Search

Collaborative filtering aims at predicting a user's interest for a given item based on a collection of user profiles. This article views collaborative filtering as a problem highly related to information retrieval, drawing an analogy between the concepts of users and items in recommender systems and queries and documents in text retrieval. We present a probabilistic user-to-item relevance framework that introduces the concept of relevance into the related problem of collaborative filtering. Three different models are derived, namely, a user-based, an item-based, and a unified relevance model, and we estimate their rating predictions from three sources: the user's own ratings for different items, other users' ratings for the same item, and ratings from different but similar users for other but similar items. To reduce the data sparsity encountered when estimating the probability density function of the relevance variable, we apply the nonparametric (data-driven) density estimation technique known as the Parzen-window method (or kernel-based density estimation). Using a Gaussian window function, the similarity between users and/or items would, however, be based on Euclidean distance. Because the collaborative filtering literature has reported improved prediction accuracy when using cosine similarity, we generalize the Parzen-window method by introducing a projection kernel. Existing user-based and item-based approaches correspond to two simplified instantiations of our framework. User-based and item-based collaborative filterings represent only a partial view of the prediction problem, where the unified relevance model brings these partial views together under the same umbrella. Experimental results complement the theoretical insights with improved recommendation accuracy. The unified model is more robust to data sparsity because the different types of ratings are used in concert. [ABSTRACT FROM AUTHOR]

Collaborative filtering aims at predicting a user's interest for a given item based on a collection of user profiles. This article views collaborative filtering as a problem highly related to information retrieval, drawing an analogy between the concepts of users and items in recommender systems and queries and documents in text retrieval. We present a probabilistic user-to-item relevance framework that introduces the concept of relevance into the related problem of collaborative filtering. Three different models are derived, namely, a user-based, an item-based, and a unified relevance model, and we estimate their rating predictions from three sources: the user's own ratings for different items, other users' ratings for the same item, and ratings from different but similar users for other but similar items. To reduce the data sparsity encountered when estimating the probability density function of the relevance variable, we apply the nonparametric (data-driven) density estimation technique known as the Parzen-window method (or kernel-based density estimation). Using a Gaussian window function, the similarity between users and/or items would, however, be based on Euclidean distance. Because the collaborative filtering literature has reported improved prediction accuracy when using cosine similarity, we generalize the Parzen-window method by introducing a projection kernel. Existing user-based and item-based approaches correspond to two simplified instantiations of our framework. User-based and item-based collaborative filterings represent only a partial view of the prediction problem, where the unified relevance model brings these partial views together under the same umbrella. Experimental results complement the theoretical insights with improved recommendation accuracy. The unified model is more robust to data sparsity because the different types of ratings are used in concert. [ABSTRACT FROM AUTHOR]

Collaborative filtering (CF) is a personalization technology that generates recommendations for users based on others' evaluations. CF is used by numerous e-commerce Web sites for providing personalized recommendations. Although much research has focused on refining collaborative filtering algorithms, little is known about the effects of user and domain characteristics on the accuracy of collaborative filtering systems. In this study, the effects of two factors-product domain and users' search mode-on the accuracy of CF are investigated. The effects of those factors are tested using data collected from two experiments in two different product domains, and from two large CF datasets, EachMovie and Book-Crossing. The study shows that the search mode of the users strongly influences the accuracy of the recommendations. CF works better when users look for specific information than when they search for general information. The accuracy drops significantly when data from different modes are mixed. The study also shows that CF is more accurate for knowledge domains than for consumer product domains. The results of this study imply that for more accurate recommendations, collaborative filtering systems should be able to identify and handle users' mode of search, even within the same domain and user group. [ABSTRACT FROM AUTHOR]