Combining Content-Based and Collaborative Filters in an Online Newspaper

The explosive growth of mailing lists, Web sites and Usenet news demands eeective ltering solutions. Collaborative l-tering combines the informed opinions of humans to make personalized, accurate predictions. Content-based ltering uses the speed of computers to make complete, fast predictions. In this work, we present a new ltering approach that combines the coverage and speed of content-lters with the depth of collaborative ltering. We apply our research approach to an online newspaper, an as yet untapped opportunity for lters useful to the widespread news reading populace. We present the design of our ltering system and describe the results from preliminary experiments that suggest merits to our approach. 1 Introduction That we are in the age of information is evident quite clearly in newspapers as an information source. Nearly everywhere in North America you can have 1/2 dozen newspapers delivered to your doorstep, each with hundreds of new articles each day. Nearly everywhere in the world via the World Wide Web, you can access more than 2500 daily newspapers through their online Web sites 27], providing tens of thousands of news articles of potential interest. We need information lters to help us prioritize news articles so that we may spend more of our time reading articles of interest. Newspaper lters present the additional opportunity for personalization, which has been shown to have a strong appeal to newspaper readers 5]. Practical considerations have prevented hardcopy newspapers from obtaining any degree of personal customization, but online newspapers are not subject to the same constraints as printed matter. Online newspaper presentation can be personalized in terms of contents , layout, media (text only, text with pictures, text with video, etc.), advertisements and more. While there have been attempts at customization of newspapers and ltering of newspapers 16, 25], these attempts have not completely countered the weaknesses of human and computer lters. Both humans and computers need help in ltering infor

[1]  Hans Peter Luhn,et al.  The Automatic Creation of Literature Abstracts , 1958, IBM J. Res. Dev..

[2]  L. Bogart.  Press and public: Who reads what, when, where, and why in American newspapers , 1981 .

[3]  Christopher J. Fox,et al.  A stop list for general text , 1989, SIGF.

[4]  Susan T. Dumais,et al.  Personalized information delivery: an analysis of information filtering methods , 1992, CACM.

[5]  Douglas B. Terry,et al.  Using collaborative filtering to weave an information tapestry , 1992, CACM.

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

[7]  Pattie Maes,et al.  Social information filtering: algorithms for automating “word of mouth” , 1995, CHI '95.

[8]  Mark Rosenstein,et al.  Recommending and evaluating choices in a virtual community of use , 1995, CHI '95.

[9]  P. R. Chesnais,et al.  The Fishwrap personalized news system , 1995, Proceedings of the Second International Workshop on Community Networking 'Integrated Multimedia Services to the Home'.

[10]  Garrison W. Cottrell,et al.  Using Relevance to Train a Linear Mixture of Experts , 1996, TREC.

[11]  Krishna Bharat,et al.  The Krakatoa Chronicle: An Interactive Personalized Newspaper on the Web , 1995, World Wide Web J..

[12]  Yoav Shoham,et al.  Content-Based, Collaborative Recommendation. , 1997 .

[13]  Joshua Alspector,et al.  Comparing feature-based and clique-based user models for movie selection , 1998, DL '98.

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

[15]  Bradley N. Miller,et al.  Using filtering agents to improve prediction quality in the GroupLens research collaborative filtering system , 1998, CSCW '98.

[16]  Ahmad M. Ahmad Wasfi.  Collecting user access patterns for building user profiles and collaborative filtering , 1998, IUI '99.

[17]  William W. Cohen,et al.  Recommendation as Classification: Using Social and Content-Based Information in Recommendation , 1998, AAAI/IAAI.

[18]  Michael J. Pazzani,et al.  Learning Collaborative Information Filters , 1998, ICML.

[19]  Anuja Gokhale,et al.  Correlation Thresholds for More Accurate Collaborative Filtering Correlation Thresholds for More Accurate Collaborative Filtering , 1999 .