This paper presents an overview of the field of recommender systems and describes the current generation of recommendation methods that are usually classified into the following three main categories: content-based, collaborative, and hybrid recommendation approaches. This paper also describes various limitations of current recommendation methods and discusses possible extensions that can improve recommendation capabilities and make recommender systems applicable to an even broader range of applications. These extensions include, among others, an improvement of understanding of users and items, incorporation of the contextual information into the recommendation process, support for multcriteria ratings, and a provision of more flexible and less intrusive types of recommendations.

© ACM, 2004. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in ACM

Collaborative filtering aims at learning predictive models of user preferences, interests or behavior from community data, that is, a database of available user preferences. In this article, we describe a new family of model-based algorithms designed for this task. These algorithms rely on a statistical modelling technique that introduces latent class variables in a mixture model setting to discover user communities and prototypical interest profiles. We investigate several variations to deal with discrete and continuous response variables as well as with different objective functions. The main advantages of this technique over standard memory-based methods are higher accuracy, constant time prediction, and an explicit and compact model representation. The latter can also be used to mine for user communitites. The experimental evaluation shows that substantial improvements in accucracy over existing methods and published results can be obtained.

Recommender systems provide users with personalized suggestions for products or services. These systems often rely on Collaborating Filtering (CF), where past transactions are analyzed in order to establish connections between users and products. The two more successful approaches to CF are latent factor models, which directly profile both users and products, and neighborhood models, which analyze similarities between products or users. In this work we introduce some innovations to both approaches. The factor and neighborhood models can now be smoothly merged, thereby building a more accurate combined model. Further accuracy improvements are achieved by extending the models to exploit both explicit and implicit feedback by the users. The methods are tested on the Netflix data. Results are better than those previously published on that dataset. In addition, we suggest a new evaluation metric, which highlights the differences among methods, based on their performance at a top-K recommendation task.

We have developed a method for recommending items that combines content and collaborative data under a single probabilistic framework. We benchmark our algorithm against a nave Bayes classifier on the cold-start problem, where we wish to recommend items that no one in the community has yet rated. We systematically explore three testing methodologies using a publicly available data set, and explain how these methods apply to specific real-world applications. We advocate heuristic recommenders when benchmarking to give competent baseline performance. We introduce a new performance metric, the CROC curve, and demonstrate empirically that the various components of our testing strategy combine to obtain deeper understanding of the performance characteristics of recommender systems. Though the emphasis of our testing is on cold-start recommending, our methods for recommending and evaluation are general.

In this work we present topic diversification, a novel method designed to balance and diversify personalized recommendation lists in order to reflect the user’s complete spectrum of interests. Though being detrimental to average accuracy, we show that our method improves user satisfaction with recommendation lists, in particular for lists generated using the common item-based collaborative filtering algorithm. Our work builds upon prior research on recommender systems, looking at properties of recommendation lists as entities in their own right rather than specifically focusing on the accuracy of individual recommendations. We introduce the intra-list similarity metric to assess the topical diversity of recommendation lists and the topic diversification approach for decreasing the intra-list similarity. We evaluate our method using book recommendation data, including offline analysis on 361, 349 ratings and an online study involving more than 2, 100 subjects.

Collaborative filtering is valuable in e-commerce, and for direct recommendations for music, movies, news etc. But today’s systems use centralized databases and have several disadvantages, including privacy risks. As we move toward ubiquitous computing, there is a great potential for individuals to share all kinds of information about places and things to do, see and buy, but the privacy risks are severe. In this paper we introduce a peer-to-peer protocol for collaborative filtering which protects the privacy of individual data. A second contribution of this paper is a new collaborative filtering algorithm based on factor analysis which appears to be the most accurate method for CF to date. The new algorithm has other advantages in speed and storage over previous algorithms. It is based on a careful probabilistic model of user choice, and on a probabilistically sound approach to dealing with missing data. Our experiments on several test datasets show that the algorithm is more accurate than previously reported methods, and the improvements increase with the sparseness of the dataset. Finally, factor analysis with privacy is applicable to other kinds of statistical analyses of survey or questionaire data scientists (e.g. web surveys or questionaires).

Recommender systems leverage product and community information to target products to consumers. Researchers have developed collaborative recommenders, content-based recommenders, and a few hybrid systems. We propose a unified probabilistic framework for merging collaborative and content-based recommendations. We extend Hofmann’s (1999) aspect model to incorporate three-way co-occurrence data among users, items, and item content. The relative influence of collaboration data versus content data is not imposed as an exogenous parameter, but rather emerges naturally from the given data sources. However, global probabilistic models coupled with standard EM learning algorithms tend to drastically overfit in the sparsedata situations typical of recommendation applications. We show that secondary content information can often be used to overcome sparsity. Experiments on data from the ResearchIndex library of Computer Science publications show that appropriate mixture models incorporating secondary data produce significantly better quality recommenders than-nearest neighbors (-NN). Global probabilistic models also allow more general inferences than local methods like-NN. 1

The paper presents a multidimensional (MD) approach to recommender systems that can provide recommendations based on additional contextual information besides the typical information on users and items used in most of the current recommender systems. This approach supports multiple dimensions, extensive profiling, and hierarchical aggregation of recommendations. The paper also presents a multidimensional rating estimation method capable of selecting two-dimensional segments of ratings pertinent to the recommendation context and applying standard collaborative filtering or other traditional two-dimensional rating estimation techniques to these segments. A comparison of the multidimensional and two-dimensional rating estimation approaches is made, and the tradeoffs between the two are studied. Moreover, the paper introduces a combined rating estimation method that identifies the situations where the MD approach outperforms the standard two-dimensional approach and uses the MD approach in those situations and the standard two-dimensional approach elsewhere. Finally, the paper presents a pilot empirical study of the combined approach, using a multidimensional movie recommender system that was developed for implementing this approach and testing its performance. 1 1.

Server-based collaborative filtering systems have been very successful in e-commerce and in direct recommendation applications. In future, they have many potential applications in ubiquitous computing settings. But today's schemes have problems such as loss of privacy, favoring retail monopolies, and with hampering diffusion of innovations. We propose an alternative model in which users control all of their log data. We describe an algorithm whereby a community of users can compute a public "aggregate" of their data that does not expose individual users' data. The aggregate allows personalized recommendations to be computed by members of the community, or by outsiders. The numerical algorithm is fast, robust and accurate. Our method reduces the collaborative filtering task to an iterative calculation of the aggregate requiring only addition of vectors of user data. Then we use homomorphic encryption to allow sums of encrypted vectors to be computed and decrypted without exposing individual data. We give verification schemes for all parties in the computation. Our system can be implemented with untrusted servers, or with additional infrastructure, as a fully peer-to-peer (P2P) system. 1