Hengshu Zhu

Mining Mobile User Preferences for Personalized Context-Aware Recommendation

Recent advances in mobile devices and their sensing capabilities have enabled the collection of rich contextual information and mobile device usage records through the device logs. These context-rich logs open a venue for mining the personal preferences of mobile users under varying contexts and thus enabling the development of personalized context-aware recommendation and other related services, such as mobile online advertising. In this article, we illustrate how to extract personal context-aware preferences from the context-rich device logs, or context logs for short, and exploit these identified preferences for building personalized context-aware recommender systems. A critical challenge along this line is that the context log of each individual user may not contain sufficient data for mining his or her context-aware preferences. Therefore, we propose to first learn common context-aware preferences from the context logs of many users. Then, the preference of each user can be represented as a distribution of these common context-aware preferences. Specifically, we develop two approaches for mining common context-aware preferences based on two different assumptions, namely, context-independent and context-dependent assumptions, which can fit into different application scenarios. Finally, extensive experiments on a real-world dataset show that both approaches are effective and outperform baselines with respect to mining personal context-aware preferences for mobile users.

A cost-effective recommender system for taxi drivers

The GPS technology and new forms of urban geography have changed the paradigm for mobile services. As such, the abundant availability of GPS traces has enabled new ways of doing taxi business. Indeed, recent efforts have been made on developing mobile recommender systems for taxi drivers using Taxi GPS traces. These systems can recommend a sequence of pick-up points for the purpose of maximizing the probability of identifying a customer with the shortest driving distance. However, in the real world, the income of taxi drivers is strongly correlated with the effective driving hours. In other words, it is more critical for taxi drivers to know the actual driving routes to minimize the driving time before finding a customer. To this end, in this paper, we propose to develop a cost-effective recommender system for taxi drivers. The design goal is to maximize their profits when following the recommended routes for finding passengers. Specifically, we first design a net profit objective function for evaluating the potential profits of the driving routes. Then, we develop a graph representation of road networks by mining the historical taxi GPS traces and provide a Brute-Force strategy to generate optimal driving route for recommendation. However, a critical challenge along this line is the high computational cost of the graph based approach. Therefore, we develop a novel recursion strategy based on the special form of the net profit function for searching optimal candidate routes efficiently. Particularly, instead of recommending a sequence of pick-up points and letting the driver decide how to get to those points, our recommender system is capable of providing an entire driving route, and the drivers are able to find a customer for the largest potential profit by following the recommendations. This makes our recommender system more practical and profitable than other existing recommender systems. Finally, we carry out extensive experiments on a real-world data set collected from the San Francisco Bay area and the experimental results clearly validate the effectiveness of the proposed recommender system.

Mining Personal Context-Aware Preferences for Mobile Users

In this paper, we illustrate how to extract personal context-aware preferences from the context-rich device logs (i.e., context logs) for building novel personalized context-aware recommender systems. A critical challenge along this line is that the context log of each individual user may not contain sufficient data for mining his/her context-aware preferences. Therefore, we propose to first learn common context-aware preferences from the context logs of many users. Then, the preference of each user can be represented as a distribution of these common context-aware preferences. Specifically, we develop two approaches for mining common context-aware preferences based on two different assumptions, namely, context independent and context dependent assumptions, which can fit into different application scenarios. Finally, extensive experiments on a real-world data set show that both approaches are effective and outperform baselines with respect to mining personal context-aware preferences for mobile users.

Locating targets through mention in Twitter

With the explosive development of social networks, there are excessive amount of user-generated contents available on social media platforms. Indeed, in social networks, it is now a big challenge to promote the right information to the right audiences at the right time. To this end, in this paper, we propose an integrated study of the mention mechanism in social media platforms, such as Twitter, towards locating target audiences for specific information. The study goal is to identify effective targets with high relevance and achieve high response rate as well. Along this line, we formulate the problem of locating targets when posting promotion-oriented messages as a ranking based recommendation task, and present a context-aware recommendation framework as a solution. Specifically, we first extract four categories of features, namely content, social, location and time based features, to measure the relevance among publishers, targets and promotion messages. Then, we employ Ranking Support Vector Machine (SVM) model as the solution to our ranking based recommendation problem. By introducing two bias adjustment parameters, i.e., confidence contributions of publishers and the responsiveness of targets, our framework can effectively recommend top K proper users to mention. Finally, to validate the proposed approach, we conduct extensive experiments on a real world dataset collected from Twitter. The experimental results clearly show that our approach outperforms other baselines with a significant margin.

Towards expert finding by leveraging relevant categories in authority ranking

How to improve authority ranking is a crucial research problem for expert finding. In this paper, we propose a novel framework for expert finding based on the authority information in the target category as well as the relevant categories. First, we develop a scalable method for measuring the relevancy between categories through topic models. Then, we provide a link analysis approach for ranking user authority by considering the information in both the target category and the relevant categories. Finally, the extensive experiments on two large-scale real-world Q&A data sets clearly show that the proposed method outperforms the baseline methods with a significant margin.

Ranking user authority with relevant knowledge categories for expert finding

The problem of expert finding targets on identifying experts with special skills or knowledge for some particular knowledge categories, i.e. knowledge domains, by ranking user authority. In recent years, this problem has become increasingly important with the popularity of knowledge sharing social networks. While many previous studies have examined authority ranking for expert finding, they have a focus on leveraging only the information in the target category for expert finding. It is not clear how to exploit the information in the relevant categories of a target category for improving the quality of authority ranking. To that end, in this paper, we propose an expert finding framework based on the authority information in the target category as well as the relevant categories. Along this line, we develop a scalable method for measuring the relevancies between categories through topic models, which takes consideration of both content and user interaction based category similarities. Also, we provide a topical link analysis approach, which is multiple-category-sensitive, for ranking user authority by considering the information in both the target category and the relevant categories. Finally, in terms of validation, we evaluate the proposed expert finding framework in two large-scale real-world data sets collected from two major commercial Question Answering (Q&A) web sites. The results show that the proposed method outperforms the baseline methods with a significant margin.

Real Estate Ranking via Mixed Land-use Latent Models

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

Popularity Modeling for Mobile Apps: A Sequential Approach

The popularity information in App stores, such as chart rankings, user ratings, and user reviews, provides an unprecedented opportunity to understand user experiences with mobile Apps, learn the process of adoption of mobile Apps, and thus enables better mobile App services. While the importance of popularity information is well recognized in the literature, the use of the popularity information for mobile App services is still fragmented and under-explored. To this end, in this paper, we propose a sequential approach based on hidden Markov model (HMM) for modeling the popularity information of mobile Apps toward mobile App services. Specifically, we first propose a popularity based HMM (PHMM) to model the sequences of the heterogeneous popularity observations of mobile Apps. Then, we introduce a bipartite based method to precluster the popularity observations. This can help to learn the parameters and initial values of the PHMM efficiently. Furthermore, we demonstrate that the PHMM is a general model and can be applicable for various mobile App services, such as trend based App recommendation, rating and review spam detection, and ranking fraud detection. Finally, we validate our approach on two real-world data sets collected from the Apple Appstore. Experimental results clearly validate both the effectiveness and efficiency of the proposed popularity modeling approach.

Discerning Tactical Patterns for Professional Soccer Teams: An Enhanced Topic Model with Applications

Analyzing team tactics plays an important role in the professional soccer industry. Recently, the progressing ability to track the mobility of ball and players makes it possible to accumulate extensive match logs, which open a venue for better tactical analysis. However, traditional methods for tactical analysis largely rely on the knowledge and manual labor of domain experts. To this end, in this paper we propose an unsupervised approach to automatically discerning the typical tactics, i.e., tactical patterns, of soccer teams through mining the historical match logs. To be specific, we first develop a novel model named Team Tactic Topic Model (T3M) for learning the latent tactical patterns, which can model the locations and passing relations of players simultaneously. Furthermore, we demonstrate several potential applications enabled by the proposed T3M, such as automatic tactical pattern discovery, pass segment annotation, and spatial analysis of player roles. Finally, we implement an intelligent demo system to empirically evaluate our approach based on the data collected from La Liga 2013-2014. Indeed, by visualizing the results obtained from T3M, we can successfully observe many meaningful tactical patterns and interesting discoveries, such as using which tactics a team is more likely to score a goal and how a teams playing tactic changes in sequential matches across a season.

Diversified social influence maximization

For better viral marketing, there has been a lot of research on social influence maximization. However, the problem that who is influenced and how diverse the influenced population is, which is important in real-world marketing, has largely been neglected. To that end, in this paper, we propose to consider the magnitude of influence and the diversity of the influenced crowd simultaneously. Specifically, we formulate it as an optimization problem, i.e., diversified social influence maximization. First, we present a general framework for this problem, under which we construct a class of diversity measures to quantify the diversity of the influenced crowd. Meanwhile, we prove that a simple greedy algorithm guarantees to provide a near-optimal solution to the optimization problem. Furthermore, we relax the problem by focusing on the diversity of the nodes targeted for initial activation, and show how this relaxed form could be used to diversify the results of many heuristics, e.g., PageRank. Finally, we run extensive experiments on two real-world datasets, showing that our formulation is effective in generating diverse results.