A method for weighting multi-valued features in content-based filtering

Authors:
Manuel J. Barranco;Luis Martínez
Affiliations:
University of Jaén, Jaén, Spain;University of Jaén, Jaén, Spain
Venue:
IEA/AIE'10 Proceedings of the 23rd international conference on Industrial engineering and other applications of applied intelligent systems - Volume Part III
Year:
2010

Citing 10
Cited 0

Elements of information theory

Elements of information theory
Content-based book recommending using learning for text categorization

DL '00 Proceedings of the fifth ACM conference on Digital libraries
An information-theoretic perspective of tf—idf measures

Information Processing and Management: an International Journal
A formal study of information retrieval heuristics

Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval
Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions

IEEE Transactions on Knowledge and Data Engineering
A multigranular linguistic content-based recommendation model: Research Articles

International Journal of Intelligent Systems
Comparing and evaluating information retrieval algorithms for news recommendation

Proceedings of the 2007 ACM conference on Recommender systems
Interpreting TF-IDF term weights as making relevance decisions

ACM Transactions on Information Systems (TOIS)
Feature-Weighted User Model for Recommender Systems

UM '07 Proceedings of the 11th international conference on User Modeling
Content-based recommendation systems

The adaptive web

Quantified Score

Hi-index	0.00

Visualization

Abstract

Content-based recommender systems (CBRS) and collaborative filtering are the type of recommender systems most spread in the e-commerce arena. A CBRS works with two sets of information: (i) a set of features that describe the items to be recommended and (ii) a user's profile built from past choices that the user made over a subset of items. Based on these sets and on weighting items features the CBRS is able to recommend those items that better fits the user profile. Commonly, a CBRS deals with simple item features such as key words extracted from the item description applying a simple feature weighting model, based on the TF-IDF. However, this method does not obtain good results when features are assessed in multiple values and or domains. In this contribution we propose a higher level feature weighting method based on entropy and coefficients of correlation and contingency in order to improve the content-based filtering in settings with multi-valued features.