Smriti Bhagat

Node Classification in Social Networks

When dealing with large graphs, such as those that arise in the context of online social networks, a subset of nodes may be labeled. These labels can indicate demographic values, interest, beliefs or other characteristics of the nodes (users). A core problem is to use this information to extend the labeling so that all nodes are assigned a label (or labels).

BlurMe: inferring and obfuscating user gender based on ratings

User demographics, such as age, gender and ethnicity, are routinely used for targeting content and advertising products to users. Similarly, recommender systems utilize user demographics for personalizing recommendations and overcoming the cold-start problem. Often, privacy-concerned users do not provide these details in their online profiles. In this work, we show that a recommender system can infer the gender of a user with high accuracy, based solely on the ratings provided by users (without additional metadata), and a relatively small number of users who share their demographics. Focusing on gender, we design techniques for effectively adding ratings to a users profile for obfuscating the users gender, while having an insignificant effect on the recommendations provided to that user.

On social event organization

Online platforms, such as Meetup and Plancast, have recently become popular for planning gatherings and event organization. However, there is a surprising lack of studies on how to effectively and efficiently organize social events for a large group of people through such platforms. In this paper, we study the key computational problem involved in organization of social events, to our best knowledge, for the first time. We propose the Social Event Organization (SEO) problem as one of assigning a set of events for a group of users to attend, where the users are socially connected with each other and have innate levels of interest in those events. As a first step toward Social Event Organization, we introduce a formal definition of a restricted version of the problem and show that it is NP-hard and is hard to approximate. We propose efficient heuristic algorithms that improve upon simple greedy algorithms by incorporating the notion of phantom events and by using look-ahead estimation. Using synthetic datasets and three real datasets including those from the platforms Meetup and Plancast, we experimentally demonstrate that our greedy heuristics are scalable and furthermore outperform the baseline algorithms significantly in terms of achieving superior social welfare.

Privacy in dynamic social networks

Anonymization of social networks before they are published or shared has become an important research question. Recent work on anonymizing social networks has looked at privacy preserving techniques for publishing a single instance of the network. However, social networks evolve and a single instance is inadequate for analyzing the evolution of the social network or for performing any longitudinal data analysis. We study the problem of repeatedly publishing social network data as the network evolves, while preserving privacy of users. Publishing multiple instances of the same network independently has privacy risks, since stitching the information together may allow an adversary to identify users in the networks. We propose methods to anonymize a dynamic network such that the privacy of users is preserved when new nodes and edges are added to the published network. These methods make use of link prediction algorithms to model the evolution of the social network. Using this predicted graph to perform group-based anonymization, the loss in privacy caused by new edges can be reduced. We evaluate the privacy loss on publishing multiple social network instances using our methods.

Applying Link-Based Classification to Label Blogs

In analyzing data from social and communication networks, we encounter the problem of classifying objects where there is explicit link structure amongst the objects. We study the problem of inferring the classification of all the objects from a labeled subset, using only link-based information between objects. We abstract the above as a labeling problem on multigraphs with weighted edges. We present two classes of algorithms, based on local and global similarities. Then we focus on multigraphs induced by blog data, and carefully apply our general algorithms to specifically infer labels such as age, gender and location associated with the blog based only on the link-structure amongst them. We perform a comprehensive set of experiments with real, large-scale blog data sets and show that significant accuracy is possible from little or no non-link information, and our methods scale to millions of nodes and edges.

Recommending with an agenda: active learning of private attributes using matrix factorization

Recommender systems leverage user demographic information, such as age, gender, etc., to personalize recommendations and better place their targeted ads. Oftentimes, users do not volunteer this information due to privacy concerns, or due to a lack of initiative in filling out their online profiles. We illustrate a new threat in which a recommender learns private attributes of users who do not voluntarily disclose them. We design both passive and active attacks that solicit ratings for strategically selected items, and could thus be used by a recommender system to pursue this hidden agenda. Our methods are based on a novel usage of Bayesian matrix factorization in an active learning setting. Evaluations on multiple datasets illustrate that such attacks are indeed feasible and use significantly fewer rated items than static inference methods. Importantly, they succeed without sacrificing the quality of recommendations to users.

Applying link-based classification to label blogs

In analyzing data from social and communication networks, we encounter the problem of classifying objects where there is an explicit link structure amongst the objects. We study the problem of inferring the classification of all the objects from a labeled subset, using only the link-based information amongst the objects. We abstract the above as a labeling problem on multigraphs with weighted edges. We present two classes of algorithms, based on local and global similarities. Then we focus on multigraphs induced by blog data, and carefully apply our general algorithms to specifically infer labels such as age, gender and location associated with the blog based only on the link-structure amongst them. We perform a comprehensive set of experiments with real, large-scale blog data sets and show that significant accuracy is possible from little or no non-link information, and our methods scale to millions of nodes and edges.

Optimal recommendations under attraction, aversion, and social influence

Peoples interests are dynamically evolving, often affected by external factors such as trends promoted by the media or adopted by their friends. In this work, we model interest evolution through dynamic interest cascades: we consider a scenario where a users interests may be affected by (a) the interests of other users in her social circle, as well as (b) suggestions she receives from a recommender system. In the latter case, we model user reactions through either attraction or aversion towards past suggestions. We study this interest evolution process, and the utility accrued by recommendations, as a function of the systems recommendation strategy. We show that, in steady state, the optimal strategy can be computed as the solution of a semi-definite program (SDP). Using datasets of user ratings, we provide evidence for the existence of aversion and attraction in real-life data, and show that our optimal strategy can lead to significantly improved recommendations over systems that ignore aversion and attraction.

Privacy tradeoffs in predictive analytics

Online services routinely mine user data to predict user preferences, make recommendations, and place targeted ads. Recent research has demonstrated that several private user attributes (such as political affiliation, sexual orientation, and gender) can be inferred from such data. Can a privacy-conscious user benefit from personalization while simultaneously protecting her private attributes? We study this question in the context of a rating prediction service based on matrix factorization. We construct a protocol of interactions between the service and users that has remarkable optimality properties: it is privacy-preserving, in that no inference algorithm can succeed in inferring a users private attribute with a probability better than random guessing; it has maximal accuracy, in that no other privacy-preserving protocol improves rating prediction; and, finally, it involves a minimal disclosure, as the prediction accuracy strictly decreases when the service reveals less information. We extensively evaluate our protocol using several rating datasets, demonstrating that it successfully blocks the inference of gender, age and political affiliation, while incurring less than 5% decrease in the accuracy of rating prediction.

PNP: Fast Path Ensemble Method for Movie Design

How can we design a product or movie that will attract, for example, the interest of Pennsylvania adolescents or liberal newspaper critics? What should be the genre of that movie and who should be in the cast? In this work, we seek to identify how we can design new movies with features tailored to a specific user population. We formulate the movie design as an optimization problem over the inference of user-feature scores and selection of the features that maximize the number of attracted users. Our approach, PNP, is based on a heterogeneous, tripartite graph of users, movies, and features (e.g. actors, directors, genres), where users rate movies and features contribute to movies. We learn the preferences by leveraging user similarities defined through different types of relations, and show that our method outperforms state-of-the-art approaches, including matrix factorization and other heterogeneous graph-based analysis. We evaluate PNP on publicly available real-world data and show that it is highly scalable and effectively provides movie designs oriented towards different groups of users, including men, women, and adolescents.