Top-N recommender systems automatically recommend N items for users from huge amounts of products. Personalized Top-N recommender systems have great impact on many real world applications such as E-commerce platforms and social networks. Sometimes there is no rating information in user-item feedback matrix but only implicit purchase or browsing history, that means the user-item feedback matrix is a binary matrix, we call such feedbacks as implicit feedbacks. In our work we try to learn Top-N recommender systems with implicit feedbacks. First, we design a heterogeneous loss function to learn the model. Second, we incorporate item side information into recommender systems. We formulate a low-rank constraint minimization problem and give a closed-form solution for it. Third, we also use item side information to learn recommender systems. We use gradient descent method to learn our model. Most existing methods produce personalized top-N recommendations by minimizing a specific uniform loss such as pairwise ranking loss or pointwise recovery loss. In our first model, we propose a novel personalized Top-N recommendation approach that minimizes a combined heterogeneous loss based on linear self-recovery models. The heterogeneous loss integrates the strengths of both pairwise ranking loss and pointwise recovery loss to provide more informative recommendation predictions. We formulate the learning problem with heterogeneous loss as a constrained convex minimization problem and develop a projected stochastic gradient descent optimization algorithm to solve it. Most previous systems are only based on the user-item feedback matrix. In many applications, in addition to the user-item rating/purchase matrix, item-based side information such as product reviews, book reviews, item comments, and movie plots can be easily collected from the Internet. This abundant item-based information can be used for recommendation systems. In the second model, we propose a novel predictive collaborative filtering approach that exploits both the partially observed user-item recommendation matrix and the item-based side information to produce top-N recommender systems. The proposed approach automatically identifies the most interesting items for each user from his or her non-recommended item pool by aggregating over his or her recommended items via a low-rank coefficient matrix. Moreover, it also simultaneously builds linear regression models from the item-based side information such as item reviews to predict the item recommendation scores for the users. The proposed approach is formulated as a rank constrained joint minimization problem with integrated least squares losses, for which an efficient analytical solution can be derived. In the third model, we also propose a joint discriminative prediction model that exploits both the partially observed user-item recommendation matrix and the item-based side information to build top-N recommender systems. This joint model aggregates observed user-item recommendation activities to predict the missing/new user-item recommendation scores while simultaneously training a linear regression model to predict the user-item recommendation scores from auxiliary item features. We evaluate the proposed approach on a variety of recommendation tasks. The experimental results show that the proposed joint model is very effective for producing top-N recommendation systems. / Computer and Information Science

Recommender systems have become paramount to customize information access and reduce information overload. They serve multiple uses, ranging from suggesting products and artifacts (to consumers), to bringing people together by the connections induced by (similar) reactions to products and services. This thesis presents a graph-theoretic model that casts recommendation as a process of 'jumping connections' in a graph. In addition to emphasizing the social network aspect, this viewpoint provides a novel evaluation criterion for recommender systems. Algorithms for recommender systems are distinguished not in terms of predicted ratings of services/artifacts, but in terms of the combinations of people and artifacts that they bring together. We present an algorithmic framework drawn from random graph theory and outline an analysis for one particular form of jump called a 'hammock.' Experimental results on two datasets collected over the Internet demonstrate the validity of this approach. / Master of Science

In this thesis, a tag-based approach is proposed for predicting mashup patterns, thus deriving inspiration for potential new mashups from the community's consensus. The proposed approach applies association rule mining techniques to discover relationships between APIs and mashups based on their annotated tags. The importance of the mined relationships is advocated as a valuable source for recommending mashup candidates while mitigating common problems in recommender systems. The proposed methodology is evaluated through experimentation using a real-life dataset. Results show that the proposed mining approach achieves prediction accuracy with 60% precision and 79% recall improvement over a direct string matching approach that lacks the mining information. / Master of Science

Question and answer (Q&A) forums, as a way for seeking expertise on the Internet, have seen rapid growth in popularity in recent years. The expertise available on most such forums is voluntary, provided by individuals willing to invest their resources for no monetary remuneration. While these forums provide easy access to expertise, the expertise available is often lacking in quality and depth. Two major reasons for this are, the time investment required to participate in such forums, and the lack of a mechanism for identifying experts for specialized questions. We believe a Q&A recommender engine can ameliorate this problem significantly. The two primary contributions of this work are: a) a hierarchical Bayesian model based Q&A recommender, and b) a discussion of metrics to measure the performance of such a Q&A recommender. Two new metrics, responder load and questioner satisfaction, are suggested based on this discussion. These metrics are used to evaluate the performance of the recommender system on datasets harvested from the Yahoo! Answers website. / text

Collaboration and sharing of information are the basis of modern social web system. Users in the social web systems are establishing and joining online communities, in order to collectively share their content with a group of people having common topic of interest. Group or community activities have increased exponentially in modern social Web systems. With the explosive growth of social communities, users of social Web systems have experienced considerable difficulty with discovering communities relevant to their interests. In this study, we address the problem of recommending communities to individual users. Recommender techniques that are based solely on community affiliation, may fail to find a wide range of proper communities for users when their available data are insufficient. We regard this problem as tag-based personalized searches. Based on social tags used by members of communities, we first represent communities in a low-dimensional space, the so-called latent semantic space, by using Latent Semantic Analysis. Then, for recommending communities to a given user, we capture how each community is relevant to both user’s personal tag usage and other community members’ tagging patterns in the latent space. We specially focus on the challenging problem of recommending communities to users who have joined very few communities or having no prior community membership. Our evaluation on two heterogeneous datasets shows that our approach can significantly improve the recommendation quality.

One of the most important challenges facing us today is to personalize services based on user preferences. In order to achieve this objective, the design of Recommender Systems (RSs), which are systems designed to aid the users through different decision-making processes by providing recommendations to them, have been an active area of research. RSs may produce personalized and non-personalized recommendations. Non-personalized RSs provide general suggestions to a user, based on the number of times an item has been selected in the past. Personalized RSs, on the other hand, aim to predict the most suitable items for a specific user, based on the user’s preferences and constraints. The latter are the focus of this thesis. While Recommender Systems have been successful in many domains, a number of challenges remain. For example, most implementations consider only single criteria ratings, and consequently are unable to identify why a user prefers an item over others. Many systems classify the user into one single group or cluster which is an unrealistic approach, since in real world users share commonalities in different degrees with diverse types of users. Others require a large amount of previously gathered data about users’ interactions and preferences, in order to be successfully applied. In this study, we introduce a methodology for the creation of Personalized Multi Criteria Context Aware Recommender Systems that aims to overcome these shortcomings. Our methodology incorporates the user’s current context information, and techniques from the Multiple Criteria Decision Analysis (MCDA) field of study to analyze and model the user preferences. To this end, we create a multi criteria user preference model to assess the utility of each item for a specific user, to then recommend the items with the highest utility. The criteria considered when creating the user preference model are the user’s location, mobility level and user profile. The latter is obtained by considering the user specific needs, and generalizing the user data from a large scale demographic database. We present a case study where we applied our methodology into PeRS, a personal Recommender System to recommend events that will take place within the Ottawa/Gatineau Region. Furthermore, we conduct an offline experiment performed to evaluate our methodology, as implemented in our case study. From the experimental results we conclude that our RS is capable to accurately narrow down, and identify, the groups from a demographic database where a user may belong, and subsequently generate highly accurate recommendation lists of items that match with his/her preferences. This means that the system has the ability to understand and typify the user. Moreover, the results show that the obtained system accuracy doesn’t depend on the user profile. Therefore, the system is potentially capable to produce equally accurate recommendations for a wide range of the population.

Recommender systems are ubiquitously used by online vendors as profitable tools to boost sales and enhance the purchase experience of their consumers. In recent literature, the value created by recommender systems are discussed extensively. In contrast, few researchers look at the negative side of the recommender systems from the viewpoint of policymakers. To fill this gap, I critically investigate the welfare impact of recommender systems (RSs) during my Ph.D. study. The main focus of my Ph.D. dissertation is analyzing whether there exists a conflict of interest between the recommendations provider and its consumers in the electronic marketplace. My dissertation is composed of three parts. In Part I, I evaluate empirically whether in the real world, the profit-driven firm will choose a recommendation mechanism that hurts or is suboptimal to its consumers. In Part II, I analyze the role of personalization technology in the RSs from a unique perspective of how personalization resembles price discrimination as a profitable tool to exploit consumer surplus. In part III, I investigate the vendor’s motivation to increase the level of personalization in two-period transactions. As the RSs are designed by the firm, and the firm’s objective is to maximize profits, the RSs might not maximize consumers’ welfare. In Part I of my thesis work, I test the existence of such a conflict of interest between the firm and its consumers. I explore this question empirically with a concrete RS created by our industry collaborator for their Video-on-Demand (VoD) system. Using a large-scale dataset (300,000 users) from a randomized experiment on the VoD platform, I simulate seven RSs based on an exponential demand model with listed movie orders and prices as key inputs, estimated from the experimental dataset. The seven simulated RSs differ by the assignments of listed orders for selected recommended movies. Specifically, assignments are chosen to maximize profits, consumer surplus, social welfare, popularity (IMDB votes and IMDB ratings), and previous sales, as well as random assignments. As a result, the profit-driven recommender system generates 8% less consumer surplus than the consumer-driven RSs, providing evidence for a conflict of interest between the vendor and its consumers. Major e-vendors personalize recommendations by different algorithms that depend on how much and types of consumer information obtained. Therefore, the welfare evaluations of personalized recommendation strategies by empirical methods are hard to generalize. In Part II of my thesis, I base my analysis of personalization in RSs on a conceptual approach. Under an analytic framework of horizontal product differentiation and heterogenous consumer preferences, the resemblance of personalization to price discrimination in welfare properties is presented. Personalization is beneficial to consumers when more personalization leads to more adoption of recommendations, since it decreases search costs for more consumers. However, when the level surpasses a threshold when all consumers adopt, a more personalized RS decreases consumer surplus and only helps the firm to exploit surplus from consumers. The extreme case of perfect personalization generates the same welfare results as first-degree price discrimination where consumers get perfectly fit recommendations but are charged their willingness-to-pay. As shown in Part II, personalization is always profitable for the monopoly seller. In Part III, I investigate the vendor’s motivation to increase the level of personalization in a two-period transactions. In the first period, consumers do not observe the true quality of the recommendations and choose to accept recommended products or not based on their initial guesses. In the second period, consumers fully learn the quality. The settings of consumer uncertainty and consumer learning incentivize the firm to charge lower-than-exploiting price for recommendations to ensure consumers’ first-period adoptions of the RS. Therefore, uncertainties mediate the conflicts of interest from the vendor’s exploitive behavior even though the vendor might strategically elevate consumers’ initial evaluation to reduce such effect.

In the thesis we compare several models for prediction of user preferences. The focus is mainly on Content Based models which work with metadata about objects that are recommended. These models are compared with other models which do not use metadata for recommendation. We use three datasets and three metrics to get the results of recommendation. The goal of the thesis is to find out how can the metadata about the users and the objects enhance the standard recommender models. However, the result is that the metadata can enhance recommendation in some cases, but it varies by used metrics and dataset. This enhancement is not significant.

Sistemas de Recomendação ajudam na personalização de sites na Internet oferecendo conteúdo ou produtos específicos aos usuários. Com dispositivos móveis, aumentou o interesse do usuário em ter recomendações personalizadas de locais para ir de acordo com o seu histórico de navegação e avaliações como restaurantes e pontos turísticos. Para que as recomendações personalizadas por locais sejam mais precisas é necessário contextualizá-las de acordo com o interesse do usuário que caracterizado por locais que ele visitou e por regiões de interesse como moradia, onde trabalha ou mesmo onde passará férias. Várias técnicas de contextualização utilizaram todos os locais que o usuário visitou para geração da recomendação contextual do local, outras técnicas trabalham na arquitetura híbrida. Muitas assumem que é necessário a posição exata do usuário para que as recomendações sejam online, o que muitas vezes não é possível por limitações técnicas ou mesmo indisponibilização do usuário por questões de privacidade. O objetivo principal deste trabalho é geração de recomendações usando multicontextos de forma offline, gerando vários contextos de cada usuário. Os locais são recomendados utilizando apenas dados históricos, sem a localização exata no usuário do momento da recomendação. Para atingir este objetivo foram utilizadas técnicas de clustering para mapeamento e divisão dos contextos em regiões indicando o interesse do usuário gerando a recomendação final dos locais a partir de um método híbrido de recomendação que usa filtragem colaborativa e a recomendação contextual proposta. Os resultados mostraram que a técnica proposta apresenta recomendações melhores do que apenas a recomendação colaborativa pura e, para usuários assíduos, as recomendações são melhores do que as técnicas base usadas para comparação. / Recommender Systems help in web sites personalization, offering specific content or products to users. With mobile devices, user interest in Point-of-Interest (POI) recommendation has increased to receive recommendations about places to go according to your navigation and evaluation history in the web site. POI recommendation are improved by contextualizing according to users interest, based on places to where user has been and on regions of interest such as the region where the user lives, works or the region intends go on vacation. Many contextualization techniques use all places that user visited for generation of POI contextual recommendation. Other techniques use hybrid architectures and many of them assume that is necessary the exact point where the user is for online recommendation and this in not possible always due technical limitations or user privacy. The main objective of this work is the offline generation of recommendations using multicontexts. Places to be recommended use only user historical data, without the user current localization at the moment of recommendation. Several techniques have been used for clustering for mapping and division of contexts in regions, indicating the user interests and, finally, generating the final recommendation using a hybrid method with collaborative filtering and contextual recommendation proposed. The results indicate that the proposed technique builds better recommendations than the pure collaborative filtering technique and for heavy users the proposed technique has better results the baseline technique used for comparison.

Wang, Dingyan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (p. 97-110). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgement --- p.vi / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Overview --- p.1 / Chapter 1.2 --- Motivations --- p.6 / Chapter 1.3 --- Contributions --- p.9 / Chapter 1.4 --- Organizations of This Thesis --- p.11 / Chapter 2 --- Background --- p.14 / Chapter 2.1 --- Label Propagation Learning Framework --- p.14 / Chapter 2.1.1 --- Graph-based Semi-supervised Learning --- p.14 / Chapter 2.1.2 --- Green's Function Learning Framework --- p.16 / Chapter 2.2 --- Recommendation Methods --- p.19 / Chapter 2.2.1 --- Traditional Memory-based Methods --- p.19 / Chapter 2.2.2 --- Traditional Model-based Methods --- p.20 / Chapter 2.2.3 --- Label Propagation Recommendation Models --- p.22 / Chapter 2.2.4 --- Latent Feature Recommendation Models . --- p.24 / Chapter 2.2.5 --- Social Recommendation Models --- p.25 / Chapter 2.2.6 --- Tag-based Recommendation Models --- p.25 / Chapter 3 --- Recommendation with Latent Features --- p.28 / Chapter 3.1 --- Motivation and Contributions --- p.28 / Chapter 3.2 --- Item Graph --- p.30 / Chapter 3.2.1 --- Item Graph Definition --- p.30 / Chapter 3.2.2 --- Item Graph Construction --- p.31 / Chapter 3.3 --- Label Propagation Recommendation Model with Latent Features --- p.33 / Chapter 3.3.1 --- Latent Feature Analysis --- p.33 / Chapter 3.3.2 --- Probabilistic Matrix Factorization --- p.35 / Chapter 3.3.3 --- Similarity Consistency Between Global and Local Views (SCGL) --- p.39 / Chapter 3.3.4 --- Item-based Green's Function Recommendation Based on SCGL --- p.41 / Chapter 3.4 --- Experiments --- p.41 / Chapter 3.4.1 --- Dataset --- p.43 / Chapter 3.4.2 --- Baseline Methods --- p.43 / Chapter 3.4.3 --- Metrics --- p.45 / Chapter 3.4.4 --- Experimental Procedure --- p.45 / Chapter 3.4.5 --- Impact of Weight Parameter u --- p.46 / Chapter 3.4.6 --- Performance Comparison --- p.48 / Chapter 3.5 --- Summary --- p.50 / Chapter 4 --- Recommendation with Social Network --- p.51 / Chapter 4.1 --- Limitation and Contributions --- p.51 / Chapter 4.2 --- A Social Recommendation Framework --- p.55 / Chapter 4.2.1 --- Social Network --- p.55 / Chapter 4.2.2 --- User Graph --- p.57 / Chapter 4.2.3 --- Social-User Graph --- p.59 / Chapter 4.3 --- Experimental Analysis --- p.60 / Chapter 4.3.1 --- Dataset --- p.61 / Chapter 4.3.2 --- Metrics --- p.63 / Chapter 4.3.3 --- Experiment Setting --- p.64 / Chapter 4.3.4 --- Impact of Control Parameter u --- p.65 / Chapter 4.3.5 --- Performance Comparison --- p.67 / Chapter 4.4 --- Summary --- p.69 / Chapter 5 --- Recommendation with Tags --- p.71 / Chapter 5.1 --- Limitation and Contributions --- p.71 / Chapter 5.2 --- Tag-Based User Modeling --- p.75 / Chapter 5.2.1 --- Tag Preference --- p.75 / Chapter 5.2.2 --- Tag Relevance --- p.78 / Chapter 5.2.3 --- User Interest Similarity --- p.80 / Chapter 5.3 --- Tag-Based Label Propagation Recommendation --- p.83 / Chapter 5.4 --- Experimental Analysis --- p.84 / Chapter 5.4.1 --- Douban Dataset --- p.85 / Chapter 5.4.2 --- Experiment Setting --- p.86 / Chapter 5.4.3 --- Metrics --- p.87 / Chapter 5.4.4 --- Impact of Tag and Rating --- p.88 / Chapter 5.4.5 --- Performance Comparison --- p.90 / Chapter 5.5 --- Summary --- p.92 / Chapter 6 --- Conclusions and Future Work --- p.94 / Chapter 6.0.1 --- Conclusions --- p.94 / Chapter 6.0.2 --- Future Work --- p.96 / Bibliography --- p.97