Extending Recommender Systems for Disjoint User/Item Sets: The Conference Recommendation Problem

In this paper, we describe the problem of recommending conference sessions to attendees and show how novel extensions to traditional model-based recommender systems, as suggested in Adomavicius and Tuzhilin [CHECK END OF SENTENCE], can address this problem. We introduce Recommendation Engine by CONjoint Decomposition of ITems and USers (RECONDITUS)-a technique that is an extension of preference-based recommender systems to recommend items from a new disjoint set to users from a new disjoint set. The assumption being that preferences exhibited by users with known usage behavior (e.g., past conference session attendance), which can be abstracted by projections of user and item matrices, will be similar to ones of new (different) users where the basic environment and item domain are the same (e.g., new conference). RECONDITUS requires no item ratings, but operates on observed user behavior such as past conference session attendance. The RECONDITUS paradigm consists of projections of both user and item data, and the learning of relationships in projected space. Once established, the relationships enable predicting new relationships and provide associated recommendations. The approach can encompass several traditional data mining problems where both clustering and prediction are necessary. RECONDITUS has been evaluated using data from the Oracle OpenWorld conference.

[1]  Luo Si,et al.  Preference-based Graphic Models for Collaborative Filtering , 2002, UAI.

[2]  Greg Linden,et al.  Amazon . com Recommendations Item-to-Item Collaborative Filtering , 2001 .

[3]  John Riedl,et al.  An algorithmic framework for performing collaborative filtering , 1999, SIGIR '99.

[4]  H. Sebastian Seung,et al.  Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[5]  Michael W. Berry,et al.  Survey of Text Mining: Clustering, Classification, and Retrieval , 2007 .

[6]  Pablo Tamayo,et al.  Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[7]  Paul Resnick,et al.  Recommender systems , 1997, CACM.

[8]  Gediminas Adomavicius,et al.  Incorporating contextual information in recommender systems using a multidimensional approach , 2005, TOIS.

[9]  艾文.  到行业中去——Oracle OpenWorld后记 , 2004 .

[10]  H. Sebastian Seung,et al.  Algorithms for Non-negative Matrix Factorization , 2000, NIPS.

[11]  David Heckerman,et al.  Empirical Analysis of Predictive Algorithms for Collaborative Filtering , 1998, UAI.

[12]  Kurt Hornik,et al.  Text Mining Infrastructure in R , 2008 .

[13]  George Karypis,et al.  Item-based top-N recommendation algorithms , 2004, TOIS.

[14]  Harald Bergstriim.  Mathematical Theory of Probability and Statistics , 1966 .

[15]  Paul Resnick,et al.  Recommender Systems - Introduction to the Special Section , 1997, Commun. ACM.

[16]  John Riedl,et al.  Application of Dimensionality Reduction in Recommender System - A Case Study , 2000 .

[17]  Luo Si,et al.  Collaborative filtering with decoupled models for preferences and ratings , 2003, CIKM '03.

[18]  Gediminas Adomavicius,et al.  Context-aware recommender systems , 2008, RecSys '08.

[19]  David Maxwell Chickering,et al.  Dependency Networks for Collaborative Filtering and Data Visualization , 2000, UAI.

[20]  John Riedl,et al.  Item-based collaborative filtering recommendation algorithms , 2001, WWW '01.

[21]  Adi Ben-Israel,et al.  Generalized inverses: theory and applications , 1974 .

[22]  Ole Winther,et al.  Bayesian Non-negative Matrix Factorization , 2009, ICA.

[23]  Gediminas Adomavicius,et al.  Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions , 2005, IEEE Transactions on Knowledge and Data Engineering.

[24]  Thomas Hofmann,et al.  Latent semantic models for collaborative filtering , 2004, TOIS.

[25]  John Riedl,et al.  GroupLens: an open architecture for collaborative filtering of netnews , 1994, CSCW '94.