Personalization is well studied in search and recommendation, but personalized question answering remains underexplored due to challenges in inferring preferences from long, noisy, implicit contexts and generating responses that are both accurate and aligned with user expectations. To address this, we propose Pathways of Thoughts (PoT), an inference-stage method that applies to any large language model (LLM) without task-specific fine-tuning. PoT models the thinking as an iterative decision process, where the model dynamically selects among cognitive operations such as reasoning, revision, personalization, and clarification. This enables exploration of multiple reasoning trajectories, producing diverse candidate responses that capture different perspectives. PoT then aggregates and reweights these candidates according to inferred user preferences, yielding a final personalized response that benefits from the complementary strengths of diverse reasoning paths. Experiments on the LaMP-QA benchmark show that PoT consistently outperforms competitive baselines, achieving up to a 10.8% relative improvement. Human evaluation further validates these improvements, with annotators preferring PoT in 66% of cases compared to the best-performing baseline and reporting ties in 15% of cases.

Low-dimensional visualizations, or "projection maps" are widely used in scientific research and creative industries to interpret largescale and complex datasets.These visualizations not only support the understanding of existing knowledge spaces but are often used implicitly to guide exploration into unknown areas.While such visualizations can be created through various methods such as TSNE or UMAP, there is no systematic way to leverage them for

Recent advances in general-purpose AI underscore the urgent need to align AI systems with human goals and values. Yet, the lack of a clear, shared understanding of what constitutes "alignment" limits meaningful progress and cross-disciplinary collaboration. In this position paper, we argue that the research community should explicitly define and critically reflect on "alignment" to account for the bidirectional and dynamic relationship between humans and AI. Through a systematic review of over 400 papers spanning HCI, NLP, ML, and more, we examine how alignment is currently defined and operationalized. Building on this analysis, we introduce the Bidirectional Human-AI Alignment framework, which not only incorporates traditional efforts to align AI with human values but also introduces the critical, underexplored dimension of aligning humans with AI – supporting cognitive, behavioral, and societal adaptation to rapidly advancing AI technologies. Our findings reveal significant gaps in current literature, especially in long-term interaction design, human value modeling, and mutual understanding. We conclude with three central challenges and actionable recommendations to guide future research toward more nuanced, reciprocal, and human-AI alignment approaches.

Large Language Models (LLMs) have demonstrated superior results across a wide range of tasks, and Retrieval-augmented Generation (RAG) is an effective way to enhance the performance by locating relevant information and placing it into the context window of the LLM.However, the relationship between retrievers and LLM in a RAG is still under-investigated.Most existing work treats the retriever and the LLM as independent components and leaves a gap between retrieving human-"friendly" information and assembling a LLM-"friendly" context.In this work, we examine a novel bridge mechanism.We validate the ranking and selection assumptions of retrievers in the context of RAG and propose a framework that chains together supervised and reinforcement learning to train a bridge model that optimizes the connection between the retriever and the LLM.Empirical results demonstrate the effectiveness of our method in both questionanswering and personalized generation tasks.

Although remote working is increasingly adopted during the pandemic, many are concerned by the low-efficiency in the remote working. Missing in text-based communication are non-verbal cues such as facial expressions and body language, which hinders the effective communication and negatively impacts the work outcomes. Prevalent on social media platforms, emojis, as alternative non-verbal cues, are gaining popularity in the virtual workspaces well. In this paper, we study how emoji usage influences developer participation and issue resolution in virtual workspaces. To this end, we collect GitHub issues for a one-year period and apply causal inference techniques to measure the causal effect of emojis on the outcome of issues, controlling for confounders such as issue content, repository, and author information. We find that emojis can significantly reduce the resolution time of issues and attract more user participation. We also compare the heterogeneous effect on different types of issues. These findings deepen our understanding of the developer communities, and they provide design implications on how to facilitate interactions and broaden developer participation.

This workshop initiates a new study field which may be named AI behavioral science. It discusses recent findings, methodologies, applications, and potential societal impacts that are related to analyzing, understanding, and directing the behaviors of AI models, especially those built upon large language models. This half-day workshop includes several keynote and invited talks, a poster session, and a panel discussion.

This is the second workshop in the series which discusses the cutting-edge developments in research and applications of personalizing large language models (LLMs) and adapting them to the demands of diverse user populations and societal needs. The full-day workshop plan includes several keynotes and invited talks, a poster session and a panel discussion.

This workshop discusses the cutting-edge developments in research and applications of personalizing large language models (LLMs) and adapting them to the demands of diverse user populations and societal needs. The full-day workshop includes several keynotes and invited talks, a poster session and a panel discussion.

Facilitated by large language models (LLMs), personalized text generation has become a rapidly growing research direction. Most existing studies focus on designing specialized models for a particular domain, or they require fine-tuning the LLMs to generate personalized text. We consider a typical scenario in which the large language model, which generates personalized output, is frozen and can only be accessed through APIs. Under this constraint, all one can do is to improve the input text (i.e., text prompts) sent to the LLM, a procedure that is usually done manually. In this paper, we propose a novel method to automatically revise prompts for personalized text generation. The proposed method takes the initial prompts generated by a state-of-the-art, multistage framework for personalized generation and rewrites a few critical components that summarize and synthesize the personal context. The prompt rewriter employs a training paradigm that chains together supervised learning (SL) and reinforcement learning (RL), where SL reduces the search space of RL and RL facilitates end-to-end training of the rewriter. Using datasets from three representative domains, we demonstrate that the rewritten prompts outperform both the original prompts and the prompts optimized via supervised learning or reinforcement learning alone. In-depth analysis of the rewritten prompts shows that they are not only human readable, but also able to guide manual revision of prompts when there is limited resource to employ reinforcement learning to train the prompt rewriter, or when it is costly to deploy an automatic prompt rewriter for inference.

In this paper, we first define information loss that occurs in the hypergraph expansion and then propose a novel framework, named MILEAGE, to evaluate hypergraph expansion methods by measuring their degree of information loss. MILEAGE employs the following four steps: (1) expanding a hypergraph; (2) performing the unsupervised representation learning on the expanded graph; (3) reconstructing a hypergraph based on vector representations obtained; and (4) measuring MILEAGE-score (i.e., mileage) by comparing the reconstructed and the original hypergraphs. To demonstrate the usefulness of MILEAGE, we conduct experiments via downstream tasks on three levels (i.e., node, hyperedge, and hypergraph): node classification, hyperedge prediction, and hypergraph classification on eight real-world hypergraph datasets. Through the extensive experiments, we observe that information loss through hypergraph expansion has a negative impact on downstream tasks and MILEAGE can effectively evaluate hypergraph expansion methods through the information loss and recommend a new method that resolves the problems of existing ones.

Modern technology innovations feature a successive and even recurrent procedure. Intervals between old and new generations of technology are shrinking, and the Internet and Web services have facilitated the fast adoption of an innovation even before the convergence of its predecessor. While the adoption and diffusion of innovations have been studied for decades, most theories and analyses focus on single and one-time innovations. Meanwhile, limited work has investigated successive innovations while lacking user-level analysis, possibly due to the unavailability of fine-grained adoption behavior data. In this study, we present the first large-scale analysis of the adoption of recurrent innovations in the context of mobile app updates, investigating how millions of users consume various versions of thousands of apps on their mobile devices. Our analysis reveals novel patterns of crowd and individual adoption behaviors, which suggest the need for new categories of adopters to be added on top of the Rogers model of innovation diffusion. We show that standard machine learning models are able to pick up various sources of signals to predict whether users in these different categories will adopt a new version of an app and how soon they will adopt it.

While many organizations have shifted to working remotely during the COVID-19 pandemic, how the remote workforce and the remote teams are influenced by and would respond to this and future shocks remain largely unknown. Software developers have relied on remote collaborations long before the pandemic, working in virtual teams (GitHub repositories). The dynamics of these repositories through the pandemic provide a unique opportunity to understand how remote teams react under shock. This work presents a systematic analysis. We measure the overall effect of the early pandemic on public GitHub repositories by comparing their sizes and productivity with the counterfactual outcomes forecasted as if there were no pandemic. We find that the productivity level and the number of active members of these teams vary significantly during different periods of the pandemic. We then conduct a finer-grained investigation and study the heterogeneous effects of the shock on individual teams. We find that the resilience of a team is highly correlated to certain properties of the team before the pandemic. Through a bootstrapped regression analysis, we reveal which types of teams are robust or fragile to the shock.

Recent years have witnessed a surge of research interest in graph machine learning. However, the benchmark datasets available to the field are rather limited in both quantity and diversity, an issue particularly notable given the immense potential applications of graph learning. The lack of diverse benchmark datasets may have biased the development of graph machine learning techniques towards narrow directions. By crowdsourcing novel tasks and datasets, this workshop aims to increase the diversity of graph learning benchmarks, identify new demands of graph machine learning in general, and gain a better synergy of how concrete techniques perform on these benchmarks. Moreover, this workshop offers a platform for discussions of best practices in curating graph learning benchmarks and data-centric approaches for graph learning.

Graph Neural Networks (GNNs) have achieved remarkable success in various applications, but their performance can be sensitive to specific data properties of the graph datasets they operate on. Current literature on understanding the limitations of GNNs has primarily employed a \emph{model-driven} approach that leverage heuristics and domain knowledge from network science or graph theory to model the GNN behaviors, which is time-consuming and highly subjective. In this work, we propose a \emph{metadata-driven} approach to analyze the sensitivity of GNNs to graph data properties, motivated by the increasing availability of graph learning benchmarks. We perform a multivariate sparse regression analysis on the metadata derived from benchmarking GNN performance across diverse datasets, yielding a set of salient data properties. To validate the effectiveness of our data-driven approach, we focus on one identified data property, the degree distribution, and investigate how this property influences GNN performance through theoretical analysis and controlled experiments. Our theoretical findings reveal that datasets with more balanced degree distribution exhibit better linear separability of node representations, thus leading to better GNN performance. We also conduct controlled experiments using synthetic datasets with varying degree distributions, and the results align well with our theoretical findings. Collectively, both the theoretical analysis and controlled experiments verify that the proposed metadata-driven approach is effective in identifying critical data properties for GNNs.

Recent developments in large language models (LLM) and generative AI have unleashed the astonishing capabilities of text-to-image generation systems to synthesize high-quality images that are faithful to a given reference text, known as a “prompt”. These systems have immediately received lots of attention from researchers, creators, and common users. Despite the plenty of efforts to improve the generative models, there is limited work on understanding the information needs of the users of these systems at scale. We conduct the first comprehensive analysis of large-scale prompt logs collected from multiple text-to-image generation systems. Our work is analogous to analyzing the query logs of Web search engines, a line of work that has made critical contributions to the glory of the Web search industry and research. Compared with Web search queries, text-to-image prompts are significantly longer, often organized into special structures that consist of the subject, form, and intent of the generation tasks and present unique categories of information needs. Users make more edits within creation sessions, which present remarkable exploratory patterns. There is also a considerable gap between the user-input prompts and the captions of the images included in the open training data of the generative models. Our findings provide concrete implications on how to improve text-to-image generation systems for creation purposes.

Graph neural networks (GNNs) have attracted increasing interests. With broad deployments of GNNs in real-world applications, there is an urgent need for understanding the robustness of GNNs under adversarial attacks, especially in realistic setups. In this work, we study the problem of attacking GNNs in a restricted and realistic setup, by perturbing the features of a small set of nodes, with no access to model parameters and model predictions. Our formal analysis draws a connection between this type of attacks and an influence maximization problem on the graph. This connection not only enhances our understanding on the problem of adversarial attack on GNNs, but also allows us to propose a group of effective and practical attack strategies. Our experiments verify that the proposed attack strategies significantly degrade the performance of three popular GNN models and outperform baseline adversarial attack strategies.

Multinomial Logit (MNL) is one of the most popular discrete choice models and has been widely used to model ranking data. However, there is a long-standing technical challenge of learning MNL from many real-world ranking data: exact calculation of the MNL likelihood of partial rankings is generally intractable. In this work, we develop a scalable method for approximating the MNL likelihood of general partial rankings in polynomial time complexity. We also extend the proposed method to learn mixture of MNL. We demonstrate that the proposed methods are particularly helpful for applications to choice-based network formation modeling, where the formation of new edges in a network is viewed as individuals making choices of their friends over a candidate set. The problem of learning mixture of MNL models from partial rankings naturally arises in such applications. And the proposed methods can be used to learn MNL models from network data without the strong assumption that temporal orders of all the edge formation are available. We conduct experiments on both synthetic and real-world network data to demonstrate that the proposed methods achieve more accurate parameter estimation and better fitness of data compared to conventional methods.

In this paper, we propose a flexible model for survival analysis using neural networks along with scalable optimization algorithms. One key technical challenge for directly applying maximum likelihood estimation (MLE) to censored data is that evaluating the objective function and its gradients with respect to model parameters requires the calculation of integrals. To address this challenge, we recognize from a novel perspective that the MLE for censored data can be viewed as a differential-equation constrained optimization problem. Following this connection, we model the distribution of event time through an ordinary differential equation and utilize efficient ODE solvers and adjoint sensitivity analysis to numerically evaluate the likelihood and the gradients. Using this approach, we are able to 1) provide a broad family of continuous-time survival distributions without strong structural assumptions, 2) obtain powerful feature representations using neural networks, and 3) allow efficient estimation of the model in large-scale applications using stochastic gradient descent. Through both simulation studies and real-world data examples, we demonstrate the effectiveness of the proposed method in comparison to existing state-of-the-art deep learning survival analysis models. The implementation of the proposed SODEN approach has been made publicly available at https://github.com/jiaqima/SODEN.

User mobility prediction is widely considered by the research community. Many studies have explored various algorithms to predict where a user is likely to visit based on their contexts and trajectories. Most of existing studies focus on specific targets of predictions. While successful cases are often reported, few discussions have been done on what happens if the prediction targets vary: whether coarser locations are easier to be predicted, and whether predicting the immediate next location on the trajectory is easier than predicting the destination. On the other hand, while spatiotemporal tags and content information are commonly used in current prediction tasks, few have utilized the finer grained, on-device user behavioral data, which are supposed to be more informative and indicative of user intentions. In this paper, we conduct a systematic study on the mobility prediction using a large-scale real-world dataset that contains plentiful contextual information. Based on a series of learning models, including a Markov model, two recurrent neural network models, and a multi-modal learning method, we perform extensive experiments to comprehensively investigate the predictability of different types of granularities of targets and the effectiveness of different types of signals. The results provide insightful knowledge on what can be predicted along with how, which sheds light on the real-world mobility prediction from a relatively general perspective.

Cristina Garbacea, Mengtian Guo, Samuel Carton, Qiaozhu Mei. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

We consider the problem of listwise learning-to-rank (LTR) on data with \textit{partitioned preference}, where a set of items are sliced into ordered and disjoint partitions, but the ranking of items within a partition is unknown. The Plackett-Luce (PL) model has been widely used in listwise LTR methods. However, given $N$ items with $M$ partitions, calculating the likelihood of data with partitioned preference under the PL model has a time complexity of $O(N+S!)$, where $S$ is the maximum size of the top $M-1$ partitions. This computational challenge restrains existing PL-based listwise LTR methods to only a special case of partitioned preference, \textit{top-$K$ ranking}, where the exact order of the top $K$ items is known. In this paper, we exploit a random utility model formulation of the PL model and propose an efficient approach through numerical integration for calculating the likelihood. This numerical approach reduces the aforementioned time complexity to $O(N+MS)$, which allows training deep-neural-network-based ranking models with a large output space. We demonstrate that the proposed method outperforms well-known LTR baselines and remains scalable through both simulation experiments and applications to real-world eXtreme Multi-Label (XML) classification tasks. The proposed method also achieves state-of-the-art performance on XML datasets with relatively large numbers of labels per sample.

Despite enormous successful applications of graph neural networks (GNNs), theoretical understanding of their generalization ability, especially for node-level tasks where data are not independent and identically-distributed (IID), has been sparse. The theoretical investigation of the generalization performance is beneficial for understanding fundamental issues (such as fairness) of GNN models and designing better learning methods. In this paper, we present a novel PAC-Bayesian analysis for GNNs under a non-IID semi-supervised learning setup. Moreover, we analyze the generalization performances on different subgroups of unlabeled nodes, which allows us to further study an accuracy-(dis)parity-style (un)fairness of GNNs from a theoretical perspective. Under reasonable assumptions, we demonstrate that the distance between a test subgroup and the training set can be a key factor affecting the GNN performance on that subgroup, which calls special attention to the training node selection for fair learning. Experiments across multiple GNN models and datasets support our theoretical results.

Sequential recommendation techniques are considered to be a promising way of providing better user experience in mobile commerce by learning sequential interests within user historical interaction behaviors. However, the recently increasing focus on privacy concerns, such as the General Data Protection Regulation (GDPR), can significantly affect the deployment of state-of-the-art sequential recommendation techniques, because user behavior data are no longer allowed to be arbitrarily used without the user’s explicit permission. To address the issue, this paper proposes DeepRec, an on-device deep learning framework of mining interaction behaviors for sequential recommendation without sending any raw data or intermediate results out of the device, preserving user privacy maximally. DeepRec constructs a global model using data collected before GDPR and fine-tunes a personal model continuously on individual mobile devices using data collected after GDPR. DeepRec employs the model pruning and embedding sparsity techniques to reduce the computation and network overhead, making the model training process practical on computation-constraint mobile devices. Evaluation results show that DeepRec can achieve comparable recommendation accuracy to existing centralized recommendation approaches with small computation overhead and up to 10x reduction in network overhead.

The past decades witnessed the fast and wide deployment of Internet. The Internet has bred the ubiquitous computing environment that is spanning the cloud, edge, mobile devices, and IoT. Software running over such a ubiquitous computing environment environment is eating the world. A recently emerging trend of Internet-based software systems is “ resource adaptive ,” i.e., software systems should be robust and intelligent enough to the changes of heterogeneous resources, both physical and logical, provided by their running environment. To keep pace of such a trend, we argue that some considerations should be taken into account for the future operating system design and implementation. From the structural perspective, rather than the “monolithic OS” that manages the aggregated resources on the single machine, the OS should be dynamically composed over the distributed resources and flexibly adapt to the resource and environment changes. Meanwhile, the OS should leverage advanced machine/deep learning techniques to derive configurations and policies and automatically learn to tune itself and schedule resources. This article envisions our recent thinking of the new OS abstraction, namely, ServiceOS , for future resource-adaptive intelligent software systems. The idea of ServiceOS is inspired by the delivery model of “ Software-as-a-Service ” that is supported by the Service-Oriented Architecture (SOA). The key principle of ServiceOS is based on resource disaggregation, resource provisioning as a service, and learning-based resource scheduling and allocation. The major goal of this article is not providing an immediately deployable OS. Instead, we aim to summarize the challenges and potentially promising opportunities and try to provide some practical implications for researchers and practitioners.

Sentiment classification typically relies on a large amount of labeled data. In practice, the availability of labels is highly imbalanced among different languages. To tackle this problem, cross-lingual sentiment classification approaches aim to transfer knowledge learned from one language that has abundant labeled examples (i.e., the source language, usually English) to another language with fewer labels (i.e., the target language). The source and the target languages are usually bridged through off-the-shelf machine translation tools. Through such a channel, cross-language sentiment patterns can be successfully learned from English and transferred into the target languages. This approach, however, often fails to capture sentiment knowledge specific to the target language. In this paper, we employ emojis, which are widely available in many languages, as a new channel to learn both the cross-language and the language-specific sentiment patterns. We propose a novel representation learning method that uses emoji prediction as an instrument to learn respective sentiment-aware representations for each language. The learned representations are then integrated to facilitate cross-lingual sentiment classification.

Millions of drivers worldwide have enjoyed financial benefits and work schedule flexibility through a ride-sharing economy, but meanwhile they have suffered from the lack of a sense of identity and career achievement. Equipped with social identity and contest theories, financially incentivized team competitions have been an effective instrument to increase drivers' productivity, job satisfaction, and retention, and to improve revenue over cost for ride-sharing platforms. While these competitions are overall effective, the decisive factors behind the treatment effects and how they affect the outcomes of individual drivers have been largely mysterious. In this study, we analyze data collected from more than 500 large-scale team competitions organized by a leading ride-sharing platform, building machine learning models to predict individual treatment effects. Through a careful investigation of features and predictors, we are able to reduce out-sample prediction error by more than 24%. Through interpreting the best-performing models, we discover many novel and actionable insights regarding how to optimize the design and the execution of team competitions on ride-sharing platforms. A simulated analysis demonstrates that by simply changing a few contest design options, the average treatment effect of a real competition is expected to increase by as much as 26%. Our procedure and findings shed light on how to analyze and optimize large-scale online field experiments in general.

We study the black-box attacks on graph neural networks (GNNs) under a novel and realistic constraint: attackers have access to only a subset of nodes in the network, and they can only attack a small number of them. A node selection step is essential under this setup. We demonstrate that the structural inductive biases of GNN models can be an effective source for this type of attacks. Specifically, by exploiting the connection between the backward propagation of GNNs and random walks, we show that the common gradient-based white-box attacks can be generalized to the black-box setting via the connection between the gradient and an importance score similar to PageRank. In practice, we find attacks based on this importance score indeed increase the classification loss by a large margin, but they fail to significantly increase the mis-classification rate. Our theoretical and empirical analyses suggest that there is a discrepancy between the loss and mis-classification rate, as the latter presents a diminishing-return pattern when the number of attacked nodes increases. Therefore, we propose a greedy procedure to correct the importance score that takes into account of the diminishing-return pattern. Experimental results show that the proposed procedure can significantly increase the mis-classification rate of common GNNs on real-world data without access to model parameters nor predictions.

Smartphone applications (a.k.a., apps) have become indispensable in our everyday life and work. In practice, accomplishing a task on smartphones may require the user to navigate among various apps. Unlike Web pages that are inherently interconnected through hyperlinks, apps are usually isolated building blocks, and the lack of direct links between apps has compromised the efficiency of task completion and user experience. In this article, we present the first in-depth empirical study of page-level access behaviors of smartphone users based on a comprehensive dataset collected through an extensive user study. We propose a model to distinguish informational pages and transitional pages , based on which we can extract page-level inter-app navigation. Surprisingly, the transitional pages account for quite substantial time cost and manual actions when navigating from the current informational page to the desirable informational page. We reveal that developing “ tunnels ” between “isolated” apps under specific usage scenarios has a huge potential to reduce the cost of navigation. Our analysis provides some practical implications on how to improve app-navigation experience from both the operating system’s perspective and the developer’s<?brk?> perspective.

Cristina Garbacea, Samuel Carton, Shiyan Yan, Qiaozhu Mei. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

We consider a family of problems that are concerned about making predictions for the majority of unlabeled, graph-structured data samples based on a small proportion of labeled samples. Relational information among the data samples, often encoded in the graph/network structure, is shown to be helpful for these semi-supervised learning tasks. However, conventional graph-based regularization methods and recent graph neural networks do not fully leverage the interrelations between the features, the graph, and the labels. In this work, we propose a flexible generative framework for graph-based semi-supervised learning, which approaches the joint distribution of the node features, labels, and the graph structure. Borrowing insights from random graph models in network science literature, this joint distribution can be instantiated using various distribution families. For the inference of missing labels, we exploit recent advances of scalable variational inference techniques to approximate the Bayesian posterior. We conduct thorough experiments on benchmark datasets for graph-based semi-supervised learning. Results show that the proposed methods outperform state-of-the-art models under most settings.

Sentiment analysis has various application scenarios in software engineering (SE), such as detecting developers' emotions in commit messages and identifying their opinions on Q&A forums. However, commonly used out-of-the-box sentiment analysis tools cannot obtain reliable results on SE tasks and the misunderstanding of technical jargon is demonstrated to be the main reason. Then, researchers have to utilize labeled SE-related texts to customize sentiment analysis for SE tasks via a variety of algorithms. However, the scarce labeled data can cover only very limited expressions and thus cannot guarantee the analysis quality. To address such a problem, we turn to the easily available emoji usage data for help. More specifically, we employ emotional emojis as noisy labels of sentiments and propose a representation learning approach that uses both Tweets and GitHub posts containing emojis to learn sentiment-aware representations for SE-related texts. These emoji-labeled posts can not only supply the technical jargon, but also incorporate more general sentiment patterns shared across domains. They as well as labeled data are used to learn the final sentiment classifier. Compared to the existing sentiment analysis methods used in SE, the proposed approach can achieve significant improvement on representative benchmark datasets. By further contrast experiments, we find that the Tweets make a key contribution to the power of our approach. This finding informs future research not to unilaterally pursue the domain-specific resource, but try to transform knowledge from the open domain through ubiquitous signals such as emojis.

Sentiment classification typically relies on a large amount of labeled data. In practice, the availability of labels is highly imbalanced among different languages, e.g., more English texts are labeled than texts in any other languages, which creates a considerable inequality in the quality of related information services received by users speaking different languages. To tackle this problem, cross-lingual sentiment classification approaches aim to transfer knowledge learned from one language that has abundant labeled examples (i.e., the source language, usually English) to another language with fewer labels (i.e., the target language). The source and the target languages are usually bridged through off-the-shelf machine translation tools. Through such a channel, cross-language sentiment patterns can be successfully learned from English and transferred into the target languages. This approach, however, often fails to capture sentiment knowledge specific to the target language, and thus compromises the accuracy of the downstream classification task. In this paper, we employ emojis, which are widely available in many languages, as a new channel to learn both the cross-language and the language-specific sentiment patterns. We propose a novel representation learning method that uses emoji prediction as an instrument to learn respective sentiment-aware representations for each language. The learned representations are then integrated to facilitate cross-lingual sentiment classification. The proposed method demonstrates state-of-the-art performance on benchmark datasets, which is sustained even when sentiment labels are scarce.

OBJECTIVE: Active Learning (AL) attempts to reduce annotation cost (ie, time) by selecting the most informative examples for annotation. Most approaches tacitly (and unrealistically) assume that the cost for annotating each sample is identical. This study introduces a cost-aware AL method, which simultaneously models both the annotation cost and the informativeness of the samples and evaluates both via simulation and user studies. MATERIALS AND METHODS: We designed a novel, cost-aware AL algorithm (Cost-CAUSE) for annotating clinical named entities; we first utilized lexical and syntactic features to estimate annotation cost, then we incorporated this cost measure into an existing AL algorithm. Using the 2010 i2b2/VA data set, we then conducted a simulation study comparing Cost-CAUSE with noncost-aware AL methods, and a user study comparing Cost-CAUSE with passive learning. RESULTS: Our cost model fit empirical annotation data well, and Cost-CAUSE increased the simulation area under the learning curve (ALC) scores by up to 5.6% and 4.9%, compared with random sampling and alternate AL methods. Moreover, in a user annotation task, Cost-CAUSE outperformed passive learning on the ALC score and reduced annotation time by 20.5%-30.2%. DISCUSSION: Although AL has proven effective in simulations, our user study shows that a real-world environment is far more complex. Other factors have a noticeable effect on the AL method, such as the annotation accuracy of users, the tiredness of users, and even the physical and mental condition of users. CONCLUSION: Cost-CAUSE saves significant annotation cost compared to random sampling.

As a special kind of "big data," text data can be regarded as data reported by human sensors. Since humans are far more intelligent than physical sensors, text data contains useful information and knowledge about the real world, making it possible to make predictions about real-world phenomena based on text. As all application domains involve humans, text-based prediction has widespread applications, especially for optimization of decision making. While the problem of text-based prediction resembles text classification when formulated as a supervised learning problem, it is more challenging because the variable to be predicted may not be directly derivable from the text and thus there is a semantic gap between the target variable and the surface features that are often used for representing text data in conventional approaches. In this paper, we propose to bridge this gap by using knowledge graph to construct more effective features for text representation. We propose a two-step filtering algorithm to enhance such a knowledge-aware text representation for a family of entity-centric text regression tasks where the response variable can be treated as an attribute of a group of central entities. We evaluate the proposed algorithm by using two revenue prediction tasks based on reviews. The results show that the proposed algorithm can effectively leverage knowledge graphs to construct interpretable features, leading to significant improvement of the prediction accuracy over traditional features.

We introduce an adversarial method for producing high-recall explanations of neural text classifier decisions.Building on an existing architecture for extractive explanations via hard attention, we add an adversarial layer which scans the residual of the attention for remaining predictive signal.Motivated by the important domain of detecting personal attacks in social media comments, we additionally demonstrate the importance of manually setting a semantically appropriate "default" behavior for the model by explicitly manipulating its bias term.We develop a validation set of humanannotated personal attacks to evaluate the impact of these changes.

Much work has been done recently on learning word embeddings from large corpora, which attempts to find the coordinates of words in a static and high dimensional semantic space. In reality, such corpora often span a sufficiently long time period, during which the meanings of many words may have changed. The co-evolution of word meanings may also result in a distortion of the semantic space, making these static embeddings unable to accurately represent the dynamics of semantics. In this paper, we present a novel computational method to capture such changes and to model the evolution of word semantics. Distinct from existing approaches that learn word embeddings independently from time periods and then align them, our method explicitly establishes the stable topological structure of word semantics and identifies the surprising changes in the semantic space over time through a principled statistical method. Empirical experiments on large-scale real-world corpora demonstrate the effectiveness of the proposed approach, which outperforms the state-of-the-art by a large margin.

Based on a large data set of emoji using behavior collected from smartphone users over the world, this paper investigates gender-specific usage of emojis. We present various interesting findings that evidence a considerable difference in emoji usage by female and male users. Such a difference is significant not just in a statistical sense; it is sufficient for a machine learning algorithm to accurately infer the gender of a user purely based on the emojis used in their messages. In real world scenarios where gender inference is a necessity, models based on emojis have unique advantages over existing models that are based on textual or contextual information. Emojis not only provide language-independent indicators, but also alleviate the risk of leaking private user information through the analysis of text and metadata.

How to improve the quality of conversations in online communities has attracted considerable attention recently. Having engaged, civil, and reactive online conversations has a critical effect on the social life of Internet users. In this study, we are particularly interested in identifying a post in a multi-party conversation that is unlikely to be further replied to, which therefore kills that thread of the conversation. For this purpose, we propose a deep learning model called the ConverNet. ConverNet is attractive due to its capability of modeling the internal structure of a long conversation and its appropriate encoding of the contextual information of the conversation, through effective integration of attention mechanisms. Empirical experiments on real-world datasets demonstrate the effectiveness of the proposed model. For the widely concerned topic, our analysis also offers implications for how to improve the quality and user experience of online conversations, or how to engage users in a conversation with a chatbot.

Objective: Medical word sense disambiguation (WSD) is challenging and often requires significant training with data labeled by domain experts. This work aims to develop an interactive learning algorithm that makes efficient use of expert's domain knowledge in building high-quality medical WSD models with minimal human effort. Methods: We developed an interactive learning algorithm with expert labeling instances and features. An expert can provide supervision in 3 ways: labeling instances, specifying indicative words of a sense, and highlighting supporting evidence in a labeled instance. The algorithm learns from these labels and iteratively selects the most informative instances to ask for future labels. Our evaluation used 3 WSD corpora: 198 ambiguous terms from Medical Subject Headings (MSH) as MEDLINE indexing terms, 74 ambiguous abbreviations in clinical notes from the University of Minnesota (UMN), and 24 ambiguous abbreviations in clinical notes from Vanderbilt University Hospital (VUH). For each ambiguous term and each learning algorithm, a learning curve that plots the accuracy on the test set against the number of labeled instances was generated. The area under the learning curve was used as the primary evaluation metric. Results: Our interactive learning algorithm significantly outperformed active learning, the previous fastest learning algorithm for medical WSD. Compared to active learning, it achieved 90% accuracy for the MSH corpus with 42% less labeling effort, 35% less labeling effort for the UMN corpus, and 16% less labeling effort for the VUH corpus. Conclusions: High-quality WSD models can be efficiently trained with minimal supervision by inviting experts to label informative instances and provide domain knowledge through labeling/highlighting contextual features.

The prevalence of smart mobile devices has promoted the popularity of mobile applications (a.k.a. apps). Supporting mobility has become a promising trend in software engineering research. This article presents an empirical study of behavioral service profiles collected from millions of users whose devices are deployed with Wandoujia, a leading Android app-store service in China. The dataset of Wandoujia service profiles consists of two kinds of user behavioral data from using 0.28 million free Android apps, including (1) app management activities (i.e., downloading, updating, and uninstalling apps) from over 17 million unique users and (2) app network usage from over 6 million unique users. We explore multiple aspects of such behavioral data and present patterns of app usage. Based on the findings as well as derived knowledge, we also suggest some new open opportunities and challenges that can be explored by the research community, including app development, deployment, delivery, revenue, etc.

Modern search engines aggregate results from different verticals : webpages, news, images, video, shopping, knowledge cards, local maps, and so on. Unlike “ten blue links,” these search results are heterogeneous in nature and not even arranged in a list on the page. This revolution directly challenges the conventional “ranked list” formulation in ad hoc search. Therefore, finding proper presentation for a gallery of heterogeneous results is critical for modern search engines. We propose a novel framework that learns the optimal page presentation to render heterogeneous results onto search result page (SERP). Page presentation is broadly defined as the strategy to present a set of items on SERP, much more expressive than a ranked list. It can specify item positions, image sizes, text fonts, and any other styles as long as variations are within business and design constraints. The learned presentation is content aware, i.e., tailored to specific queries and returned results. Simulation experiments show that the framework automatically learns eye-catchy presentations for relevant results. Experiments on real data show that simple instantiations of the framework already outperform leading algorithm in federated search result presentation. It means the framework can learn its own result presentation strategy purely from data, without even knowing the “probability ranking principle.”

Effectively making sense of short texts is a critical task for many real world applications such as search engines, social media services, and recommender systems. The task is particularly challenging as a short text contains very sparse information, often too sparse for a machine learning algorithm to pick up useful signals. A common practice for analyzing short text is to first expand it with external information, which is usually harvested from a large collection of longer texts. In literature, short text expansion has been done with all kinds of heuristics. We propose an end-to-end solution that automatically learns how to expand short text to optimize a given learning task. A novel deep memory network is proposed to automatically find relevant information from a collection of longer documents and reformulate the short text through a gating mechanism. Using short text classification as a demonstrating task, we show that the deep memory network significantly outperforms classical text expansion methods with comprehensive experiments on real world data sets.

We consider the task of identifying attitudes towards a given set of entities from text. Conventionally, this task is decomposed into two separate subtasks: target detection that identifies whether each entity is mentioned in the text, either explicitly or implicitly, and polarity classification that classifies the exact sentiment towards an identified entity (the target) into positive, negative, or neutral.

Information cascades, effectively facilitated by most social network platforms, are recognized as a major factor in almost every social success and disaster in these networks. Can cascades be predicted? While many believe that they are inherently unpredictable, recent work has shown that some key properties of information cascades, such as size, growth, and shape, can be predicted by a machine learning algorithm that combines many features. These predictors all depend on a bag of hand-crafting features to represent the cascade network and the global network structures. Such features, always carefully and sometimes mysteriously designed, are not easy to extend or to generalize to a different platform or domain.

App marketplaces host millions of mobile apps that are downloaded billions of times. Investigating how people manage mobile apps in their everyday lives creates a unique opportunity to understand the behavior and preferences of mobile device users, infer the quality of apps, and improve user experience. Existing literature provides very limited knowledge about app management activities, due to the lack of app usage data at scale. This article takes the initiative to analyze a very large app management log collected through a leading Android app marketplace. The dataset covers 5 months of detailed downloading, updating, and uninstallation activities, which involve 17 million anonymized users and 1 million apps. We present a surprising finding that the metrics commonly used to rank apps in app stores do not truly reflect the users’ real attitudes. We then identify behavioral patterns from the app management activities that more accurately indicate user preferences of an app even when no explicit rating is available. A systematic statistical analysis is designed to evaluate machine learning models that are trained to predict user preferences using these behavioral patterns, which features an inverse probability weighting method to correct the selection biases in the training process.

Existing retrieval systems rely on a single active query to pull documents from the index. Relevance feedback may be used to iteratively refine the query, but only one query is active at a time. If the user's information need has multiple aspects, the query must represent the union of these aspects. We consider a new paradigm of retrieval where multiple queries are kept ``active'' simultaneously. In the presence of rate limits, the active queries take turns accessing the index to retrieve another ``page'' of results. Turns are assigned by a multi-armed bandit based on user feedback. This allows the system to explore which queries return more relevant results and to exploit the best ones. In empirical tests, query pools outperform solo, combined queries. Significant improvement is observed both when the subtopic queries are known in advance and when the queries are generated in a user-interactive process.

Emojis have been widely used to simplify emotional expression and enrich user experience. As an interesting practice of ubiquitous computing, emojis are adopted by Internet users from many different countries, on many devices (particularly popular on smartphones), and in many applications. The "ubiquitous" usage of emojis enables us to study and compare user behaviors and preferences across countries and cultures. We present an analysis on how smartphone users use emojis based on a very large data set collected from a popular emoji keyboard. The data set contains a complete month of emoji usage of 3.88 million active users from 212 countries and regions. We demonstrate that the categories and frequencies of emojis used by these users provide rich signals for the identification and the understanding of cultural differences of smartphone users. Users from different countries present significantly different preferences on emojis, which complies with the well-known Hofstede's cultural dimensions model.

Selecting and prioritizing major device models are critical for mobile app developers to select testbeds and optimize resources such as marketing and quality-assurance resources. The heavily fragmented distribution of Android devices makes it challenging to select a few major device models out of thousands of models available on the market. Currently app developers usually rely on some reported or estimated general market share of device models. However, these estimates can be quite inaccurate, and more problematically, can be irrelevant to the particular app under consideration. To address this issue, we propose PRADA, the first approach to prioritizing Android device models for individual apps, based on mining large-scale usage data. PRADA adapts the concept of operational profiling (popularly used in software reliability engineering) for mobile apps -- the usage of an app on a specific device model reflects the importance of that device model for the app. PRADA includes a collaborative filtering technique to predict the usage of an app on different device models, even if the app is entirely new (without its actual usage in the market yet), based on the usage data of a large collection of apps. We empirically demonstrate the effectiveness of PRADA over two popular app categories, i.e., Game and Media, covering over 3.86 million users and 14,000 device models collected through a leading Android management app in China.

A key challenge of entity set expansion is that multifaceted input seeds can lead to significant incoherence in the result set. In this paper, we present a novel solution to handling multifaceted seeds by combining existing user-generated ontologies with a novel word-similarity metric based on skip-grams. By blending the two resources we are able to produce sparse word ego-networks that are centered on the seed terms and are able to capture semantic equivalence among words. We demonstrate that the resulting networks possess internally-coherent clusters, which can be exploited to provide non-overlapping expansions, in order to reflect different semantic classes of the seeds. Empirical evaluation against state-of-the-art baselines shows that our solution, EgoSet, is able to not only capture multiple facets in the input query, but also generate expansions for each facet with higher precision.

Modern search engines aggregate results from different verticals: webpages, news, images, video, shopping, knowledge cards, local maps, etc. Unlike "ten blue links", these search results are heterogeneous in nature and not even arranged in a list on the page. This revolution directly challenges the conventional "ranked list" formulation in ad hoc search. Therefore, finding proper presentation for a gallery of heterogeneous results is critical for modern search engines.

Smartphone users have adopted an explosive number of mobile applications (a.k.a., apps) in the recent years. App marketplaces for iOS, Android and Windows Phone platforms host millions of apps which have been downloaded for more than 100 billion times. Investigating how people manage mobile apps in their everyday lives creates a unique opportunity to understand the behavior and preferences of mobile users, to infer the quality of apps, and to improve the user experience. Existing literature provides very limited knowledge about app management activities, due to the lack of user behavioral data at scale. This paper takes the initiative to analyze a very large app management log collected through a leading Android app marketplace. The data set covers five months of detailed downloading, updating, and uninstallation activities, involving 17 million anonymized users and one million apps. We present a surprising finding that the metrics commonly used by app stores to rank apps do not truly reflect the users' real attitudes towards the apps. We then identify useful patterns from the app management activities that much more accurately predict the user preferences of an app even when no user rating is available.

We study the problem of visualizing large-scale and high-dimensional data in a low-dimensional (typically 2D or 3D) space. Much success has been reported recently by techniques that first compute a similarity structure of the data points and then project them into a low-dimensional space with the structure preserved. These two steps suffer from considerable computational costs, preventing the state-of-the-art methods such as the t-SNE from scaling to large-scale and high-dimensional data (e.g., millions of data points and hundreds of dimensions). We propose the LargeVis, a technique that first constructs an accurately approximated K-nearest neighbor graph from the data and then layouts the graph in the low-dimensional space. Comparing to t-SNE, LargeVis significantly reduces the computational cost of the graph construction step and employs a principled probabilistic model for the visualization step, the objective of which can be effectively optimized through asynchronous stochastic gradient descent with a linear time complexity. The whole procedure thus easily scales to millions of high-dimensional data points. Experimental results on real-world data sets demonstrate that the LargeVis outperforms the state-of-the-art methods in both efficiency and effectiveness. The hyper-parameters of LargeVis are also much more stable over different data sets.

OBJECTIVE: Traditional Chinese medicine (TCM) is a unique and complex medical system that has developed over thousands of years. This article studies the problem of automatically extracting meaningful relations of entities from TCM literature, for the purposes of assisting clinical treatment or poly-pharmacology research and promoting the understanding of TCM in Western countries. METHODS: Instead of separately extracting each relation from a single sentence or document, we propose to collectively and globally extract multiple types of relations (eg, herb-syndrome, herb-disease, formula-syndrome, formula-disease, and syndrome-disease relations) from the entire corpus of TCM literature, from the perspective of network mining. In our analysis, we first constructed heterogeneous entity networks from the TCM literature, in which each edge is a candidate relation, then used a heterogeneous factor graph model (HFGM) to simultaneously infer the existence of all the edges. We also employed a semi-supervised learning algorithm estimate the model's parameters. RESULTS: We performed our method to extract relations from a large dataset consisting of more than 100,000 TCM article abstracts. Our results show that the performance of the HFGM at extracting all types of relations from TCM literature was significantly better than a traditional support vector machine (SVM) classifier (increasing the average precision by 11.09%, the recall by 13.83%, and the F1-measure by 12.47% for different types of relations, compared with a traditional SVM classifier). CONCLUSION: This study exploits the power of collective inference and proposes an HFGM based on heterogeneous entity networks, which significantly improved our ability to extract relations from TCM literature.

OBJECTIVE: ClinicalTrials.gov serves critical functions of disseminating trial information to the public and helping the trials recruit participants. This study assessed the readability of trial descriptions at ClinicalTrials.gov using multiple quantitative measures. MATERIALS AND METHODS: The analysis included all 165,988 trials registered at ClinicalTrials.gov as of April 30, 2014. To obtain benchmarks, the authors also analyzed 2 other medical corpora: (1) all 955 Health Topics articles from MedlinePlus and (2) a random sample of 100,000 clinician notes retrieved from an electronic health records system intended for conveying internal communication among medical professionals. The authors characterized each of the corpora using 4 surface metrics, and then applied 5 different scoring algorithms to assess their readability. The authors hypothesized that clinician notes would be most difficult to read, followed by trial descriptions and MedlinePlus Health Topics articles. RESULTS: Trial descriptions have the longest average sentence length (26.1 words) across all corpora; 65% of their words used are not covered by a basic medical English dictionary. In comparison, average sentence length of MedlinePlus Health Topics articles is 61% shorter, vocabulary size is 95% smaller, and dictionary coverage is 46% higher. All 5 scoring algorithms consistently rated CliniclTrials.gov trial descriptions the most difficult corpus to read, even harder than clinician notes. On average, it requires 18 years of education to properly understand these trial descriptions according to the results generated by the readability assessment algorithms. DISCUSSION AND CONCLUSION: Trial descriptions at CliniclTrials.gov are extremely difficult to read. Significant work is warranted to improve their readability in order to achieve CliniclTrials.gov's goal of facilitating information dissemination and subject recruitment.

The popularity of smartphones and tablet computers in recent years makes mobile apps burst. Mobile apps have become the main consumers of the Internet-based services. Compared to traditional applications in the desktop computing era, mobile devices with their apps bring new opportunities and challenges to service computing community, in various aspects like service publication, discovery, interaction, composition, et al. In this paper, we propose a novel data-driven, content-based mobile apps composition approach, called Mash Droid, by leveraging a novel In-App Search mechanism, i.e., Discovering relevant services for the data and content in apps. Rather than existing techniques that usually integrate fixed Web services, our approach relies on the dynamic service discovery and flexible data exchange between several apps. The unique feature of our approach is enabling the data communication channel between apps by the content index services provided by a leading Android appstore, Wandoujia, which now has over 1,000,000 apps and 200 million users. We employ the In-App Search mechanism to define a Restful-style app model and resource-oriented app description model. Based on the models, we design a framework for dynamically discovering relevant apps that could be composed with current app's contexts. We implement a prototype to demonstrate our approach.

he prevalence of smart devices has promoted the popular- ity of mobile applications (a.k.a. apps) in recent years. A number of interesting and important questions remain unan- swered, such as why a user likes/dislikes an app, how an app becomes popular or eventually perishes, how a user selects apps to install and interacts with them, how frequently an app is used and how much traffic it generates, etc. This paper presents an empirical analysis of app usage behaviors collected from millions of users of Wandoujia, a leading An- droid app marketplace in China. The dataset covers two types of user behaviors of using over 0.2 million Android apps, including (1) app management activities (i.e., installa- tion, updating, and uninstallation) of over 0.8 million unique users and (2) app network traffic from over 2 million unique users. We explore multiple aspects of such behavior data and present interesting patterns of app usage. The results provide many useful implications to the developers, users, and disseminators of mobile apps.

We present the problem of click-through prediction for advertising in Twitter timeline, which displays a stream of Tweets from accounts a user choose to follow. Traditional computational advertising usually appears in two forms: sponsored search that places ads onto the search result page when a query is issued to a search engine, and contextual advertising that places ads onto a regular, usually static Web page. Compared with these two paradigms, placing ads into a Tweet stream is particularly challenging given the nature of the data stream: the context into which an ad can be placed updates dynamically and never replicates. Every ad is therefore placed into a unique context. This makes the information available for training a machine learning model extremely sparse. In this study, we propose a learning-to-rank method which not only addresses the sparsity of training signals but also can be trained and updated online. The proposed method is evaluated using both offline experiments and online A/B tests, which involve very large collections of Twitter data and real Twitter users. Results of the experiments prove the effectiveness and efficiency of our solution, and its superiority over the current production model adopted by Twitter.

Unsupervised text embedding methods, such as Skip-gram and Paragraph Vector, have been attracting increasing attention due to their simplicity, scalability, and effectiveness. However, comparing to sophisticated deep learning architectures such as convolutional neural networks, these methods usually yield inferior results when applied to particular machine learning tasks. One possible reason is that these text embedding methods learn the representation of text in a fully unsupervised way, without leveraging the labeled information available for the task. Although the low dimensional representations learned are applicable to many different tasks, they are not particularly tuned for any task. In this paper, we fill this gap by proposing a semi-supervised representation learning method for text data, which we call the predictive text embedding (PTE). Predictive text embedding utilizes both labeled and unlabeled data to learn the embedding of text. The labeled information and different levels of word co-occurrence information are first represented as a large-scale heterogeneous text network, which is then embedded into a low dimensional space through a principled and efficient algorithm. This low dimensional embedding not only preserves the semantic closeness of words and documents, but also has a strong predictive power for the particular task. Compared to recent supervised approaches based on convolutional neural networks, predictive text embedding is comparable or more effective, much more efficient, and has fewer parameters to tune.

This paper studies the problem of embedding very large information networks into low-dimensional vector spaces, which is useful in many tasks such as visualization, node classification, and link prediction. Most existing graph embedding methods do not scale for real world information networks which usually contain millions of nodes. In this paper, we propose a novel network embedding method called the ``LINE,'' which is suitable for arbitrary types of information networks: undirected, directed, and/or weighted. The method optimizes a carefully designed objective function that preserves both the local and global network structures. An edge-sampling algorithm is proposed that addresses the limitation of the classical stochastic gradient descent and improves both the effectiveness and the efficiency of the inference. Empirical experiments prove the effectiveness of the LINE on a variety of real-world information networks, including language networks, social networks, and citation networks. The algorithm is very efficient, which is able to learn the embedding of a network with millions of vertices and billions of edges in a few hours on a typical single machine. The source code of the LINE is available online\footnote{\url{https://github.com/tangjianpku/LINE}}.

Many previous techniques identify trending topics in social media, even topics that are not pre-defined. We present a technique to identify trending rumors, which we define as topics that include disputed factual claims. Putting aside any attempt to assess whether the rumors are true or false, it is valuable to identify trending rumors as early as possible. It is extremely difficult to accurately classify whether every individual post is or is not making a disputed factual claim. We are able to identify trending rumors by recasting the problem as finding entire clusters of posts whose topic is a disputed factual claim.

OBJECTIVE: This paper describes the University of Michigan's nine-year experience in developing and using a full-text search engine designed to facilitate information retrieval (IR) from narrative documents stored in electronic health records (EHRs). The system, called the Electronic Medical Record Search Engine (EMERSE), functions similar to Google but is equipped with special functionalities for handling challenges unique to retrieving information from medical text. MATERIALS AND METHODS: Key features that distinguish EMERSE from general-purpose search engines are discussed, with an emphasis on functions crucial to (1) improving medical IR performance and (2) assuring search quality and results consistency regardless of users' medical background, stage of training, or level of technical expertise. RESULTS: Since its initial deployment, EMERSE has been enthusiastically embraced by clinicians, administrators, and clinical and translational researchers. To date, the system has been used in supporting more than 750 research projects yielding 80 peer-reviewed publications. In several evaluation studies, EMERSE demonstrated very high levels of sensitivity and specificity in addition to greatly improved chart review efficiency. DISCUSSION: Increased availability of electronic data in healthcare does not automatically warrant increased availability of information. The success of EMERSE at our institution illustrates that free-text EHR search engines can be a valuable tool to help practitioners and researchers retrieve information from EHRs more effectively and efficiently, enabling critical tasks such as patient case synthesis and research data abstraction. CONCLUSION: EMERSE, available free of charge for academic use, represents a state-of-the-art medical IR tool with proven effectiveness and user acceptance.

Topic models such as the latent Dirichlet allocation (LDA) have become a standard staple in the modeling toolbox of machine learning. They have been applied to a vast variety of data sets, contexts, and tasks to varying degrees of success. However, to date there is almost no formal theory explicating the LDA’s behavior, and despite its familiarity there is very little systematic analysis of and guidance on the properties of the data that affect the inferential performance of the model. This paper seeks to address this gap, by providing a systematic analysis of factors which characterize the LDA’s performance. We present theorems elucidating the posterior contraction rates of the topics as the amount of data increases, and a thorough supporting empirical study using synthetic and real data sets, including news and web-based articles and tweet messages. Based on these results we provide practical guidance on how to identify suitable data sets for topic models, and how to specify particular model parameters.

Hashtags have been widely used to annotate topics in tweets (short posts on Twitter.com). In this paper, we study the problems of real-time prediction of bursting hashtags. Will a hashtag burst in the near future? If it will, how early can we predict it, and how popular will it become? Based on empirical analysis of data collected from Twitter, we propose solutions to these challenging problems. The performance of different features and possible solutions are evaluated.

We consider a scenario where a searcher requires both high precision and high recall from an interactive retrieval process. Such scenarios are very common in real life, exemplified by medical search, legal search, market research, and literature review. When access to the entire data set is available, an active learning loop could be used to ask for additional relevance feedback labels in order to refine a classifier. When data is accessed via search services, however, only limited subsets of the corpus can be considered, subsets defined by queries. In that setting, relevance feedback has been used in a query enhancement loop that updates a query.

Topic modeling has been proved to be an effective method for exploratory text mining. It is a common assumption of most topic models that a document is generated from a mixture of topics. In real-world scenarios, individual documents usually concentrate on several salient topics instead of covering a wide variety of topics. A real topic also adopts a narrow range of terms instead of a wide coverage of the vocabulary. Understanding this sparsity of information is especially important for analyzing user-generated Web content and social media, which are featured as extremely short posts and condensed discussions. In this paper, we propose a dual-sparse topic model that addresses the sparsity in both the topic mixtures and the word usage. By applying a "Spike and Slab" prior to decouple the sparsity and smoothness of the document-topic and topic-word distributions, we allow individual documents to select a few focused topics and a topic to select focused terms, respectively. Experiments on different genres of large corpora demonstrate that the dual-sparse topic model outperforms both classical topic models and existing sparsity-enhanced topic models. This improvement is especially notable on collections of short documents.

OBJECTIVE: We describe experiments designed to determine the feasibility of distinguishing known from novel associations based on a clinical dataset comprised of International Classification of Disease, V.9 (ICD-9) codes from 1.6 million patients by comparing them to associations of ICD-9 codes derived from 20.5 million Medline citations processed using MetaMap. Associations appearing only in the clinical dataset, but not in Medline citations, are potentially novel. METHODS: Pairwise associations of ICD-9 codes were independently identified in both the clinical and Medline datasets, which were then compared to quantify their degree of overlap. We also performed a manual review of a subset of the associations to validate how well MetaMap performed in identifying diagnoses mentioned in Medline citations that formed the basis of the Medline associations. RESULTS: The overlap of associations based on ICD-9 codes in the clinical and Medline datasets was low: only 6.6% of the 3.1 million associations found in the clinical dataset were also present in the Medline dataset. Further, a manual review of a subset of the associations that appeared in both datasets revealed that co-occurring diagnoses from Medline citations do not always represent clinically meaningful associations. DISCUSSION: Identifying novel associations derived from large clinical datasets remains challenging. Medline as a sole data source for existing knowledge may not be adequate to filter out widely known associations. CONCLUSIONS: In this study, novel associations were not readily identified. Further improvements in accuracy and relevance for tools such as MetaMap are needed to realize their expected utility.

The spreading of innovations among individuals and organizations in a social network has been extensively studied. Although the recent studies among the social computing and data mining communities have produced various insightful conclusions about the diffusion process of innovations by focusing on the properties and evolution of social network structures, less attention has been paid to the interrelationships among the multiple innovations being diffused, such as the competitive and collaborative relationships between innovations. In this paper, we take a formal quantitative approach to address how different pieces of innovations socialize with each other and how the interrelationships among innovations affect users' adoption behavior, which provides a novel perspective of understanding the diffusion of innovations. Networks of innovations are constructed by mining large scale text collections in an unsupervised fashion. We are particularly interested in the following questions: what are the meaningful metrics on the network of innovations? What effects do these metrics exert on the diffusion of innovations? Do these effects vary among users with different adoption preferences or communication styles? While existing studies primarily address social influence, we provide a detailed discussion of how innovations interrelate and influence the diffusion process.

New challenges have been presented to classical topic models when applied to social media, as user-generated content suffers from significant problems of data sparseness. A variety of heuristic adjustments to these models have been proposed, many of which are based on the use of context information to improve the performance of topic modeling. Existing contextualized topic models rely on arbitrary manipulation of the model structure, by incorporating various context variables into the generative process of classical topic models in an ad hoc manner. Such manipulations usually result in much more complicated model structures, sophisticated inference procedures, and low generalizability to accommodate arbitrary types or combinations of contexts. In this paper we explore a different direction. We propose a general solution that is able to exploit multiple types of contexts without arbitrary manipulation of the structure of classical topic models. We formulate different types of contexts as multiple views of the partition of the corpus. A co-regularization framework is proposed to let these views collaborate with each other, vote for the consensus topics, and distinguish them from view-specific topics. Experiments with real-world datasets prove that the proposed method is both effective and flexible to handle arbitrary types of contexts.

Twitter has become one of the most popular platforms for users to share information in real time. However, as an individual tweet is short and lacks sufficient contextual information, users cannot effectively understand or consume information on Twitter, which can either make users less engaged or even detached from using Twitter. In order to provide informative context to a Twitter user, we propose the task of Twitter context summarization, which generates a succinct summary from a large but noisy Twitter context tree. Traditional summarization techniques only consider text information, which is insufficient for Twitter context summarization task, since text information on Twitter is very sparse. Given that there are rich user interactions in Twitter, we thus study how to improve summarization methods by leveraging such signals. In particular, we study how user influence models, which project user interaction information onto a Twitter context tree, can help Twitter context summarization within a supervised learning framework. To evaluate our methods, we construct a data set by asking human editors to manually select the most informative tweets as a summary. Our experimental results based on this editorial data set show that Twitter context summarization is a promising research topic and pairwise user influence signals can significantly improve the task performance.

Conventional studies of online information seeking behavior usually focus on the use of search engines or question answering (Q&A) websites. Recently, the fast growth of online social platforms such as Twitter and Facebook has made it possible for people to utilize them for information seeking by asking questions to their friends or followers. We anticipate a better understanding of Web users' information needs by investigating research questions about these questions. How are they distinctive from daily tweeted conversations? How are they related to search queries? Can users' information needs on one platform predict those on the other?

OBJECTIVES: This study was to assess whether active learning strategies can be integrated with supervised word sense disambiguation (WSD) methods, thus reducing the number of annotated samples, while keeping or improving the quality of disambiguation models. METHODS: We developed support vector machine (SVM) classifiers to disambiguate 197 ambiguous terms and abbreviations in the MSH WSD collection. Three different uncertainty sampling-based active learning algorithms were implemented with the SVM classifiers and were compared with a passive learner (PL) based on random sampling. For each ambiguous term and each learning algorithm, a learning curve that plots the accuracy computed from the test set as a function of the number of annotated samples used in the model was generated. The area under the learning curve (ALC) was used as the primary metric for evaluation. RESULTS: Our experiments demonstrated that active learners (ALs) significantly outperformed the PL, showing better performance for 177 out of 197 (89.8%) WSD tasks. Further analysis showed that to achieve an average accuracy of 90%, the PL needed 38 annotated samples, while the ALs needed only 24, a 37% reduction in annotation effort. Moreover, we analyzed cases where active learning algorithms did not achieve superior performance and identified three causes: (1) poor models in the early learning stage; (2) easy WSD cases; and (3) difficult WSD cases, which provide useful insight for future improvements. CONCLUSIONS: This study demonstrated that integrating active learning strategies with supervised WSD methods could effectively reduce annotation cost and improve the disambiguation models.

Diversified ranking is a fundamental task in machine learning. It is broadly applicable in many real world problems, e.g., information retrieval, team assembling, product search, etc. In this paper, we consider a generic setting where we aim to diversify the top-k ranking list based on an arbitrary relevance function and an arbitrary similarity function among all the examples. We formulate it as an optimization problem and show that in general it is NP-hard. Then, we show that for a large volume of the parameter space, the proposed objective function enjoys the diminishing returns property, which enables us to design a scalable, greedy algorithm to find the near-optimal solution. Experimental results on real data sets demonstrate the effectiveness of the proposed algorithm.

As a new paradigm of online communities, microfinance sites such as Kiva.org have attracted much public attention. To understand lender motivations on Kiva, we classify the lenders' self-stated motivations into ten categories with human coders and machine learning based classifiers. We employ text classifiers using lexical features, along with social features based on lender activity information on Kiva, to predict the categories of lender motivation statements. Although the task appears to be much more challenging than traditional topic-based categorization, our classifiers can achieve high precision in most categories. Using the results of this classification along with Kiva teams information, we predict lending activity from lender motivation and team affiliations. Finally, we make design recommendations regarding Kiva practices which might increase pro-social lending.

Researchers and social observers have both believed that hashtags, as a new type of organizational objects of information, play a dual role in online microblogging communities (e.g., Twitter). On one hand, a hashtag serves as a bookmark of content, which links tweets with similar topics; on the other hand, a hashtag serves as the symbol of a community membership, which bridges a virtual community of users. Are the real users aware of this dual role of hashtags? Is the dual role affecting their behavior of adopting a hashtag? Is hashtag adoption predictable? We take the initiative to investigate and quantify the effects of the dual role on hashtag adoption. We propose comprehensive measures to quantify the major factors of how a user selects content tags as well as joins communities. Experiments using large scale Twitter datasets prove the effectiveness of the dual role, where both the content measures and the community measures significantly correlate to hashtag adoption on Twitter. With these measures as features, a machine learning model can effectively predict the future adoption of hashtags that a user has never used before.

The prevalence of Web 2.0 techniques has led to the boom of various online communities, where topics spread ubiquitously among user-generated documents. Working together with this diffusion process is the evolution of topic content, where novel contents are introduced by documents which adopt the topic. Unlike explicit user behavior (e.g., buying a DVD), both the diffusion paths and the evolutionary process of a topic are implicit, making their discovery challenging. In this paper, we track the evolution of an arbitrary topic and reveal the latent diffusion paths of that topic in a social community. A novel and principled probabilistic model is proposed which casts our task as an joint inference problem, which considers textual documents, social influences, and topic evolution in a unified way. Specifically, a mixture model is introduced to model the generation of text according to the diffusion and the evolution of the topic, while the whole diffusion process is regularized with user-level social influences through a Gaussian Markov Random Field. Experiments on both synthetic data and real world data show that the discovery of topic diffusion and evolution benefits from this joint inference, and the probabilistic model we propose performs significantly better than existing methods.

OBJECTIVE: A full-text search engine can be a useful tool for augmenting the reuse value of unstructured narrative data stored in electronic health records (EHR). A prominent barrier to the effective utilization of such tools originates from users' lack of search expertise and/or medical-domain knowledge. To mitigate the issue, the authors experimented with a 'collaborative search' feature through a homegrown EHR search engine that allows users to preserve their search knowledge and share it with others. This feature was inspired by the success of many social information-foraging techniques used on the web that leverage users' collective wisdom to improve the quality and efficiency of information retrieval. DESIGN: The authors conducted an empirical evaluation study over a 4-year period. The user sample consisted of 451 academic researchers, medical practitioners, and hospital administrators. The data were analyzed using a social-network analysis to delineate the structure of the user collaboration networks that mediated the diffusion of knowledge of search. RESULTS: The users embraced the concept with considerable enthusiasm. About half of the EHR searches processed by the system (0.44 million) were based on stored search knowledge; 0.16 million utilized shared knowledge made available by other users. The social-network analysis results also suggest that the user-collaboration networks engendered by the collaborative search feature played an instrumental role in enabling the transfer of search knowledge across people and domains. CONCLUSION: Applying collaborative search, a social information-foraging technique popularly used on the web, may provide the potential to improve the quality and efficiency of information retrieval in healthcare.

The growth of the web has directly influenced the increase in the availability of relational data. One of the key problems in mining such data is computing the similarity between objects with heterogeneous feature types. For example, publications have many heterogeneous features like text, citations, authorship information, venue information, etc. In most approaches, similarity is estimated using each feature type in isolation and then combined in a linear fashion. However, this approach does not take advantage of the dependencies between the different feature spaces. In this paper, we propose a novel approach to combine the different sources of similarity using a regularization framework over edges in multiple graphs. We show that the objective function induced by the framework is convex. We also propose an efficient algorithm using coordinate descent to solve the optimization problem. We extrinsically evaluate the performance of the proposed unified similarity measure on two different tasks, clustering and classification. The proposed similarity measure outperforms three baselines and a state-of-the-art classification algorithm on a variety of standard, large data sets.

User generated information in online communities has been characterized with the mixture of a text stream and a network structure both changing over time. A good example is a web-blogging community with the daily blog posts and a social network of bloggers.

Information networks are widely used to characterize the relationships between data items such as text documents. Many important retrieval and mining tasks rely on ranking the data items based on their centrality or prestige in the network. Beyond prestige, diversity has been recognized as a crucial objective in ranking, aiming at providing a non-redundant and high coverage piece of information in the top ranked results. Nevertheless, existing network-based ranking approaches either disregard the concern of diversity, or handle it with non-optimized heuristics, usually based on greedy vertex selection.

In this paper we present a general framework to study sequences of search activities performed by a user. Our framework provides (i) a vocabulary to discuss types of features, models, and tasks, (ii) straightforward feature re-use across problems, (iii) realistic baselines for many sequence analysis tasks we study, and (iv) a simple mechanism to develop baselines for sequence analysis tasks beyond those studied in this paper. Using this framework we study a set of fourteen sequence analysis tasks with a range of features and models. While we show that most tasks benefit from features based on recent history, we also identify two categories of "sequence-resistant" tasks for which simple classes of local features perform as well as richer features and models.

This paper studies web object classification problem with the novel exploration of social tags. Automatically classifying web objects into manageable semantic categories has long been a fundamental preprocess for indexing, browsing, searching, and mining these objects. The explosive growth of heterogeneous web objects, especially non-textual objects such as products, pictures, and videos, has made the problem of web classification increasingly challenging. Such objects often suffer from a lack of easy-extractable features with semantic information, interconnections between each other, as well as training examples with category labels.

Social tagging is becoming increasingly popular in many Web 2.0 applications where users can annotate resources (e.g. Web pages) with arbitrary keywords (i.e. tags). A tag recommendation module can assist users in tagging process by suggesting relevant tags to them. It can also be directly used to expand the set of tags annotating a resource. The benefits are twofold: improving user experience and enriching the index of resources. However, the former one is not emphasized in previous studies, though a lot of work has reported that different users may describe the same concept in different ways. We address the problem of personalized tag recommendation for text documents. In particular, we model personalized tag recommendation as a "query and ranking" problem and propose a novel graph-based ranking algorithm for interrelated multi-type objects. When a user issues a tagging request, both the document and the user are treated as a part of the query. Tags are then ranked by our graph-based ranking algorithm which takes into consideration both relevance to the document and preference of the user. Finally, the top ranked tags are presented to the user as suggestions. Experiments on a large-scale tagging data set collected from Del.icio.us have demonstrated that our proposed algorithm significantly outperforms algorithms which fail to consider the diversity of different users' interests.

Promotion is one of the key ingredients in marketing. It is often desirable to find merit in an object (e.g., product, person, organization, or service) and promote it in an appropriate community. In this paper, we propose a novel functionality, called promotion analysis through ranking , for promoting a given object by leveraging highly ranked results. Since the object may not be highly ranked in the global space, our goal is to discover promotive subspaces in which the object becomes prominent. To achieve this goal, the notion of promotiveness is formulated. We show that this functionality is practical and useful in a wide variety of applications such as business intelligence. However, computing promotive subspaces is challenging due to the explosion of search space and high aggregation cost. For efficient computation, we propose a PromoRank framework, and develop three efficient optimization techniques, namely subspace pruning, object pruning, and promotion cube, which are seamlessly integrated into the framework. Our empirical evaluation on two real data sets confirms the effectiveness of promotion analysis, and that our proposed algorithms significantly outperform baseline solutions.

Topic modeling has been a key problem for document analysis. One of the canonical approaches for topic modeling is Probabilistic Latent Semantic Indexing, which maximizes the joint probability of documents and terms in the corpus. The major disadvantage of PLSI is that it estimates the probability distribution of each document on the hidden topics independently and the number of parameters in the model grows linearly with the size of the corpus, which leads to serious problems with overfitting. Latent Dirichlet Allocation (LDA) is proposed to overcome this problem by treating the probability distribution of each document over topics as a hidden random variable. Both of these two methods discover the hidden topics in the Euclidean space. However, there is no convincing evidence that the document space is Euclidean, or flat. Therefore, it is more natural and reasonable to assume that the document space is a manifold, either linear or nonlinear. In this paper, we consider the problem of topic modeling on intrinsic document manifold. Specifically, we propose a novel algorithm called Laplacian Probabilistic Latent Semantic Indexing (LapPLSI) for topic modeling. LapPLSI models the document space as a submanifold embedded in the ambient space and directly performs the topic modeling on this document manifold in question. We compare the proposed LapPLSI approach with PLSI and LDA on three text data sets. Experimental results show that LapPLSI provides better representation in the sense of semantic structure.

Generating alternative queries, also known as query suggestion, has long been proved useful to help a user explore and express his information need. In many scenarios, such suggestions can be generated from a large scale graph of queries and other accessory information, such as the clickthrough. However, how to generate suggestions while ensuring their semantic consistency with the original query remains a challenging problem.

A common task in many text mining applications is to generate a multi-faceted overview of a topic in a text collection. Such an overview not only directly serves as an informative summary of the topic, but also provides a detailed view of navigation to different facets of the topic. Existing work has cast this problem as a categorization problem and requires training examples for each facet. This has three limitations: (1) All facets are predefined, which may not fit the need of a particular user. (2) Training examples for each facet are often unavailable. (3) Such an approach only works for a predefined type of topics. In this paper, we break these limitations and study a more realistic new setup of the problem, in which we would allow a user to flexibly describe each facet with keywords for an arbitrary topic and attempt to mine a multi-faceted overview in an unsupervised way. We attempt a probabilistic approach to solve this problem. Empirical experiments on different genres of text data show that our approach can effectively generate a multi-faceted overview for arbitrary topics; the generated overviews are comparable with those generated by supervised methods with training examples. They are also more informative than unstructured flat summaries. The method is quite general, thus can be applied to multiple text mining tasks in different application domains.

Recent work on language models for information retrieval has shown that smoothing language models is crucial for achieving good retrieval performance. Many different effective smoothing methods have been proposed, which mostly implement various heuristics to exploit corpus structures. In this paper, we propose a general and unified optimization framework for smoothing language models on graph structures. This framework not only provides a unified formulation of the existing smoothing heuristics, but also serves as a road map for systematically exploring smoothing methods for language models. We follow this road map and derive several different instantiations of the framework. Some of the instantiations lead to novel smoothing methods. Empirical results show that all such instantiations are effective with some outperforming the state of the art smoothing methods.

How many pages are there on the Web? 5B? 20B? More? Less? Big bets on clusters in the clouds could be wiped out if a small cache of a few million urls could capture much of the value. Language modeling techniques are applied to MSN's search logs to estimate entropy. The perplexity is surprisingly small: millions, not billions.

In this paper, we formally define the problem of topic modeling with network structure (TMN). We propose a novel solution to this problem, which regularizes a statistical topic model with a harmonic regularizer based on a graph structure in the data. The proposed method bridges topic modeling and social network analysis, which leverages the power of both statistical topic models and discrete regularization. The output of this model well summarizes topics in text, maps a topic on the network, and discovers topical communities. With concrete selection of a topic model and a graph-based regularizer, our model can be applied to text mining problems such as author-topic analysis, community discovery, and spatial text mining. Empirical experiments on two different genres of data show that our approach is effective, which improves text-oriented methods as well as network-oriented methods. The proposed model is general; it can be applied to any text collections with a mixture of topics and an associated network structure.

Multinomial distributions over words are frequently used to model topics in text collections. A common, major challenge in applying all such topic models to any text mining problem is to label a multinomial topic model accurately so that a user can interpret the discovered topic. So far, such labels have been generated manually in a subjective way. In this paper, we propose probabilistic approaches to automatically labeling multinomial topic models in an objective way. We cast this labeling problem as an optimization problem involving minimizing Kullback-Leibler divergence between word distributions and maximizing mutual information between a label and a topic model. Experiments with user study have been done on two text data sets with different genres.The results show that the proposed labeling methods are quite effective to generate labels that are meaningful and useful for interpreting the discovered topic models. Our methods are general and can be applied to labeling topics learned through all kinds of topic models such as PLSA, LDA, and their variations.

Many variants of language models have been proposed for information retrieval. Most existing models are based on multinomial distribution and would score documents based on query likelihood computed based on a query generation probabilistic model. In this paper, we propose and study a new family of query generation models based on Poisson distribution. We show that while in their simplest forms, the new family of models and the existing multinomial models are equivalent. However, based on different smoothing methods, the two families of models behave differently. We show that the Poisson model has several advantages, including naturally accommodating per-term smoothing and modeling accurate background more efficiently. We present several variants of the new model corresponding to different smoothing methods, and evaluate them on four representative TREC test collections. The results show that while their basic models perform comparably, the Poisson model can out perform multinomial model with per-term smoothing. The performance can be further improved with two-stage smoothing.

In this paper, we define the problem of topic-sentiment analysis on Weblogs and propose a novel probabilistic model to capture the mixture of topics and sentiments simultaneously. The proposed Topic-Sentiment Mixture (TSM) model can reveal the latent topical facets in a Weblog collection, the subtopics in the results of an ad hoc query, and their associated sentiments. It could also provide general sentiment models that are applicable to any ad hoc topics. With a specifically designed HMM structure, the sentiment models and topic models estimated with TSM can be utilized to extract topic life cycles and sentiment dynamics. Empirical experiments on different Weblog datasets show that this approach is effective for modeling the topic facets and sentiments and extracting their dynamics from Weblog collections. The TSM model is quite general; it can be applied to any text collections with a mixture of topics and sentiments, thus has many potential applications, such as search result summarization, opinion tracking, and user behavior prediction.

Using frequent patterns to analyze data has been one of the fundamental approaches in many data mining applications. Research in frequent pattern mining has so far mostly focused on developing efficient algorithms to discover various kinds of frequent patterns, but little attention has been paid to the important next step—interpreting the discovered frequent patterns. Although the compression and summarization of frequent patterns has been studied in some recent work, the proposed techniques there can only annotate a frequent pattern with nonsemantical information (e.g., support), which provides only limited help for a user to understand the patterns. In this article, we study the novel problem of generating semantic annotations for frequent patterns. The goal is to discover the hidden meanings of a frequent pattern by annotating it with in-depth, concise, and structured information. We propose a general approach to generate such an annotation for a frequent pattern by constructing its context model, selecting informative context indicators, and extracting representative transactions and semantically similar patterns. This general approach can well incorporate the user's prior knowledge, and has potentially many applications, such as generating a dictionary-like description for a pattern, finding synonym patterns, discovering semantic relations, and summarizing semantic classes of a set of frequent patterns. Experiments on different datasets show that our approach is effective in generating semantic pattern annotations.

As a fundamental data mining task, frequent pattern mining has widespread applications in many different domains. Research in frequent pattern mining has so far mostly focused on developing efficient algorithms to discover various kinds of frequent patterns, but little attention has been paid to the important nextstep - interpreting the discovered frequent patterns. Although some recent work has studied the compression and summarization of frequent patterns, the proposed techniques can only annotate a frequent pattern with non-semantical information (e.g. support), which provides only limited help for a user to understand the patterns.In this paper, we propose the novel problem of generating semantic annotations for frequent patterns. The goal is to annotate a frequent pattern with in-depth, concise, and structured information that can better indicate the hidden meanings of the pattern. We propose a general approach to generate such anannotation for a frequent pattern by constructing its context model, selecting informative context indicators, and extracting representative transactions and semantically similar patterns. This general approach has potentially many applications such as generating a dictionary-like description for a pattern, finding synonym patterns, discovering semantic relations, and summarizing semantic classes of a set of frequent patterns. Experiments on different datasets show that our approach is effective in generating semantic pattern annotations.

Contextual text mining is concerned with extracting topical themes from a text collection with context information (e.g., time and location) and comparing/analyzing the variations of themes over different contexts. Since the topics covered in a document are usually related to the context of the document, analyzing topical themes within context can potentially reveal many interesting theme patterns. In this paper, we generalize some of these models proposed in the previous work and we propose a new general probabilistic model for contextual text mining that can cover several existing models as special cases. Specifically, we extend the probabilistic latent semantic analysis (PLSA) model by introducing context variables to model the context of a document. The proposed mixture model, called contextual probabilistic latent semantic analysis (CPLSA) model, can be applied to many interesting mining tasks, such as temporal text mining, spatiotemporal text mining, author-topic analysis, and cross-collection comparative analysis. Empirical experiments show that the proposed mixture model can discover themes and their contextual variations effectively.

In this paper, we study the problem of discovering interesting patterns through user’s interactive feedback. We assume a set of candidate patterns (i.e., frequent patterns) has already been mined. Our goal is to help a particular user effectively discover interesting patterns according to his specific interest. Without requiring a user to explicitly construct a prior knowledge to measure the interestingness of patterns, we learn the user’s prior knowledge from his interactive feedback. We propose two models to represent a user’s prior: the log-linear model and biased belief model. The former is designed for item-set patterns, whereas the latter is also applicable to sequential and structural patterns. To learn these models, we present a two-stage approach, progressive shrinking and clustering, to select sample patterns for feedback. The experimental results on real and synthetic data sets demonstrate the effectiveness of our approach.

Mining subtopics from weblogs and analyzing their spatiotemporal patterns have applications in multiple domains. In this paper, we define the novel problem of mining spatiotemporal theme patterns from weblogs and propose a novel probabilistic approach to model the subtopic themes and spatiotemporal theme patterns simultaneously. The proposed model discovers spatiotemporal theme patterns by (1) extracting common themes from weblogs; (2) generating theme life cycles for each given location; and (3) generating theme snapshots for each given time period. Evolution of patterns can be discovered by comparative analysis of theme life cycles and theme snapshots. Experiments on three different data sets show that the proposed approach can discover interesting spatiotemporal theme patterns effectively. The proposed probabilistic model is general and can be used for spatiotemporal text mining on any domain with time and location information.

Temporal Text Mining (TTM) is concerned with discovering temporal patterns in text information collected over time. Since most text information bears some time stamps, TTM has many applications in multiple domains, such as summarizing events in news articles and revealing research trends in scientific literature. In this paper, we study a particular TTM task -- discovering and summarizing the evolutionary patterns of themes in a text stream. We define this new text mining problem and present general probabilistic methods for solving this problem through (1) discovering latent themes from text; (2) constructing an evolution graph of themes; and (3) analyzing life cycles of themes. Evaluation of the proposed methods on two different domains (i.e., news articles and literature) shows that the proposed methods can discover interesting evolutionary theme patterns effectively.

How to personalize an e-learning system has become an interesting question. The purpose of this paper is to demonstrate an innovative idea of personalizing an e-learning system driven by knowledge flow and show the stratified architecture and the current progress made in Peking University.