular, each node is represented by compositing multiple latent vectors, each of which is optimally selected from a distinct set. Since (generalized) matrix factorization unifies several well-known embedding methods with high-order proximity preserved, we propose a \underline{N}etwork \underline{R}epresentation \underline{L}ightening framework based on \underline{M}atrix \underline{F}actorization (NRL-MF) to learn binary and quantized codes. We also propose an alternating optimization algorithm for efficient parameter learning, even for the generalized matrix factorization case. We finally evaluate NRL-MF on four real-world information network datasets with respect to the tasks of node classification and node recommendation. The results show that NRL-MF significantly outperforms competing baselines in both tasks, and that quantized representations indeed incur much smaller information loss than binarized codes.

The accumulated massive job and employee data at various platforms such as LinkedIn and Glassdoor are very valuable for many online applications such as job/employee search and recommendations. In order to exploit these data, an interesting and practical problem is how to learn effective job and employee representations, which could be further utilized for many computing tasks such as searching for similar jobs and employees. Yet this problem is very challenging because these user-generated job and employee data are semi-structured and created without standards, which makes them very messy, sparse, and difficult to model. Developing novel and advanced methods to learn job and employee representations has become an urgent need. To this end, in this paper, we develop a novel neural network model for job and employee embeddings. Our proposed approach consists of three components to model career data from three levels of granularity: job content, job context, and job sequence. We fine-tune a transformer model to learn the semantics of massive text in job content, build a shallow neural network to accumulate contextual information in job sequences, and develop an RNN encoder-decoder model to learn representations of employees’ career paths. To evaluate the proposed method, we conduct two experimental tasks: job similarity and employee similarity searches. The experimental results with a real-world dataset demonstrate the superiority of the developed approach.

Electric Vehicles (EVs) have been emerging as a promising low-carbon transport target. While a large number of public charging stations are available, the use of these stations is often imbalanced, causing many problems to Charging Station Operators (CSOs). To this end, in this paper, we propose a Multi-Agent Graph Convolutional Reinforcement Learning (MAGC) framework to enable CSOs to achieve more effective use of these stations by providing dynamic pricing for each of the continuously arising charging requests with optimizing multiple long-term commercial goals. Specifically, we first formulate this charging station request-specific dynamic pricing problem as a mixed competitive-cooperative multi-agent reinforcement learning task, where each charging station is regarded as an agent. Moreover, by modeling the whole charging market as a dynamic heterogeneous graph, we devise a multi-view heterogeneous graph attention networks to integrate complex interplay between agents induced by their diversified relationships. Then, we propose a shared meta generator to generate individual customized dynamic pricing policies for large-scale yet diverse agents based on the extracted meta characteristics. Finally, we design a contrastive heterogeneous graph pooling representation module to learn a condensed yet effective state action representation to facilitate policy learning of large-scale agents. Extensive experiments on two real-world datasets demonstrate the effectiveness of MAGC and empirically show that the overall use of stations can be improved if all the charging stations in a charging market embrace our dynamic pricing policy.

With an increasing popularity, Multiplayer Online Battle Arena (MOBA) games where two opposing teams compete against each other, have played a major role in E-sports tournaments. Among game analysis, real-time winning prediction is an important but challenging problem, which is mainly due to the complicated coupling of the overall Confrontation1, the excessive noise of the player’s Movement, and unclear optimization goals. Existing research is difficult to solve this problem in a dynamic, comprehensive and systematic way. In this study, we design a unified framework, namely Winning Tracker (WT), for solving this problem. Specifically, offense and defense extractors are developed to extract the Confrontation of both sides. A well-designed trajectory representation algorithm is applied to extracting individual’s Movement information. Moreover, we design a hierarchical attention mechanism to capture team-level strategies and facilitate the interpretability of the framework. To optimize accurately, we adopt a multi-task learning method to design short-term and long-term goals, which are used to represent immediate state and make end-state prediction respectively. Intensive experiments on a real-world data set demonstrate that our proposed method WT outperforms state-of-the-art algorithms. Furthermore, our work has been practically deployed in real MOBA games, and provided case studies reflecting its outstanding commercial value.

Crime poses a major threat to human life and property, which has been recognized as one of the most crucial problems in our society. Predicting the number of crime incidents in each region of a city before they happen is of great importance to fight against crime. There has been a great deal of research focused on crime prediction, ranging from introducing diversified data sources to exploring various prediction models. However, most of the existing approaches fail to offer fine-scale prediction results and take little notice of the intricate spatial-temporal-categorical correlations contained in crime incidents. In this article, we propose a tailor-made framework called CrimeTensor to predict the number of crime incidents belonging to different categories within each target region via tensor learning with spatiotemporal consistency. In particular, we model the crime data as a tensor and present an objective function which tries to take full advantage of the spatial, temporal, and categorical correlations contained in crime incidents. Moreover, a well-designed optimization algorithm which transforms the objective into a compact form and then applies CP decomposition to find the optimal solution is elaborated to solve the objective function. Furthermore, we develop an enhanced framework which takes a set of pre-selected regions to conduct prediction so as to further improve the computational efficiency of the optimization algorithm. Finally, extensive experiments are performed on both proprietary and public datasets and our framework significantly outperforms all the baselines in terms of each evaluation metric.

Social recommendation has emerged to leverage social connections among users for predicting users’ unknown preferences, which could alleviate the data sparsity issue in collaborative filtering based recommendation. Early approaches relied on utilizing each user’s first-order social neighbors’ interests for better user modeling, and failed to model the social influence diffusion process from the global social network structure. Recently, we propose a preliminary work of a neural influenceDiffusionNetwork (i.e., DiffNet) for social recommendation L. Wu, P. Sun, Y. Fu, R. Hong, X. Wang, and M. Wang, “A neural influence diffusion model for social recommendation,” inProc. Int. ACM SIGIR Conf. Res. Develop. Inf. Retrieval, 2019, pp. 235–244.. DiffNet models the recursive social diffusion process for each user, such that the influence diffusion hidden in the higher-order social network is captured in the user embedding process. Despite the superior performance of DiffNet, we argue that, as users play a central role in both user-user social network and user-item interest network, only modeling the influence diffusion process in the social network would neglect the latent collaborative interests of users hidden in the user-item interest network. To this end, in this paper, we propose DiffNet++, an improved algorithm of DiffNet that models the neural influence diffusion and interest diffusion in a unified framework. By reformulating the social recommendation as a heterogeneous graph with social network and interest network as input, DiffNet++ advances DiffNet by injecting both the higher-order user latent interest reflected in the user-item graph and higher-order user influence reflected in the user-user graph for user embedding learning. This is achieved by iteratively aggregating each user’s embedding from three aspects: the user’s previous embedding, the influence aggregation of social neighbors from the social network, and the interest aggregation of item neighbors from the user-item interest network. Furthermore, we design a multi-level attention network that learns how to attentively aggregate user embeddings from these three aspects. Finally, extensive experimental results on four real-world datasets clearly show the effectiveness of our proposed model. We release the source code athttps://github.com/PeiJieSun/diffnet.

Approximate nearest neighbor search (ANNs) plays an important role in many applications ranging from information retrieval, recommender systems to machine translation. Several ANN indexes, such as hashing and quantization, have been designed to update for the evolving database, but there exists a remarkable performance gap between them and retrained indexes on the entire database. To close the gap, we propose an online additive quantization algorithm (online AQ) to dynamically update quantization codebooks with the incoming streaming data. Then we derive the regret bound to theoretically guarantee the performance of the online AQ algorithm. Moreover, to improve the learning efficiency, we develop a randomized block beam search algorithm for assigning each data to the codewords of the codebook. Finally, we extensively evaluate the proposed online AQ algorithm on four real-world datasets, showing that it remarkably outperforms the state-of-the-art baselines.

As an important means to solve information overload, recommender systems have been widely applied in many fields, such as e-commerce and advertising. However, recent studies have shown that recommender systems are vulnerable to poisoning attacks; that is, injecting a group of carefully designed user profiles into the recommender system can severely affect recommendation quality. Despite the development from shilling attacks to optimization-based attacks, the imperceptibility and harmfulness of the generated data in most attacks are arduous to balance. To this end, we propose a triple adversarial learning for influence based poisoning attack (TrialAttack), a flexible end-to-end poisoning framework to generate non-notable and harmful user profiles. Specifically, given the input noise, TrialAttack directly generates malicious users through triple adversarial learning of the generator, discriminator, and influence module. Besides, to provide reliable influence for TrialAttack training, we explore a new approximation approach for estimating each fake user's influence. Through theoretical analysis, we prove that the distribution characterized by TrialAttack approximates to the rating distribution of real users under the premise of performing an efficient attack. This property allows the injected users to attack in an unremarkable way. Experiments on three real-world datasets show that TrialAttack's attack performance outperforms state-of-the-art attacks, and the generated fake profiles are more difficult to detect compared to baselines.

Recent studies have shown that recommender systems are vulnerable, and it is easy for attackers to inject well-designed malicious profiles into the system, leading to biased recommendations. We cannot deny these data's rationality, making it imperative to establish a robust recommender system. Adversarial training has been extensively studied for robust recommendations. However, traditional adversarial training adds small perturbations to the parameters (inputs), which do not comply with the poisoning mechanism in the recommender system. Thus for the practical models that are very good at learning existing data, it does not perform well. To address the above limitations, we propose adversarial poisoning training (APT). It simulates the poisoning process by injecting fake users (ERM users) who are dedicated to minimizing empirical risk to build a robust system. Besides, to generate ERM users, we explore an approximation approach to estimate each fake user's influence on the empirical risk. Although the strategy of "fighting fire with fire" seems counterintuitive, we theoretically prove that the proposed APT can boost the upper bound of poisoning robustness. Also, we deliver the first theoretical proof that adversarial training holds a positive effect on enhancing recommendation robustness. Through extensive experiments with five poisoning attacks on four real-world datasets, the results show that the robustness improvement of APT significantly outperforms baselines. It is worth mentioning that APT also improves model generalization in most cases.

The accumulated large amount of mobility data and the ability to track moving people or objects have enabled us to develop advanced mobile recommendations, which are essential to recommend a sequence of locations to an individual user on the move. In this paper, we study a particular case of mobile recommendations, route recommendations to drivers, by utilizing vehicle GPS data. Specifically, we formulate a new Route Recommendation with Relaxed Assumptions (RR-RA) problem, the goal of which is to recommend a sequence of locations to a driver based on his current location in order to maximize his business success. To make our recommendation practical and scalable for real practice, we need to produce recommendation results in a timely fashion once a request emerges. Therefore, we propose an efficient algorithm to efficiently generate recommendations. Furthermore, we identify and address a destination-oriented route recommendation (DORR) problem. Without solving DORR problem, RR-RA alone does not work well in practice because drivers may encounter the destination constraint on a daily basis. We develop a dedicated and efficient algorithm for solving DORR problem. The package of solutions for both RR-RA and DORR problems provide a comprehensive approach for route recommendations to drivers. We evaluate our methods using both real-world GPS data and synthetic data, and demonstrate the effectiveness and efficiency of proposed methods with different evaluation metrics.

Spatial co-location patterns are subsets of spatial features usually located together in geographic space. Recent literature has provided different approaches to discover co-location patterns over point spatial data. However, most approaches consider the neighborhood relationship among spatial objects as binary and are mainly designed for point spatial features, thus are not appropriate for extended spatial features such as line strings and polygons, the neighborhood relationship among which is naturally continuous. This paper adopts a buffer-based model for measuring the spatial relationship of extended objects and mining co-location patterns. While the buffer-based model has several advantages for extended spatial features, it involves high computational complexity due to the expensive buffer-level overlay operation. To tackle this challenge, we introduce a coarse-level co-location mining framework, which follows a filter-and-refine paradigm. Within the framework, we develop a serious of rigorous upper bounds based on geometric property and progressively prune search space with these upper bounds. Moreover, we develop a join-less schema to further reduce computation cost of size-k(k > 2) co-location patterns. Finally, we conduct experiments with large-scale spatial data to validate the efficiency of the developed algorithms against several state-of-art methods. All experimental results demonstrate the superiority of our methods.

Recommender systems play a fundamental role in web applications in filtering massive information and matching user interests. While many efforts have been devoted to developing more effective models in various scenarios, the exploration on the explainability of recommender systems is running behind. Explanations could help improve user experience and discover system defects. In this paper, after formally introducing the elements that are related to model explainability, we propose a novel explainable recommendation model through improving the transparency of the representation learning process. Specifically, to overcome the representation entangling problem in traditional models, we revise traditional graph convolution to discriminate information from different layers. Also, each representation vector is factorized into several segments, where each segment relates to one semantic aspect in data. Different from previous work, in our model, factor discovery and representation learning are simultaneously conducted, and we are able to handle extra attribute information and knowledge. In this way, the proposed model can learn interpretable and meaningful representations for users and items. Unlike traditional methods that need to make a trade-off between explainability and effectiveness, the performance of our proposed explainable model is not negatively affected after considering explainability. Finally, comprehensive experiments are conducted to validate the performance of our model as well as explanation faithfulness.

In this paper, we study the problem of balancing effectiveness and efficiency in automated feature selection. Feature selection is to find the optimal feature subset from large-scale feature space, and is a fundamental intelligence for machine learning and predictive analysis. After exploring many feature selection methods, we observe a computational dilemma: 1) traditional feature selection methods (e.g., K-Best, decision tree based ranking, mRMR) are mostly efficient, but difficult to identify the best subset; 2) the emerging reinforced feature selection methods automatically navigate feature space to explore the best subset, but are usually inefficient. Are automation and efficiency always apart from each other? Can we bridge the gap between effectiveness and efficiency under automation? Motivated by such a computational dilemma, this study is to develop a novel feature space navigation method. To that end, we propose an Interactive Reinforced Feature Selection (IRFS) framework that guides agents by not just self-exploration experience, but also diverse external skilled trainers to accelerate learning for feature exploration. Specifically, we formulate the feature selection problem into an interactive reinforcement learning framework. In this framework, we first model two trainers skilled at different searching strategies: (1) KBest based trainer; (2) Decision Tree based trainer. We then develop two strategies: (1) to identify assertive and hesitant agents to diversify agent training, and (2) to enable the two trainers to take the teaching role in different stages to fuse the experience of the trainers and diversify teaching process. Such a hybrid teaching strategy can help agents to learn broader knowledge, and thereafter be more effective. Finally, we present extensive experiments on real-world datasets to demonstrate the improved performances of our method: more efficient than reinforced selection and more effective than classic feature selection.

Sequential location recommendation plays an important role in many applications such as mobility prediction, route planning and location-based advertisements. In spite of evolving from tensor factorization to RNN-based neural networks, existing methods did not make effective use of geographical information and suffered from the sparsity issue. To this end, we propose a Geography-aware sequential recommender based on the Self-Attention Network (GeoSAN for short) for location recommendation. On the one hand, we propose a new loss function based on importance sampling for optimization, to address the sparsity issue by emphasizing the use of informative negative samples. On the other hand, to make better use of geographical information, GeoSAN represents the hierarchical gridding of each GPS point with a self-attention based geography encoder. Moreover, we put forward geography-aware negative samplers to promote the informativeness of negative samples. We evaluate the proposed algorithm with three real-world LBSN datasets, and show that GeoSAN outperforms the state-of-the-art sequential location recommenders by 34.9%. The experimental results further verify significant effectiveness of the new loss function, geography encoder, and geography-aware negative samplers.

With the emergencies of Web 2.0, online word of mouth (e-WOM) has become one of the most useful communication channels. Organizations increasingly rely on social media to receive customers’ feedback for product development. Despite the known effect of e-WOM on customers’ decision-making, few studies have examined the influence of e-WOM on product improvement. Extending the prior e-WOM research, this paper empirically investigates how e-WOM affects mobile application (App) improvement and examines the moderating effect of developer expertise. Based on a dataset obtained from Applause.com, we find that e-WOM volume, valence, and reviewer exposure positively affect App improvement. We also find that developer expertise negatively moderates the effect of e-WOM on App improvement. The results indicate that expert developers are less likely affected by e-WOM. This study contributes by empirically investigating the role of e-WOM used to facilitate customer-to-business communication. This study also provides practical value for firms active in the App market.

As an e-commerce feature, the personalized recommendation is invariably highly-valued by both consumers and merchants. The e-tourism has become one of the hottest industries with the adoption of recommendation systems. Several lines of evidence have confirmed the travel-product recommendation is quite different from traditional recommendations. Travel products are usually browsed and purchased relatively infrequently compared with other traditional products (e.g., books and food), which gives rise to the extreme sparsity of travel data. Meanwhile, the choice of a suitable travel product is affected by an army of factors such as departure, destination, and financial and time budgets. To address these challenging problems, in this article, we propose a Probabilistic Matrix Factorization with Multi-Auxiliary Information (PMF-MAI) model in the context of the travel-product recommendation. In particular, PMF-MAI is able to fuse the probabilistic matrix factorization on the user-item interaction matrix with the linear regression on a suite of features constructed by the multiple auxiliary information. In order to fit the sparse data, PMF-MAI is built by a whole-data based learning approach that utilizes unobserved data to increase the coupling between probabilistic matrix factorization and linear regression. Extensive experiments are conducted on a real-world dataset provided by a large tourism e-commerce company. PMF-MAI shows an overwhelming superiority over all competitive baselines on the recommendation performance. Also, the importance of features is examined to reveal the crucial auxiliary information having a great impact on the adoption of travel products.

Social networking services provide platforms for users to perform two kinds of behaviors: consumption behavior (e.g., recommending items of interest) and social link behavior (e.g., recommending potential social links). Accurately modeling and predicting users’ two kinds of behaviors are two core tasks in these platforms with various applications. Recently, with the advance of neural networks, many neural-based models have been designed to predict a single users’ behavior, i.e., social link behavior or consumption behavior. Compared to the classical shallow models, these neural-based models show better performance to drive a user’s behavior by modeling the complex patterns. However, there are few works exploiting whether it is possible to design a neural-based model to jointly predict users’ two kinds of behaviors to further enhance the prediction performance. In fact, social scientists have already shown that users’ two kinds of behaviors are not isolated; people trend to the consumption recommendation of friends on social platforms and would like to make new friends with like-minded users. While some previous works jointly model users’ two kinds of behaviors with shallow models, we argue that the correlation between users’ two kinds of behaviors are complex, which could not be well-designed with shallow linear models. To this end, in this article, we propose a neural joint behavior prediction model named Neural Joint Behavior Prediction Model (NJBP) to mutually enhance the prediction performance of these two tasks on social networking platforms. Specifically, there are two key characteristics of our proposed model: First, to model the correlation of users’ two kinds of behaviors, we design a fusion layer in the neural network to model the positive correlation of users’ two kinds of behaviors. Second, as the observed links in the social network are often very sparse, we design a new link-based loss function that could preserve the social network topology. After that, we design a joint optimization function to allow the two behaviors modeling tasks to be trained to mutually enhance each other. Finally, extensive experimental results on two real-world datasets show that our proposed method is on average 7.14% better than the best baseline on social link behavior while 6.21% on consumption behavior prediction. Compared with the pair-wise loss function on two datasets, our proposed link-based loss function improves at least 4.69% on the social link behavior prediction and 4.72% on the consumption behavior prediction.

We enhance the mobile sequential recommendation (MSR) model and address some critical issues in existing formulations by proposing three new forms of the MSR from a multi-user perspective. The multi-user MSR (MMSR) model searches optimal routes for multiple drivers at different locations while disallowing overlapping routes to be recommended. To enrich the properties of pick-up points in the problem formulation, we additionally consider the pick-up capacity as an important feature, leading to the following two modified forms of the MMSR: MMSR-m and MMSR-d. The MMSR-m sets a maximum pick-up capacity for all urban areas, while the MMSR-d allows the pick-up capacity to vary at different locations. We develop a parallel framework based on the simulated annealing to numerically solve the MMSR problem series. Also, a push-point method is introduced to improve our algorithms further for the MMSR-m and the MMSR-d, which can handle the route optimization in more practical ways. Our results on both real-world and synthetic data confirmed the superiority of our problem formulation and solutions under more demanding practical scenarios over several published benchmarks.

Online donation-based crowdfunding has brought new life to charity by soliciting small monetary contributions from crowd donors to help others in trouble or with dreams. However, a crucial issue for crowdfunding platforms as well as traditional charities is the problem of high donor attrition, i.e., many donors donate only once or very few times within a rather short lifecycle and then leave. Thus, it is an urgent task to analyze the factors of and then further predict the donors behaviors. Especially, we focus on two types of behavioral events, e.g., donation recurrence (whether one donor will make donations at some time slices in the future) and donor retention (whether she will remain on the crowdfunding platform until a future time). However, this problem has not been well explored due to many domain and technical challenges, such as the heterogeneous influence, the relevance of the two types of events, and the censoring phenomenon of retention records. In this paper, we present a focused study on donation recurrence and donor retention with the help of large-scale behavioral data collected from crowdfunding. Specifically, we propose a Joint Deep Survival model, i.e., JDS, which can integrate heterogeneous features, e.g., donor motives, projects recently donated to, social contacts, to jointly model the donation recurrence and donor retention since these two types of behavioral events are highly relevant. In addition, we model the censoring phenomenon and dependence relations of different behaviors from the survival analysis view by designing multiple innovative constraints and incorporating them into the objective functions. Finally, we conduct extensive analysis and validation experiments with large-scale data collected from Kiva.org. The experimental results clearly demonstrate the effectiveness of our proposed models for analyzing and predicting the donation recurrence and donor retention in crowdfunding.

The association-rule-based approach is one of the most common technologies for building recommender systems and it has been extensively adopted for commercial use. A variety of techniques, mainly including eligible rule selection and multiple rules combination, have been developed to create effective recommendation. Unfortunately, little attention has been paid to the scalability concern of rule-based recommendation methods. However, the computational complexity of rule-based methods shall increase drastically with the growth of both online customers and rules, which are usually several millions in typical e-commerce platforms. Moreover, the dynamic change of users’ actions requires rule-based methods make recommendations in nearly real-time, which further highlights the scalability issue of rule-based recommender systems. In this article, we present a distributed framework that can scale different association-rule-based recommendation methods in a unified way. Specifically, based on the summarization of existing rule-based approaches, a generic tree-type structure is defined to store separate kinds of patterns, and an efficient algorithm is designed for mining eligible patterns along with computing recommendation scores. To handle the ever-increasing number of online customers, a distributed framework is proposed, where two load-balanced strategies for partitioning tree are put forward to fit sparse and dense data, respectively. Extensive experiments on five real-life data sets demonstrate that the efficiency of association-rule-based recommender systems can be significantly improved by the proposed framework.

Real-time bidding (RTB) that features perimpression-level real-time ad auctions has become a popular practice in today's digital advertising industry. In RTB, click-through rate (CTR) prediction is a fundamental problem to ensure the success of an ad campaign and boost revenue. In this paper, we present a dynamic CTR prediction model designed for the Samsung demand-side platform (DSP). From our production data, we identify two key technical challenges that have not been fully addressed by the existing solutions: the dynamic nature of RTB and user information scarcity. To address both challenges, we develop a Dynamic Neural Network model. Our model effectively captures the dynamic evolutions of both users and ads and integrates auxiliary data sources (e.g., installed apps) to better model users' preferences. We put forward a novel interaction layer that fuses both explicit user responses (e.g., clicks on ads) and auxiliary data sources to generate consolidated user preference representations. We evaluate our model using a large amount of data collected from the Samsung advertising platform and compare our method against several state-of-the-art methods that are likely suitable for real-world deployment. The evaluation results demonstrate the effectiveness of our method and the potential for production. In addition, we discuss how to address a few practical engineering challenges caused by big data toward making our model in readiness for deployment.

Human mobility data refers to records of human movements, such as cellphone traces, vehicle GPS trajectories, geo-tagged posts, and photos. While successfully mining human mobility data can benefit many applications such as city planning, transportation, urban economics, and public safety, it is very challenging to model large-scale Heterogeneous Human Mobility Data (HHMD) that are generated from different resources. In this paper, we develop a general collective learning approach to model HHMD at an individual level towards identifying and quantifying the urban forms of residential communities. Specifically, our proposed method exploits two geographic regularities among HHMD. First, we jointly capture the correlations among residential communities, urban functions, temporal effects, and user mobility patterns by analogizing communities as documents and mobility patterns as words. Also, we further combine explicit LASSO analysis and significant testing into latent representation learning as a regularization term by analogizing compatible Point-of-Interests (POIs) as the meta-data of communities. In this way, we can learn the urban forms, including a mix of functions and corresponding portfolios, of residential communities from HHDM and POIs. We further leverage these learned results to address two application problems: real estate ranking and restaurant popularity prediction. Finally, we conduct intensive evaluations with a variety of real-world data, where experimental results demonstrate the effectiveness of our proposed modeling method and its successful applications for other problems.

We speed up the solution of the mobile sequential recommendation (MSR) problem that requires searching optimal routes for empty taxi cabs through mining massive taxi GPS data. We develop new methods that combine parallel computing and the simulated annealing with novel global and local searches. While existing approaches usually involve costly offline algorithms and methodical pruning of the search space, our new methods provide direct real-time search for the optimal route without the offline preprocessing. Our methods significantly reduce computational time for the high dimensional MSR problems from days to seconds based on the real-world data as well as the synthetic ones. We efficiently provide solutions to MSR problems with thousands of pick-up points without offline training, compared to the published record of 25 pick-up points.

Mobile gaming has emerged as a promising market with billion-dollar revenues. A variety of mobile game platforms and services have been developed around the world. One critical challenge for these platforms and services is to understand user churn behavior in mobile games. Accurate churn prediction will benefit many stakeholders such as game developers, advertisers, and platform operators. In this paper, we present the first large-scale churn prediction solution for mobile games. In view of the common limitations of the state-of-the-art methods built upon traditional machine learning models, we devise a novel semi-supervised and inductive embedding model that jointly learns the prediction function and the embedding function for user-app relationships. We model these two functions by deep neural networks with a unique edge embedding technique that is able to capture both contextual information and relationship dynamics. We also design a novel attributed random walk technique that takes into consideration both topological adjacency and attribute similarities. To evaluate the performance of our solution, we collect real-world data from the Samsung Game Launcher platform that includes tens of thousands of games and hundreds of millions of user-app interactions. The experimental results with this data demonstrate the superiority of our proposed model against existing state-of-the-art methods.

Information network embedding is an effective way for efficient graph analytics. However, it still faces with computational challenges in problems such as link prediction and node recommendation, particularly with increasing scale of networks. Hashing is a promising approach for accelerating these problems by orders of magnitude. However, no prior studies have been focused on seeking binary codes for information networks to preserve high-order proximity. Since matrix factorization (MF) unifies and outperforms several well-known embedding methods with high-order proximity preserved, we propose a MF-based \underlineI nformation \underlineN etwork \underlineH ashing (INH-MF) algorithm, to learn binary codes which can preserve high-order proximity. We also suggest Hamming subspace learning, which only updates partial binary codes each time, to scale up INH-MF. We finally evaluate INH-MF on four real-world information network datasets with respect to the tasks of node classification and node recommendation. The results demonstrate that INH-MF can perform significantly better than competing learning to hash baselines in both tasks, and surprisingly outperforms network embedding methods, including DeepWalk, LINE and NetMF, in the task of node recommendation. The source code of INH-MF is available online\footnote\urlhttps://github.com/DefuLian/network .

The classic mobile sequential recommendation (MSR) problem aims to provide the optimal route to taxi drivers for minimizing the potential travel distance before they meet next passengers. However, the problem is designed from the view of a single user and may lead to overlapped recommendations and cause traffic problems. Existing approaches usually contain an offline pruning process with extremely high computational cost, given a large number of pick-up points. To this end, we formalize a new multi-user MSR (MMSR) problem that locates optimal routes for a group of drivers with different starting positions. We develop two efficient methods, PSAD and PSAD-M, for solving the MMSR problem by ganging parallel computing and simulated annealing. Our methods outperform several existing approaches, especially for high-dimensional MMSR problems, with a record-breaking performance of 180x speedup using 384 cores.

Location recommendation plays an essential role in helping people find attractive places. Though recent research has studied how to recommend locations with social and geographical information, few of them addressed the cold-start problem of new users. Because mobility records are often shared on social networks, semantic information can be leveraged to tackle this challenge. A typical method is to feed them into explicit-feedback-based content-aware collaborative filtering, but they require drawing negative samples for better learning performance, as users’ negative preference is not observable in human mobility. However, prior studies have empirically shown sampling-based methods do not perform well. To this end, we propose a scalable Implicit-feedback-based Content-aware Collaborative Filtering (ICCF) framework to incorporate semantic content and to steer clear of negative sampling. We then develop an efficient optimization algorithm, scaling linearly with data size and feature size, and quadratically with the dimension of latent space. We further establish its relationship with graph Laplacian regularized matrix factorization. Finally, we evaluate ICCF with a large-scale LBSN dataset in which users have profiles and textual content. The results show that ICCF outperforms several competing baselines, and that user information is not only effective for improving recommendations but also coping with cold-start scenarios.

As the worlds of commerce and the Internet technology become more inextricably linked, a large number of user consumption series become available for online market intelligence analysis. A critical demand along this line is to predict the future product adoption state of each user, which enables a wide range of applications such as targeted marketing. Nevertheless, previous works only aimed at predicting if a user would adopt a particular product or not with a binary buy-or-not representation. The problem of tracking and predicting users' adoption rates, i.e., the frequency and regularity of using each product over time, is still under-explored. To this end, we present a comprehensive study of product adoption rate prediction in a competitive market. This task is nontrivial as there are three major challenges in modeling users' complex adoption states: the heterogeneous data sources around users, the unique user preference and the competitive product selection. To deal with these challenges, we first introduce a flexible factor-based decision function to capture the change of users' product adoption rate over time, where various factors that may influence users' decisions from heterogeneous data sources can be leveraged. Using this factor-based decision function, we then provide two corresponding models to learn the parameters of the decision function with both generalized and personalized assumptions of users' preferences. We further study how to leverage the competition among different products and simultaneously learn product competition and users' preferences with both generalized and personalized assumptions. Finally, extensive experiments on two real-world datasets show the superiority of our proposed models.

Location recommendation is an important means to help people discover attractive locations. However, extreme sparsity of user-location matrices leads to a severe challenge, so it is necessary to take implicit feedback characteristics of user mobility data into account and leverage the location’s spatial information. To this end, based on previously developed GeoMF, we propose a scalable and flexible framework, dubbed GeoMF++, for joint geographical modeling and implicit feedback-based matrix factorization. We then develop an efficient optimization algorithm for parameter learning, which scales linearly with data size and the total number of neighbor grids of all locations. GeoMF++ can be well explained from two perspectives. First, it subsumes two-dimensional kernel density estimation so that it captures spatial clustering phenomenon in user mobility data; Second, it is strongly connected with widely used neighbor additive models, graph Laplacian regularized models, and collective matrix factorization. Finally, we extensively evaluate GeoMF++ on two large-scale LBSN datasets. The experimental results show that GeoMF++ consistently outperforms the state-of-the-art and other competing baselines on both datasets in terms of NDCG and Recall. Besides, the efficiency studies show that GeoMF++ is much more scalable with the increase of data size and the dimension of latent space.

DanMu , an emerging type of user-generated comment, has become increasingly popular in recent years. Many online video platforms such as Tudou.com have provided the DanMu function. Unlike traditional online reviews such as reviews at Youtube.com that are outside the videos, DanMu is a scrolling marquee comment, which is overlaid directly on top of the video and synchronized to a specific playback time. Such comments are displayed as streams of moving subtitles overlaid on the video screen. Viewers could easily write DanMu s while watching videos, and the written DanMu s will be immediately overlaid onto the video and displayed to writers themselves and other viewers as well. Such DanMu systems have greatly enabled users to communicate with each other in a much more direct way, creating a real-time sharing experience. Although there are several unique features of DanMu and has had a great impact on online video systems, to the best of our knowledge, there is no work that has provided a comprehensive study on DanMu . In this article, as a pilot study, we analyze the unique characteristics of DanMu from various perspectives. Specifically, we first illustrate some unique distributions of DanMu s by comparing with traditional reviews (TReviews) that we collected from a real DanMu -enabled online video system. Second, we discover two interesting patterns in DanMu data: a herding effect and multiple-burst phenomena that are significantly different from those in TRviews and reveal important insights about the growth of DanMu s on a video. Towards exploring antecedents of both th herding effect and multiple-burst phenomena, we propose to further detect leading DanMu s within bursts, because those leading DanMu s make the most contribution to both patterns. A framework is proposed to detect leading DanMu s that effectively combines multiple factors contributing to leading DanMu s. Based on the identified characteristics of DanMu , finally we propose to predict the distribution of future DanMu s (i.e., the growth of DanMu s), which is important for many DanMu -enabled online video systems, for example, the predicted DanMu distribution could be an indicator of video popularity. This prediction task includes two aspects: One is to predict which videos future DanMu s will be posted for, and the other one is to predict which segments of a video future DanMu s will be posted on. We develop two sophisticated models to solve both problems. Finally, intensive experiments are conducted with a real-world dataset to validate all methods developed in this article.

The study of career development has become more important during a time of rising competition. Even with the help of newly available big data in the field of human resources, it is challenging to prospect the career development of talents in an effective manner, since the nature and structure of talent careers can change quickly. To this end, in this paper, we propose a novel survival analysis approach to model the talent career paths, with a focus on two critical issues in talent management, namely turnover and career progression. Specifically, for modeling the talent turnover behaviors, we formulate the prediction of survival status at a sequence of time intervals as a multi-task learning problem by considering the prediction at each time interval as a task. Also, we impose the ranking constraints to model both censored and uncensored data, and capture the intrinsic properties exhibited in general lifetime modeling with non-recurrent and recurrent events. Similarly, for modeling the talent career progression, each task concerns the prediction of a relative occupational level at each time interval. The ranking constraints imposed on different occupational levels can help to reduce the prediction error. Finally, we evaluate our approach with several state-of-the-art baseline methods on real-world talent data. The experimental results clearly demonstrate the effectiveness of the proposed models for predicting the turnover and career progression of talents.

Neural network based sequence-to-sequence models in an encoder-decoder framework have been successfully applied to solve Question Answering (QA) problems, predicting answers from statements and questions. However, almost all previous models have failed to consider detailed context information and unknown states under which systems do not have enough information to answer given questions. These scenarios with incomplete or ambiguous information are very common in the setting of Interactive Question Answering (IQA). To address this challenge, we develop a novel model, employing context-dependent word-level attention for more accurate statement representations and question-guided sentence-level attention for better context modeling. We also generate unique IQA datasets to test our model, which will be made publicly available. Employing these attention mechanisms, our model accurately understands when it can output an answer or when it requires generating a supplementary question for additional input depending on different contexts. When available, user's feedback is encoded and directly applied to update sentence-level attention to infer an answer. Extensive experiments on QA and IQA datasets quantitatively demonstrate the effectiveness of our model with significant improvement over state-of-the-art conventional QA models.

Precisely recommending relevant items from massive candidates to a large number of users is an indispensable yet computationally expensive task in many online platforms (e.g., Amazon.com and Netflix.com). A promising way is to project users and items into a Hamming space and then recommend items via Hamming distance. However, previous studies didn't address the cold-start challenges and couldn't make the best use of preference data like implicit feedback. To fill this gap, we propose a Discrete Content-aware Matrix Factorization (DCMF) model, 1) to derive compact yet informative binary codes at the presence of user/item content information; 2) to support the classification task based on a local upper bound of logit loss; 3) to introduce an interaction regularization for dealing with the sparsity issue. We further develop an efficient discrete optimization algorithm for parameter learning. Based on extensive experiments on three real-world datasets, we show that DCFM outperforms the state-of-the-arts on both regression and classification tasks.

Crowdfunding is an emerging Internet fundraising mechanism by raising monetary contributions from the crowd for projects or ventures. In these platforms, the dynamics, i.e., daily funding amount on campaigns and perks (backing options with rewards), are the most concerned issue for creators, backers and platforms. However, tracking the dynamics in crowdfunding is very challenging and still under-explored. To that end, in this paper, we present a focused study on this important problem. A special goal is to forecast the funding amount for a given campaign and its perks in the future days. Specifically, we formalize the dynamics in crowdfunding as a hierarchical time series, i.e., campaign level and perk level. Specific to each level, we develop a special regression by modeling the decision making process of the crowd (visitors and backing probability) and exploring various factors that impact the decision; on this basis, an enhanced switching regression is proposed at each level to address the heterogeneity of funding sequences. Further, we employ a revision matrix to combine the two-level base forecasts for the final forecasting. We conduct extensive experiments on a real-world crowdfunding data collected from Indiegogo.com. The experimental results clearly demonstrate the effectiveness of our approaches on tracking the dynamics in crowdfunding.

P2P lending is an emerging Internet-based application where individuals can directly borrow money from each other. The past decade has witnessed the rapid development and prevalence of online P2P lending platforms, examples of which include Prosper, LendingClub, and Kiva. Meanwhile, extensive research has been done that mainly focuses on the studies of platform mechanisms and transaction data. In this article, we provide a comprehensive survey on the research about P2P lending, which, to the best of our knowledge, is the first focused effort in this field. Specifically, we first provide a systematic taxonomy for P2P lending by summarizing different types of mainstream platforms and comparing their working mechanisms in detail. Then, we review and organize the recent advances on P2P lending from various perspectives (e.g., economics and sociology perspective, and data-driven perspective). Finally, we propose our opinions on the prospects of P2P lending and suggest some future research directions in this field. Meanwhile, throughout this paper, some analysis on real-world data collected from Prosper and Kiva are also conducted.

Social networking services have been prevalent at many online communities such as Twitter.com and Weibo.com, where millions of users keep interacting with each other every day. One interesting and important problem in the social networking services is to rank users based on their vitality in a timely fashion. An accurate ranking list of user vitality could benefit many parties in social network services such as the ads providers and site operators. Although it is very promising to obtain a vitality-based ranking list of users, there are many technical challenges due to the large scale and dynamics of social networking data. In this paper, we propose a unique perspective to achieve this goal, which is quantifying user vitality by analyzing the dynamic interactions among users on social networks. Examples of social network include but are not limited to social networks in microblog sites and academical collaboration networks. Intuitively, if a user has many interactions with his friends within a time period and most of his friends do not have many interactions with their friends simultaneously, it is very likely that this user has high vitality. Based on this idea, we develop quantitative measurements for user vitality and propose our first algorithm for ranking users based vitality. Also, we further consider the mutual influence between users while computing the vitality measurements and propose the second ranking algorithm, which computes user vitality in an iterative way. Other than user vitality ranking, we also introduce a vitality prediction problem, which is also of great importance for many applications in social networking services. Along this line, we develop a customized prediction model to solve the vitality prediction problem. To evaluate the performance of our algorithms, we collect two dynamic social network data sets. The experimental results with both data sets clearly demonstrate the advantage of our ranking and prediction methods.

Advances in real-time location systems have enabled us to collect massive amounts of fine-grained semantically rich location traces, which provide unparalleled opportunities for understanding human activities and generating useful knowledge. This, in turn, delivers intelligence for real-time decision making in various fields, such as workflow management. Indeed, it is a new paradigm to model workflows through knowledge discovery in location traces. To that end, in this paper, we provide a focused study of workflow modeling by integrated analysis of indoor location traces in the hospital environment. In particular, we develop a workflow modeling framework that automatically constructs the workflow states and estimates the parameters describing the workflow transition patterns. More specifically, we propose effective and efficient regularizations for modeling the indoor location traces as stochastic processes. First, to improve the interpretability of the workflow states, we use the geography relationship between the indoor rooms to define a prior of the workflow state distribution. This prior encourages each workflow state to be a contiguous region in the building. Second, to further improve the modeling performance, we show how to use the correlation between related types of medical devices to reinforce the parameter estimation for multiple workflow models. In comparison with our preliminary work [11], we not only develop an integrated workflow modeling framework applicable to general indoor environments, but also improve the modeling accuracy significantly. We reduce the average log-loss by up to 11 percent.

Sociologists have long converged that the evolution of a Social Networking Service(SNS) is driven by the interplay between users' preferences (reflected in user-item interaction behavior) and the social network structure (reflected in user-user interaction behavior). Nevertheless, traditional approaches either modeled these two kinds of behaviors in isolation or relied on a static assumption of a SNS. Thus, it is still unclear how do the roles of the dynamic social network structure and users' historical preferences affect the evolution of SNSs. Furthermore, can transforming the underlying social theories in the platform evolution modeling process benefit both behavior prediction tasks? In this paper, we incorporate the underlying social theories to explain and model the evolution of users' two kinds of behaviors in SNSs. Specifically, we present two kinds of representations for users' behaviors: a direct (latent) representation that presumes users' behaviors are represented directly (latently) by their historical behaviors. Under each representation, we associate each user's two kinds of behaviors with two vectors at each time. Then, for each representation, we propose the corresponding learning model to fuse the interplay between users' two kinds of behaviors. Finally, extensive experimental results demonstrate the effectiveness of our proposed models for both user preference prediction and social link suggestion.

Although weighted tensor factorization tailored to implicit feedback has shown its superior performance in temporal-aware location recommendation, it suffers from three critical challenges. First, it doesn't distinguish the confidence of negative preference for time-dependent unvisited locations from that for fully unvisited ones. Second, discontinuity arises from time discretization, and thus an infinitely large margin may exist between different bins of time. Third, geographical constraints of neighbor locations are not taken into account. To address these challenges, we propose a regularized content-aware tensor factorization (RCTF) algorithm, which exploits three strategies to address the corresponding challenges. First, it introduces a novel interaction regularization, second, it represents each bin of time by a derived feature vector from eigen decomposition of a time-bin similarity matrix, to capture the proximity of neighbor bins of time, third, it encodes geographical information of locations by discrete spatial distributions, so that spatial proximity constraints can be satisfied by simply feeding them into location content. The proposed algorithm is then evaluated for time-aware location recommendation on two large scale location-based social network datasets. The experimental results show the superiority of the proposed algorithm to several competing time-aware recommendation baselines, and verify the significant benefit of three strategies in the proposed algorithm.

Group-oriented services such as group recommendations aim to provide services for a group of users. For these applications, how to aggregate the preferences of different group members is the toughest yet most important problem. Inspired by game theory, in this paper, we propose to explore the idea of Nash equilibrium to simulate the selections of members in a group by a game process. Along this line, we first compute the preferences (group-dependent optimal selections) of each individual member in a given group scene, i.e., an equilibrium solution of this group, with the help of two pruning approaches. Then, to get the aggregated unitary preference of each group from all group members, we design a matrix factorization-based method which aggregates the preferences in latent space and estimates the final group preference in rating space. After obtaining the group preference, group-oriented services (e.g., group recommendation) can be directly provided. Finally, we construct extensive experiments on two real-world data sets from multiple aspects. The results clearly demonstrate the effectiveness of our method.

The emergence of Location-based Social Network (LBSN) services provides a wonderful opportunity to build personalized Point-of-Interest (POI) recommender systems. Although a personalized POI recommender system can significantly facilitate users' outdoor activities, it faces many challenging problems, such as the hardness to model user's POI decision making process and the difficulty to address data sparsity and user/location cold-start problem. To cope with these challenges, we define three types of friends (i.e., social friends, location friends, and neighboring friends) in LBSN, and develop a two-step framework to leverage the information of friends to improve POI recommendation accuracy and address cold-start problem. Specifically, we first propose to learn a set of potential locations that each individual's friends have checked-in before and this individual is most interested in. Then we incorporate three types of check-ins (i.e., observed check-ins, potential check-ins and other unobserved check-ins) into matrix factorization model using two different loss functions (i.e., the square error based loss and the ranking error based loss). To evaluate the proposed model, we conduct extensive experiments with many state-of-the-art baseline methods and evaluation metrics on two real-world data sets. The experimental results demonstrate the effectiveness of our methods.

P2P lending is an emerging wealth-management service for individuals, which allows lenders to directly bid and invest on the loans created by borrowers. In these platforms, lenders often pursue multiple objectives (e.g., non-default probability, fully-funded probability and winning-bid probability) when they select loans to invest. How to automatically assess loans from these objectives and help lenders select loan portfolios is a very important but challenging problem. To that end, in this paper, we present a holistic study on portfolio selections in P2P lending. Specifically, we first propose to adapt gradient boosting decision tree, which combines both static features and dynamic features, to assess loans from multiple objectives. Then, we propose two strategies, i.e., weighted objective optimization strategy and multi-objective optimization strategy, to select portfolios for lenders. For each lender, the first strategy attempts to provide one optimal portfolio while the second strategy attempts to provide a Pareto-optimal portfolio set. Further, we design two algorithms, namely DPA and EVA, which can efficiently resolve the optimizations in these two strategies, respectively. Finally, extensive experiments on a large-scale real-world data set demonstrate the effectiveness of our solutions.

Days on Market (DOM) refers to the number of days a property is on the active market, which is an important measurement of market liquidity in real estate industry. Indeed, at the micro level, DOM is not only a special concern of house sellers, but also a useful indicator for potential buyers to evaluate the popularity of a house. At the macro level, DOM is an important indicator of real estate market status. However, it is very challenging to measure DOM, since there are a variety of factors which can impact on the DOM of a property. To this end, in this paper, we aim to measure real estate liquidity by examining multiple factors in a holistic manner. A special goal is to predict the DOM of a given property listing. Specifically, we first extract key features from multiple types of heterogeneous real estate-related data, such as house profiles and geo-social information of residential communities. Then, based on these features, we develop a multi-task learning based regression approach for predicting the DOM of real estates. This approach can effectively learn district-aware models for different property listings by considering multiple factors. Finally, we conduct extensive experiments on real-world real estate data collected in Beijing and develop a prototype system for practical use. The experimental results clearly validate the effectiveness of the proposed approach for measuring liquidity in real estate markets.

With the rapid development of Location-based Social Network (LBSN) services, a large number of Point-of-Interests (POIs) have been available, which consequently raises a great demand of building personalized POI recommender systems. A personalized POI recommender system can significantly help users to find their preferred POIs and assist POI owners to attract more customers. However, due to the complexity of users' checkin decision making process that is influenced by many different factors such as POI distance and region's prosperity, and the dynamics of user's preference, POI recommender systems usually suffer from many challenges. Although different latent factor based methods (e.g., probabilistic matrix factorization) have been proposed, most of them do not successfully incorporate both geographical influence and temporal effect together into latent factor models. To this end, in this paper, we propose a new Spatial-Temporal Probabilistic Matrix Factorization (STPMF) model that models a user's preference for POI as the combination of his geographical preference and other general interest in POI. Furthermore, in addition to static general interest of user, we capture the temporal dynamics of user's interest as well by modeling checkin data in a unique way. To evaluate the proposed STPMF model, we conduct extensive experiments with many state-of-the-art baseline methods and evaluation metrics on two real-world data sets. The experimental results clearly demonstrate the effectiveness of our proposed STPMF model.

It is traditionally a challenge for home buyers to understand, compare, and contrast the investment value of real estate. Although a number of appraisal methods have been developed to value real properties, the performances of these methods have been limited by traditional data sources for real estate appraisal. With the development of new ways of collecting estate-related mobile data, there is a potential to leverage geographic dependencies of real estate for enhancing real estate appraisal. Indeed, the geographic dependencies of the investment value of an estate can be from the characteristics of its own neighborhood (individual), the values of its nearby estates (peer), and the prosperity of the affiliated latent business area (zone). To this end, in this paper, we propose a geographic method, named ClusRanking, for real estate appraisal by leveraging the mutual enforcement of ranking and clustering power. ClusRanking is able to exploit geographic individual, peer, and zone dependencies in a probabilistic ranking model. Specifically, we first extract the geographic utility of estates from geography data, estimate the neighborhood popularity of estates by mining taxicab trajectory data, and model the influence of latent business areas. Also, we fuse these three influential factors and predict real estate investment value. Moreover, we simultaneously consider individual, peer and zone dependencies, and derive an estate-specific ranking likelihood as the objective function. Furthermore, we propose an improved method named CR-ClusRanking by incorporating checkin information as a regularization term which reduces the performance volatility of real estate ranking system. Finally, we conduct a comprehensive evaluation with the real estate-related data of Beijing, and the experimental results demonstrate the effectiveness of our proposed methods.

With the rapid development of Location-based Social Network (LBSN) services, a large number of Point-Of-Interests (POIs) have been available, which consequently raises a great demand of building personalized POI recommender systems. A personalized POI recommender system can significantly assist users to find their preferred POIs and help POI owners to attract more customers. However, it is very challenging to develop a personalized POI recommender system because a user's checkin decision making process is very complex and could be influenced by many factors such as social network and geographical distance. In the literature, a variety of methods have been proposed to tackle this problem. Most of these methods model user's preference for POIs with integrated approaches and consider all candidate POIs as a whole space. However, by carefully examining a longitudinal real-world checkin data, we find that the whole space of users' checkins actually consists of two parts: social friend space and user interest space. The social friend space denotes the set of POI candidates that users' friends have checked-in before and the user interest space refers to the set of POI candidates that are similar to users' historical checkins, but are not visited by their friends yet. Along this line, we develop separate models for the both spaces to recommend POIs. Specifically, in social friend space, we assume users would repeat their friends' historical POIs due to the preference propagation through social networks, and propose a new Social Friend Probabilistic Matrix Factorization (SFPMF) model. In user interest space, we propose a new User Interest Probabilistic Matrix Factorization (UIPMF) model to capture the correlations between a new POI and one user's historical POIs. To evaluate the proposed models, we conduct extensive experiments with many state-of-the-art baseline methods and evaluation metrics on the real-world data set. The experimental results firmly demonstrate the effectiveness of our proposed models.

A large number of online reviews have been accumulated on the Web, such as Amazon.com and Cnet.com. It is increasingly challenging to digest these reviews for both consumers and firms as the volume of reviews increases. A promising direction to ease such a burden is to automatically identify aspects of a product and reveal each individual's ratings on them from these reviews. The identified and rated aspects can help consumers understand the pros and cons of a product and make their purchase decisions, and help firms learn user feedbacks and improve their products and marketing strategy. While different methods have been introduced to tackle this problem in the past, few of them successfully model the intrinsic connection between aspect and aspect rating particularly in short reviews. To this end, in this paper, we first propose the Aspect Identification and Rating (AIR) model to model observed textual reviews and overall ratings in a generative way, where the sampled aspect rating influences the sampling of sentimental words on this aspect. Furthermore, we enhance AIR model to particularly address one unique characteristic of short reviews that aspects mentioned in reviews may be quite unbalanced, and develop another model namely AIRS. Within AIRS model, we allow an aspect to directly affect the sampling of a latent rating on this aspect in order to capture the mutual influence between aspect and aspect rating through the whole generative process. Finally, we examine our two models and compare them with other methods based on multiple real world data sets, including hotel reviews, beer reviews and app reviews. Experimental results clearly demonstrate the effectiveness and improvement of our models. Other potential applications driven by our results are also shown in the experiments.

Location recommendation plays an essential role in helping people find places they are likely to enjoy. Though some recent research has studied how to recommend locations with the presence of social network and geographical information, few of them addressed the cold-start problem, specifically, recommending locations for new users. Because the visits to locations are often shared on social networks, rich semantics (e.g., tweets) that reveal a person's interests can be leveraged to tackle this challenge. A typical way is to feed them into traditional explicit-feedback content-aware recommendation methods (e.g., LibFM). As a user's negative preferences are not explicitly observable in most human mobility data, these methods need draw negative samples for better learning performance. However, prior studies have empirically shown that sampling-based methods don't perform as well as a method that considers all unvisited locations as negative but assigns them a lower confidence. To this end, we propose an Implicit-feedback based Content-aware Collaborative Filtering (ICCF) framework to incorporate semantic content and steer clear of negative sampling. For efficient parameter learning, we develop a scalable optimization algorithm, scaling linearly with the data size and the feature size. Furthermore, we offer a good explanation to ICCF, such that the semantic content is actually used to refine user similarity based on mobility. Finally, we evaluate ICCF with a large-scale LBSN dataset where users have profiles and text content. The results show that ICCF outperforms LibFM of the best configuration, and that user profiles and text content are not only effective at improving recommendation but also helpful for coping with the cold-start problem.

Mixed land use refers to the effort of putting residential, commercial and recreational uses in close proximity to one another. This can contribute economic benefits, support viable public transit, and enhance the perceived security of an area. It is naturally promising to investigate how to rank real estate from the viewpoint of diverse mixed land use, which can be reflected by the portfolio of community functions in the observed area. To that end, in this paper, we develop a geographical function ranking method, named FuncDivRank, by incorporating the functional diversity of communities into real estate appraisal. Specifically, we first design a geographic function learning model to jointly capture the correlations among estate neighborhoods, urban functions, temporal effects, and user mobility patterns. In this way we can learn latent community functions and the corresponding portfolios of estates from human mobility data and Point of Interest (POI) data. Then, we learn the estate ranking indicator by simultaneously maximizing ranking consistency and functional diversity, in a unified probabilistic optimization framework. Finally, we conduct a comprehensive evaluation with real-world data. The experimental results demonstrate the enhanced performance of the proposed method for real estate appraisal.

With the rapid development of smart TV industry, a large number of TV programs have been available for meeting various user interests, which consequently raise a great demand of building personalized TV recommender systems. Indeed, a personalized TV recommender system can greatly help users to obtain their preferred programs and assist TV and channel providers to attract more audiences. While different methods have been proposed for TV recommendations, most of them neglect the mixture of watching groups behind an individual TV. In other words, there may be different groups of audiences at different times in front of a TV. For instance, watching groups of a TV may consist of children, wife and husband, husband, wife, etc in many US household. To this end, in this paper, we propose a Mixture Probabilistic Matrix Factorization (mPMF) model to learn the program preferences of televisions, which assumes that the preference of a given television can be regarded as the mixed preference of different watching groups. Specifically, the latent vector of a television is drawn from a mixture of Gaussian and the mixture number is the estimated number of watching groups behind the television. To evaluate the proposed mPMF model, we conduct extensive experiments with many state-of-the-art baseline methods and evaluation metrics on a real-world data set. The experimental results clearly demonstrate the effectiveness of our model.

Ranking fraud in the mobile App market refers to fraudulent or deceptive activities which have a purpose of bumping up the Apps in the popularity list. Indeed, it becomes more and more frequent for App developers to use shady means, such as inflating their Apps' sales or posting phony App ratings, to commit ranking fraud. While the importance of preventing ranking fraud has been widely recognized, there is limited understanding and research in this area. To this end, in this paper, we provide a holistic view of ranking fraud and propose a ranking fraud detection system for mobile Apps. Specifically, we first propose to accurately locate the ranking fraud by mining the active periods, namely leading sessions, of mobile Apps. Such leading sessions can be leveraged for detecting the local anomaly instead of globalanomaly of App rankings. Furthermore, we investigate three types of evidences, i.e., ranking based evidences, rating based evidences and review based evidences, by modeling Apps' ranking, rating and review behaviors through statistical hypotheses tests. In addition, we propose an optimization based aggregation method to integrate all the evidences for fraud detection. Finally, we evaluate the proposed system with real-world App data collected from the iOS App Store for a long time period. In the experiments, we validate the effectiveness of the proposed system, and show the scalability of the detection algorithm as well as some regularity of ranking fraud activities.

Recent years have witnessed the rapid prevalence of online serials, which play an important role in our daily entertainment. A critical demand along this line is to predict the popularity of online serials, which can enable a wide range of applications, such as online advertising, and serial recommendation. However, compared with traditional online media such as user-generated content (UGC), online serials have unique characteristics of sequence dependence, release date dependence as well as unsynchronized update regularity. Therefore, the popularity prediction for online serials is a nontrivial task and still under-addressed. To this end, in this paper we present a comprehensive study for predicting the popularity of online serials with autoregressive models. Specifically, we first introduce a straightforward yet effective Naive Autoregressive (NAR) model based on the correlations of serial episodes. Furthermore, we develop a sophisticated model, namely Transfer Autoregressive (TAR) model, to capture the dynamic behaviors of audiences, which can achieve better prediction performance than the NAR model. Indeed, the two models can reveal the popularity generation from different perspectives. In addition, as a derivative of the TAR model, we also design a novel metric, namely favor, for evaluating the quality of online serials. Finally, extensive experiments on two real-world data sets clearly show that both models are effective and outperform baselines in terms of the popularity prediction for online serials. And the new metric performs better than other metrics for quality estimation.

Ranking residential real estates based on investment values can provide decision making support for home buyers and thus plays an important role in estate marketplace. In this paper, we aim to develop methods for ranking estates based on investment values by mining users' opinions about estates from online user reviews and offline moving behaviors (e.g., Taxi traces, smart card transactions, check-ins). While a variety of features could be extracted from these data, these features are Interco related and redundant. Thus, selecting good features and integrating the feature selection into the fitting of a ranking model are essential. To this end, in this paper, we first strategically mine the fine-grained discrminative features from user reviews and moving behaviors, and then propose a probabilistic sparse pair wise ranking method for estates. Specifically, we first extract the explicit features from online user reviews which express users' opinions about point of interests (POIs) near an estate. We also mine the implicit features from offline moving behaviors from multiple perspectives (e.g., Direction, volume, velocity, heterogeneity, topic, popularity, etc.). Then we learn an estate ranking predictor by combining a pair wise ranking objective and a sparsity regularization in a unified probabilistic framework. And we develop an effective solution for the optimization problem. Finally, we conduct a comprehensive performance evaluation with real world estate related data, and the experimental results demonstrate the competitive performance of both features and the proposed model.

Technology prospecting is a process to evaluate the potential business values of high tech companies from the technology perspective. In this paper, we provide a new view-angle to understand technology prospecting by studying the evolving distributions of technologies in the companies. Specifically, we first exploit topic models to learn technological context in the form of probabilistic distributions of assignees and locations from large-scale patent documents. Then, we develop a matching solution to measure the relationships between patent topics and the description documents of technology terms. In this way, we can obtain the distribution of technologies for each company. In addition, we are able to assess the technology prospecting of a company by a designed indicator, which allows to compare the levels of discrepancies between the emerging technology distributions available as Garner Hype Cycles and the distribution of technologies of the company. Finally, experimental results on real-world patent data show the effectiveness of our approach for technology prospecting.

The management of large-scale distributed information systems relies on the effective use and modeling of monitoring data collected at various points in the distributed information systems. A promising approach is to discover invariant relationships among the monitoring data and generate invariant networks, where a node is a monitoring data source (metric) and a link indicates an invariant relationship between two monitoring data. Such an invariant network representation can help system experts to localize and diagnose the system faults by examining those broken invariant relationships and their related metrics, because system faults usually propagate among the monitoring data and eventually lead to some broken invariant relationships. However, at one time, there are usually a lot of broken links (invariant relationships) within an invariant network. Without proper guidance, it is difficult for system experts to manually inspect this large number of broken links. Thus, a critical challenge is how to effectively and efficiently rank metrics (nodes) of invariant networks according to the anomaly levels of metrics. The ranked list of metrics will provide system experts with useful guidance for them to localize and diagnose the system faults. To this end, we propose to model the nodes and the broken links as a Markov Random Field (MRF), and develop an iteration algorithm to infer the anomaly of each node based on belief propagation (BP). Finally, we validate the proposed algorithm on both real-world and synthetic data sets to illustrate its effectiveness.

The problem of software artifact retrieval has the goal to effectively locate software artifacts, such as a piece of source code, in a large code repository. This problem has been traditionally addressed through the textual query. In other words, information retrieval techniques will be exploited based on the textual similarity between queries and textual representation of software artifacts, which is generated by collecting words from comments, identifiers, and descriptions of programs. However, in addition to these semantic information, there are rich information embedded in source codes themselves. These source codes, if analyzed properly, can be a rich source for enhancing the efforts of software artifact retrieval. To this end, in this paper, we develop a feature extraction method on source codes. Specifically, this method can capture both the inherent information in the source codes and the semantic information hidden in the comments, descriptions, and identifiers of the source codes. Moreover, we design a heterogeneous metric learning approach, which allows to integrate code features and text features into the same latent semantic space. This, in turn, can help to measure the artifact similarity by exploiting the joint power of both code and text features. Finally, extensive experiments on real-world data show that the proposed method can help to improve the performances of software artifact retrieval with a significant margin.

P2P lending is an online platform to make borrowing and investment transactions. A central question on these platforms is how to align the right products with the right investors, thus helping investors to make better decisions. Along this line, tremendous efforts have been devoted to modeling the credits of products and borrowers from an economic perspective. However, these global models are only exploratory in nature and are not practical. In this paper, we focus on the personalized investment recommendation by reconstructing the two steps for investment decision making: what to buy and how much money to pay. Specifically, we first generate a candidate investment recommendation list for each investor that tackles "what to buy" problem. In this process, we consider various unique properties of investment recommendation. Furthermore, according to the portfolio theory, we optimize the shares of each recommended candidate by incorporating the investments an investor currently holds, thus solving the "how much money to pay" problem. Finally, extensive experimental results on a large-scale real world dataset show the effectiveness of our model under various evaluation metrics.

Many of today's online news websites have enabled users to specify different types of emotions (e.g., Angry and shocked) they have after reading news. Compared with traditional user feedbacks such as comments and ratings, these specific emotion annotations are more accurate for expressing users' personal emotions. In this paper, we propose to exploit these users' emotion annotations for online news in order to track the evolution of emotions, which plays an important role in various online services. A critical challenge is how to model emotions with respect to time spans. To this end, we propose a time-aware topic modeling perspective for solving this problem. Specifically, we first develop a model named emotion-Topic over Time (eToT), in which we represent the topics of news as a Beta distribution over time and a multinomial distribution over emotions. Whilee ToT can uncover the latent relationship among news, emotion and time directly, it cannot capture the dynamics of topics. Therefore, we further develop another model named emotion based Dynamic Topic Model (eDTM), where we explore the state space model for tracking the dynamics of topics. In addition, we demonstrate that both eToT and eDTM could enable several potential applications, such as emotion prediction, emotion-based news recommendations and emotion anomaly detections. Finally, we validate the proposed models with extensive experiments with a real-world data set.

It is traditionally a challenge for home buyers to understand, compare and contrast the investment values of real estates. While a number of estate appraisal methods have been developed to value real property, the performances of these methods have been limited by the traditional data sources for estate appraisal. However, with the development of new ways of collecting estate-related mobile data, there is a potential to leverage geographic dependencies of estates for enhancing estate appraisal. Indeed, the geographic dependencies of the value of an estate can be from the characteristics of its own neighborhood (individual), the values of its nearby estates (peer), and the prosperity of the affiliated latent business area (zone). To this end, in this paper, we propose a geographic method, named ClusRanking, for estate appraisal by leveraging the mutual enforcement of ranking and clustering power. ClusRanking is able to exploit geographic individual, peer, and zone dependencies in a probabilistic ranking model. Specifically, we first extract the geographic utility of estates from geography data, estimate the neighborhood popularity of estates by mining taxicab trajectory data, and model the influence of latent business areas via ClusRanking. Also, we use a linear model to fuse these three influential factors and predict estate investment values. Moreover, we simultaneously consider individual, peer and zone dependencies, and derive an estate-specific ranking likelihood as the objective function. Finally, we conduct a comprehensive evaluation with real-world estate related data, and the experimental results demonstrate the effectiveness of our method.

Advances in real-time location system (RTLS) solutions have enabled us to collect massive amounts of fine-grained semantically rich location traces, which provide unparalleled opportunities for understanding human activities and discovering useful knowledge. This, in turn, delivers intelligence for real-time decision making in various fields, such as workflow management. Indeed, it is a new paradigm for workflow modeling by the knowledge discovery in location traces. To that end, in this paper, we provide a focused study of workflow modeling by the integrated analysis of indoor location traces in the hospital environment. In comparison with conventional workflow modeling based on passive workflow logs, one salient feature of our approach is that it can proactively unravel the workflow patterns hidden in the location traces, by automatically constructing the workflow states and estimating parameters describing the transition patterns of moving objects. Specifically, to determine a meaningful granularity for the model, the workflow states are first constructed as regions associated with specific healthcare activities. Then, we transform the original indoor location traces to the sequences of workflow states and model the workflow transition patterns by finite state machines. Furthermore, we leverage the correlations in the location traces between related types of medical devices to reinforce the modeling performance and enable more applications. The results show that the proposed framework can not only model the workflow patterns effectively, but also have managerial applications in workflow monitoring, auditing, and inspection of workflow compliance, which are critical in the healthcare industry.

With the rapid prevalence of smart mobile devices, the number of mobile Apps available has exploded over the past few years. To facilitate the choice of mobile Apps, existing mobile App recommender systems typically recommend popular mobile Apps to mobile users. However, mobile Apps are highly varied and often poorly understood, particularly for their activities and functions related to privacy and security. Therefore, more and more mobile users are reluctant to adopt mobile Apps due to the risk of privacy invasion and other security concerns. To fill this crucial void, in this paper, we propose to develop a mobile App recommender system with privacy and security awareness. The design goal is to equip the recommender system with the functionality which allows to automatically detect and evaluate the security risk of mobile Apps. Then, the recommender system can provide App recommendations by considering both the Apps' popularity and the users' security preferences. Specifically, a mobile App can lead to security risk because insecure data access permissions have been implemented in this App. Therefore, we first develop the techniques to automatically detect the potential security risk for each mobile App by exploiting the requested permissions. Then, we propose a flexible approach based on modern portfolio theory for recommending Apps by striking a balance between the Apps' popularity and the users' security concerns, and build an App hash tree to efficiently recommend Apps. Finally, we evaluate our approach with extensive experiments on a large-scale data set collected from Google Play. The experimental results clearly validate the effectiveness of our approach.

If properly analyzed, the multi-aspect rating data could be a source of rich intelligence for providing personalized restaurant recommendations. Indeed, while recommender systems have been studied for various applications and many recommendation techniques have been developed for general or specific recommendation tasks, there are few studies for restaurant recommendation by addressing the unique challenges of the multi-aspect restaurant reviews. As we know, traditional collaborative filtering methods are typically developed for single aspect ratings. However, multi-aspect ratings are often collected from the restaurant customers. These ratings can reflect multiple aspects of the service quality of the restaurant. Also, geographic factors play an important role in restaurant recommendation. To this end, in this paper, we develop a generative probabilistic model to exploit the multi-aspect ratings of restaurants for restaurant recommendation. Also, the geographic proximity is integrated into the probabilistic model to capture the geographic influence. Moreover, the profile information, which contains customer/restaurant-independent features and the shared features, is also integrated into the model. Finally, we conduct a comprehensive experimental study on a real-world data set. The experimental results clearly demonstrate the benefit of exploiting multi-aspect ratings and the improvement of the developed generative probabilistic model.

Trajectory data is becoming more and more popular nowadays and extensive studies have been conducted on trajectory data. One important research direction about trajectory data is the anomaly detection which is to find all anomalies based on trajectory patterns in a road network. In this paper, we introduce a road segment-based anomaly detection problem, which is to detect the abnormal road segments each of which has its “real” traffic deviating from its “expected” traffic and to infer the major causes of anomalies on the road network. First, a deviation-based method is proposed to quantify the anomaly of reach road segment. Second, based on the observation that one anomaly from a road segment can trigger other anomalies from the road segments nearby, a diffusion-based method based on a heat diffusion model is proposed to infer the major causes of anomalies on the whole road network. To validate our methods, we conduct intensive experiments on a large real-world GPS dataset of about 23,000 taxis in Shenzhen, China to demonstrate the performance of our algorithms.

Recent years have witnessed an increased interest in recommender systems. Despite significant progress in this field, there still remain numerous avenues to explore. Indeed, this paper provides a study of exploiting online travel information for personalized travel package recommendation. A critical challenge along this line is to address the unique characteristics of travel data, which distinguish travel packages from traditional items for recommendation. To that end, in this paper, we first analyze the characteristics of the existing travel packages and develop a tourist-area-season topic (TAST) model. This TAST model can represent travel packages and tourists by different topic distributions, where the topic extraction is conditioned on both the tourists and the intrinsic features (i.e., locations, travel seasons) of the landscapes. Then, based on this topic model representation, we propose a cocktail approach to generate the lists for personalized travel package recommendation. Furthermore, we extend the TAST model to the tourist-relation-area-season topic (TRAST) model for capturing the latent relationships among the tourists in each travel group. Finally, we evaluate the TAST model, the TRAST model, and the cocktail recommendation approach on the real-world travel package data. Experimental results show that the TAST model can effectively capture the unique characteristics of the travel data and the cocktail approach is, thus, much more effective than traditional recommendation techniques for travel package recommendation. Also, by considering tourist relationships, the TRAST model can be used as an effective assessment for travel group formation.

Advances in tourism economics have enabled us to collect massive amounts of travel tour data. If properly analyzed, this data could be a source of rich intelligence for providing real-time decision making and for the provision of travel tour recommendations. However, tour recommendation is quite different from traditional recommendations, because the tourist’s choice is affected directly by the travel costs, which includes both financial and time costs. To that end, in this article, we provide a focused study of cost-aware tour recommendation. Along this line, we first propose two ways to represent user cost preference. One way is to represent user cost preference by a two-dimensional vector. Another way is to consider the uncertainty about the cost that a user can afford and introduce a Gaussian prior to model user cost preference. With these two ways of representing user cost preference, we develop different cost-aware latent factor models by incorporating the cost information into the probabilistic matrix factorization (PMF) model, the logistic probabilistic matrix factorization (LPMF) model, and the maximum margin matrix factorization (MMMF) model, respectively. When applied to real-world travel tour data, all the cost-aware recommendation models consistently outperform existing latent factor models with a significant margin.

Ranking fraud in the mobile App market refers to fraudulent or deceptive activities which have a purpose of bumping up the Apps in the popularity list. Indeed, it becomes more and more frequent for App develops to use shady means, such as inflating their Apps' sales or posting phony App ratings, to commit ranking fraud. While the importance of preventing ranking fraud has been widely recognized, there is limited understanding and research in this area. To this end, in this paper, we provide a holistic view of ranking fraud and propose a ranking fraud detection system for mobile Apps. Specifically, we investigate two types of evidences, ranking based evidences and rating based evidences, by modeling Apps' ranking and rating behaviors through statistical hypotheses tests. In addition, we propose an optimization based aggregation method to integrate all the evidences for fraud detection. Finally, we evaluate the proposed system with real-world App data collected from the Apple's App Store for a long time period. In the experiments, we validate the effectiveness of the proposed system, and show the scalability of the detection algorithm as well as some regularity of ranking fraud activities.

The management of large-scale distributed information systems relies on the effective use and modeling of monitoring data collected at various points in the distributed information systems. A traditional approach to model monitoring data is to discover invariant relationships among the monitoring data. Indeed, we can discover all invariant relationships among all pairs of monitoring data and generate invariant networks, where a node is a monitoring data source (metric) and a link indicates an invariant relationship between two monitoring data. Such an invariant network representation can help system experts to localize and diagnose the system faults by examining those broken invariant relationships and their related metrics, since system faults usually propagate among the monitoring data and eventually lead to some broken invariant relationships. However, at one time, there are usually a lot of broken links (invariant relationships) within an invariant network. Without proper guidance, it is difficult for system experts to manually inspect this large number of broken links. To this end, in this article, we propose the problem of ranking metrics according to the anomaly levels for a given invariant network, while this is a nontrivial task due to the uncertainties and the complex nature of invariant networks. Specifically, we propose two types of algorithms for ranking metric anomaly by link analysis in invariant networks. Along this line, we first define two measurements to quantify the anomaly level of each metric, and introduce the m R ank algorithm. Also, we provide a weighted score mechanism and develop the g R ank algorithm, which involves an iterative process to obtain a score to measure the anomaly levels. In addition, some extended algorithms based on m R ank and g R ank algorithms are developed by taking into account the probability of being broken as well as noisy links. Finally, we validate all the proposed algorithms on a large number of real-world and synthetic data sets to illustrate the effectiveness and efficiency of different algorithms.

Rapid growth in the development of real-time location system solutions has led to an increased interest in indoor location-aware services, such as hospital asset management. Although there are extensive studies in the literature on the analysis of outdoor location traces, the studies of indoor location traces are less touched and fragmented. To that end, in this paper, we provide a focused study of indoor location traces collected by the sensors attached to medical devices in a hospital environment. Along this line, we first introduce some unique properties of these indoor location traces. We show that they can capture the movement patterns of the medical devices, which are tightly coupled with the work flow in the controlled hospital environment. Based on this observation, we propose a stochastic model for context-aware anomaly detection in indoor location traces, which exploits the hospital work flow and models the movements of medical devices as transitions in finite state machines. In detail, we first develop a density-based method to identify the hotspots filled with high-level abnormal activities in the indoor environment. The discovered hotspots serve as the context for nearby trajectories. Then, we introduce an N-gram based method for measuring the degree of anomaly based on the detected hotspots, which is able to predict the missing events possibly due to the devices being stolen. Besides, to address the noisy nature of the indoor sensor networks, we also propose an iterative algorithm to estimate the transition probabilities. This algorithm allows to effectively recover the missing location records which are critical for the abnormality estimation. Finally, the experimental results on the real-world date sets validate the effectiveness of the proposed context-aware anomaly detection method for identifying abnormal events.

In this paper, we present a systematic perspective study on choosing and evaluating the initial seed items that will be recommended to the cold start users. We first construct an item consumption correlation network to capture the existing users' general consumption behaviors. Then, we formalize initial items recommendation as the influential seed set selection problem. Along this line, we present several methods, each of which selects seed items according to different rules. Finally, the experimental results on two real-world data sets verify that with different seed items, the users' consumption numbers will be quite different. Meanwhile, the results also provide many deep insights into these selection methods and their recommended seed items.

It is well known that the distance metric plays an important role in the clustering process. Indeed, many clustering problems can be treated as an optimization problem of a criterion function defined over one distance metric. While many distance metrics have been developed, it is not clear that how these distance metrics can impact on the clustering/optimization process. To that end, in this paper, we study the impact of a set of popular cosine-based distance metrics on K-means clustering. Specifically, by revealing the common order-preserving property, we first show that K-means has exactly the same cluster assignment for these metrics during the E-step. Next, by both theoretical and empirical studies, we prove that the cluster centroid is a good approximator of their respective optimal centers in the M-step. As such, we identify a problem with K-means: it cannot differentiate these metrics. To explore the nature of these metrics, we propose an evolutionary K-means framework that integrates K-means and genetic algorithms. This framework not only enables inspection of arbitrary distance metrics, but also can be used to investigate different formulations of the optimization problem. Finally, this framework is used in extensive experiments on real-world data sets. The results validate our theoretical findings on the characteristics and interrelationships of these metrics. Most importantly, this paper furthers our understanding of the impact of the distance metrics on the optimization process of K-means.

Advances in GPS tracking technology have enabled us to install GPS tracking devices in city taxis to collect a large amount of GPS traces under operational time constraints. These GPS traces provide unparallel opportunities for us to uncover taxi driving fraud activities. In this paper, we develop a taxi driving fraud detection system, which is able to systematically investigate taxi driving fraud. In this system, we first provide functions to find two aspects of evidences: travel route evidence and driving distance evidence. Furthermore, a third function is designed to combine the two aspects of evidences based on Dempster-Shafer theory. To implement the system, we first identify interesting sites from a large amount of taxi GPS logs. Then, we propose a parameter-free method to mine the travel route evidences. Also, we introduce route mark to represent a typical driving path from an interesting site to another one. Based on route mark, we exploit a generative statistical model to characterize the distribution of driving distance and identify the driving distance evidences. Finally, we evaluate the taxi driving fraud detection system with large scale real-world taxi GPS logs. In the experiments, we uncover some regularity of driving fraud activities and investigate the motivation of drivers to commit a driving fraud by analyzing the produced taxi fraud data.

As the worlds of commerce, entertainment, travel, and Internet technology become more inextricably linked, new types of business data become available for creative use and formal analysis. Indeed, this paper provides a study of exploiting online travel information for personalized travel package recommendation. A critical challenge along this line is to address the unique characteristics of travel data, which distinguish travel packages from traditional items for recommendation. To this end, we first analyze the characteristics of the travel packages and develop a Tourist-Area-Season Topic (TAST) model, which can extract the topics conditioned on both the tourists and the intrinsic features (i.e. locations, travel seasons) of the landscapes. Based on this TAST model, we propose a cocktail approach on personalized travel package recommendation. Finally, we evaluate the TAST model and the cocktail approach on real-world travel package data. The experimental results show that the TAST model can effectively capture the unique characteristics of the travel data and the cocktail approach is thus much more effective than traditional recommendation methods for travel package recommendation.

The increasing availability of large-scale location traces creates unprecedent opportunities to change the paradigm for knowledge discovery in transportation systems. A particularly promising area is to extract useful business intelligence, which can be used as guidance for reducing inefficiencies in energy consumption of transportation sectors, improving customer experiences, and increasing business performances. However, extracting business intelligence from location traces is not a trivial task. Conventional data analytic tools are usually not customized for handling large, complex, dynamic, and distributed nature of location traces. To that end, we develop a taxi business intelligence system to explore the massive taxi location traces from different business perspectives with various data mining functions. Since we implement the system using the real-world taxi GPS data, this demonstration will help taxi companies to improve their business performances by understanding the behaviors of both drivers and customers. In addition, several identified technical challenges also motivate data mining people to develop more sophisticate techniques in the future.

Advances in tourism economics have enabled us to collect massive amounts of travel tour data. If properly analyzed, this data can be a source of rich intelligence for providing real-time decision making and for the provision of travel tour recommendations. However, tour recommendation is quite different from traditional recommendations, because the tourist's choice is directly affected by the travel cost, which includes the financial cost and the time. To that end, in this paper, we provide a focused study of cost-aware tour recommendation. Along this line, we develop two cost-aware latent factor models to recommend travel packages by considering both the travel cost and the tourist's interests. Specifically, we first design a cPMF model, which models the tourist's cost with a 2-dimensional vector. Also, in this cPMF model, the tourist's interests and the travel cost are learnt by exploring travel tour data. Furthermore, in order to model the uncertainty in the travel cost, we further introduce a Gaussian prior into the cPMF model and develop the GcPMF model, where the Gaussian prior is used to express the uncertainty of the travel cost. Finally, experiments on real-world travel tour data show that the cost-aware recommendation models outperform state-of-the-art latent factor models with a significant margin. Also, the GcPMF model with the Gaussian prior can better capture the impact of the uncertainty of the travel cost, and thus performs better than the cPMF model.

Rating sparsity is a critical issue for collaborative filtering. For example, the well-known Netflix Movie rating data contain ratings of only about 1% user-item pairs. One way to address this rating sparsity problem is to develop more effective methods for training rating prediction models. To this end, in this paper, we introduce a collective training paradigm to automatically and effectively augment the training ratings. Essentially, the collective training paradigm builds multiple different Collaborative Filtering (CF) models separately, and augments the training ratings of each CF model by using the partial predictions of other CF models for unknown ratings. Along this line, we develop two algorithms, Bi-CF and Tri-CF, based on collective training. For Bi-CF and Tri-CF, we collectively and iteratively train two and three different CF models via iteratively augmenting training ratings for individual CF model. We also design different criteria to guide the selection of augmented training ratings for Bi-CF and Tri-CF. Finally, the experimental results show that Bi-CF and Tri-CF algorithms can significantly outperform baseline methods, such as neighborhood-based and SVD-based models.

In this study, we formalize a multifocal learning problem, where training data are partitioned into several different focal groups and the prediction model will be learned within each focal group. The multifocal learning problem is motivated by numerous real-world learning applications. For instance, for the same type of problems encountered in a customer service center, the problem descriptions from different customers can be quite different. Experienced customers usually give more precise and focused descriptions about the problem. In contrast, inexperienced customers usually provide diverse descriptions. In this case, the examples from the same class in the training data can be naturally in different focal groups. Therefore, it is necessary to identify those natural focal groups and exploit them for learning at different focuses. Along this line, the key development challenge is how to identify those focal groups in the training data. As a case study, we exploit multifocal learning for profiling customer problems. Also, we provide an empirical study about how the performance of multifocal learning is affected by the quality of focal groups. The results on real-world customer problem logs show that multifocal learning can significantly boost the performance of many existing classification algorithms, such as Support Vector Machines (SVMs), for classifying customer problems and there is strong correlation between the quality of focal groups and the learning performance.

The increasing availability of large-scale location traces creates unprecedent opportunities to change the paradigm for identifying abnormal moving activities. Indeed, various aspects of abnormality of moving patterns have recently been exploited, such as wrong direction and wandering. However, there is no recognized way of combining different aspects into an unified evolving abnormality score which has the ability to capture the evolving nature of abnormal moving trajectories. To that end, in this paper, we provide an evolving trajectory outlier detection method, named TOP-EYE, which continuously computes the outlying score for each trajectory in an accumulating way. Specifically, in TOP-EYE, we introduce a decay function to mitigate the influence of the past trajectories on the evolving outlying score, which is defined based on the evolving moving direction and density of trajectories. This decay function enables the evolving computation of accumulated outlying scores along the trajectories. An advantage of TOP-EYE is to identify evolving outliers at very early stage with relatively low false alarm rate. Finally, experimental results on real-world location traces show that TOP-EYE can effectively capture evolving abnormal trajectories.

The increasing availability of large-scale location traces creates unprecedent opportunities to change the paradigm for knowledge discovery in transportation systems. A particularly promising area is to extract energy-efficient transportation patterns (green knowledge), which can be used as guidance for reducing inefficiencies in energy consumption of transportation sectors. However, extracting green knowledge from location traces is not a trivial task. Conventional data analysis tools are usually not customized for handling the massive quantity, complex, dynamic, and distributed nature of location traces. To that end, in this paper, we provide a focused study of extracting energy-efficient transportation patterns from location traces. Specifically, we have the initial focus on a sequence of mobile recommendations. As a case study, we develop a mobile recommender system which has the ability in recommending a sequence of pick-up points for taxi drivers or a sequence of potential parking positions. The goal of this mobile recommendation system is to maximize the probability of business success. Along this line, we provide a Potential Travel Distance (PTD) function for evaluating each candidate sequence. This PTD function possesses a monotone property which can be used to effectively prune the search space. Based on this PTD function, we develop two algorithms, LCP and SkyRoute, for finding the recommended routes. Finally, experimental results show that the proposed system can provide effective mobile sequential recommendation and the knowledge extracted from location traces can be used for coaching drivers and leading to the efficient use of energy.

In this study, we formalize a multi-focal learning problem, where training data are partitioned into several different focal groups and the prediction model will be learned within each focal group. The multi-focal learning problem is motivated by numerous real-world learning applications. For instance, for the same type of problems encountered in a customer service center, the problem descriptions from different customers can be quite different. The experienced customers usually give more precise and focused descriptions about the problem. In contrast, the inexperienced customers usually provide more diverse descriptions. In this case, the examples from the same class in the training data can be naturally in different focal groups. As a result, it is necessary to identify those natural focal groups and exploit them for learning at different focuses. The key developmental challenge is how to identify those focal groups in the training data. As a case study, we exploit multi-focal learning for profiling problems in customer service centers. The results show that multifocal learning can significantly boost the learning accuracies of existing learning algorithms, such as Support Vector Machines (SVMs), for classifying customer problems.