Nima Dokoohaki

Social Trust-Aware Recommendation System: A T-Index Approach

Collaborative Filtering based on similarity suffers from a variety of problems such as sparsity and scalability. In this paper, we propose an ontological model of trust between users on a social network to address the limitations of similarity measure in Collaborative Filtering algorithms. For enhancing the constructed network of users based on trust, we introduce an estimate of a user’s trustworthiness called T-index to identify and select neighbors in an effective manner. We employ T-index to store raters of an item in a so-called TopTrustee list which provides information about users who might not be accessible within a predefined maximum path length. An empirical evaluation shows that our solution improves both prediction accuracy and coverage of recommendations collected along few edges that connect users on a social network by exploiting T-index. We also analyze effect of T-index on structure of trust network to justify the results.

Diversifying Product Review Rankings: Getting the Full Picture

E-commerce Web sites owe much of their popularity to consumer reviews provided together with product descriptions. On-line customers spend hours and hours going through heaps of textual reviews to build confidence in products they are planning to buy. At the same time, popular products have thousands of user-generated reviews. Current approaches to present them to the user or recommend an individual review for a product are based on the helpfulness or usefulness of each review. In this paper we look at the top-k reviews in a ranking to give a good summary to the user with each review complementing the others. To this end we use Latent Dirichlet Allocation to detect latent topics within reviews and make use of the assigned star rating for the product as an indicator of the polarity expressed towards the product and the latent topics within the review. We present a framework to cover different ranking strategies based on theusers need: Summarizing all reviews, focus on a particular latent topic, or focus on positive, negative or neutral aspects. We evaluated the system using manually annotated review data from a commercial review Web site.

Achieving optimal privacy in trust-aware social recommender systems

Collaborative filtering (CF) recommenders are subject to numerous shortcomings such as centralized processing, vulnerability to shilling attacks, and most important of all privacy. To overcome these obstacles, researchers proposed for utilization of interpersonal trust between users, to alleviate many of these crucial shortcomings. Till now, attention has been mainly paid to strong points about trust-aware recommenders such as alleviating profile sparsity or calculation cost efficiency, while least attention has been paid on investigating the notion of privacy surrounding the disclosure of individual ratings and most importantly protection of trust computation across social networks forming the backbone of these systems. To contribute to addressing problem of privacy in trust-aware recommenders, within this paper, first we introduce a framework for enabling privacy-preserving trust-aware recommendation generation. While trust mechanism aims at elevating recommenders accuracy, to preserve privacy, accuracy of the system needs to be decreased. Since within this context, privacy and accuracy are conflicting goals we show that a Pareto set can be found as an optimal setting for both privacy-preserving and trust-enabling mechanisms. We show that this Pareto set, when used as the configuration for measuring the accuracy of base collaborative filtering engine, yields an optimized tradeoff between conflicting goals of privacy and accuracy. We prove this concept along with applicability of our framework by experimenting with accuracy and privacy factors, and we show through experiment how such optimal set can be inferred.

Structural Determination of Ontology-Driven Trust Networks in Semantic Social Institutions and Ecosystems

Social institutions and ecosystems are growing across the web and social trust networks formed within these systems create an extraordinary test-bed to study relation dependant notions such as trust, reputation and belief. In order to capture, model and represent the semantics of trust relationships forming the trust networks, main components of relationships are represented and described using ontologies. This paper investigates how effective design of trust ontologies can improve the structure of trust networks created and implemented within semantic web-driven social institutions and systems. Based on the context of our research, we represent a trust ontology that captures the semantics of the structure of trust networks based on the context of social institutions and ecosystems on semantic web.

Diversifying customer review rankings

E-commerce Web sites owe much of their popularity to consumer reviews accompanying product descriptions. On-line customers spend hours and hours going through heaps of textual reviews to decide which products to buy. At the same time, each popular product has thousands of user-generated reviews, making it impossible for a buyer to read everything. Current approaches to display reviews to users or recommend an individual review for a product are based on the recency or helpfulness of each review. In this paper, we present a framework to rank product reviews by optimizing the coverage of the ranking with respect to sentiment or aspects, or by summarizing all reviews with the top-K reviews in the ranking. To accomplish this, we make use of the assigned star rating for a product as an indicator for a reviews sentiment polarity and compare bag-of-words (language model) with topic models (latent Dirichlet allocation) as a mean to represent aspects. Our evaluation on manually annotated review data from a commercial review Web site demonstrates the effectiveness of our approach, outperforming plain recency ranking by 30% and obtaining best results by combining language and topic model representations.

Mechanizing social trust-aware recommenders with T-index augmented trustworthiness

Social Networks have dominated growth and popularity of the Web to an extent which has never been witnessed before. Such popularity puts forward issue of trust to the participants of Social Networks. Collaborative Filtering Recommenders have been among many systems which have begun taking full advantage of Social Trust phenomena for generating more accurate predictions. For analyzing the evolution of constructed networks of trust, we utilize Collaborative Filtering enhanced with T-index as an estimate of a users trustworthiness to identify and select neighbors in an effective manner. Our empirical evaluation demonstrates how T-index improves the Trust Network structure by generating connections to more trustworthy users. We also show that exploiting T-index results in better prediction accuracy and coverage of recommendations collected along few edges that connect users on a network.

Trust and Privacy Enabled Service Composition using Social Experience

In this paper, we present a framework for automatic selection and composition of services which exploits trustworthiness of services as a metric for measuring the quality of service composition. Trustworthiness is defined in terms of service reputation extracted from user profiles. The profiles are, in particular, extracted and inferred from a social network which accumulates users past experience with corresponding services. Using our privacy inference model we, first, prune social network to hide privacy sensitive contents and, then, utilize a trust inference based algorithm to measure reputation score of each individual service, and subsequently trustworthiness of their composition.

Epidemic Trust-Based Recommender Systems

Collaborative filtering(CF) recommender systems are among the most popular approaches to solving the information overload problem in social networks by generating accurate predictions based on the ratings of similar users. Traditional CF recommenders suffer from lack of scalability while decentralized CF recommenders (DHT-based, Gossip-based etc.) have promised to alleviate this problem. Thus, in this paper we propose a decentralized approach to CF recommender systems that uses the T-Man algorithm to create and maintain an overlay network that in turn would facilitate the generation of recommendations based on local information of a node. We analyse the influence of the number of rounds and neighbors on the accuracy of prediction and item coverage and we propose a new approach to inferring trust values between a user and its neighbors. Our experiment son two datasets show an improvement of prediction accuracy relative to previous approaches while using a highly scalable, decentralized paradigm. We also analyse item coverage and show that our system is able to generate predictions for significant fraction of the users, which is comparable with the centralized approaches.

Predicting Swedish Elections with Twitter: A Case for Stochastic Link Structure Analysis

The question that whether Twitter data can be leveraged to forecast outcome of the elections has always been of great anticipation in the research community. Existing research focuses on leveraging content analysis for positivity or negativity analysis of the sentiments of opinions expressed. This is while, analysis of link structure features of social networks underlying the conversation involving politicians has been less looked. The intuition behind such study comes from the fact that density of conversations about parties along with their respective members, whether explicit or implicit, should reflect on their popularity. On the other hand, dynamism of interactions, can capture the inherent shift in popularity of accounts of politicians. Within this manuscript we present evidence of how a well-known link prediction algorithm, can reveal an authoritative structural link formation within which the popularity of the political accounts along with their neighbourhoods, shows strong correlation with the standing of electoral outcomes. As an evidence, the public time-lines of two electoral events from 2014 elections of Sweden on Twitter have been studied. By distinguishing between member and official party accounts, we report that even using a focus-crawled public dataset, structural link popularities bear strong statistical similarities with vote outcomes. In addition we report strong ranked dependence between standings of selected politicians and general election outcome, as well as for official party accounts and European election outcome.

Mining Divergent Opinion Trust Networks through Latent Dirichlet Allocation

While the focus of trust research has been mainly on defining and modeling various notions of social trust, less attention has been given to modeling opinion trust. When speaking of social trust mainly homophily (similarity) has been the most successful metric for learning trustworthy links, specially in social web applications such as collaborative filtering recommendation systems. While pure homophily such as Pearson coefficient correlation and its variations, have been favorable to finding taste distances between individuals based on their rated items, they are not necessarily useful in finding opinion distances between individuals discussing a trending topic, e.g. Arab spring. At the same time text mining techniques, such as vector-based techniques, are not capable of capturing important factors such as saliency or polarity which are possible with topical models for detecting, analyzing and suggesting aspects of people mentioning those tags or topics. Thus, in this paper we are proposing to model opinion distances using probabilistic information divergence as a metric for measuring the distances between peoples opinion contributing to a discussion in a social network. To acquire feature sets from topics discussed in a discussion we use a very successful topic modeling technique, namely Latent Dirichlet Allocation (LDA). We use the distributions resulting to model topics for generating social networks of group and individual users. Using a Twitter dataset we show that learned graphs exhibit properties of real-world like networks.