A multi agent approach to interest profiling of users

Authors:
P. H. H. Rongen;J. Schröder;F. P. M. Dignum;J. Moorman
Affiliations:
IBM Nederland - Center for Advanced Studies;IBM Nederland - Center for Advanced Studies;Institute of Information and Computing Sciences, Utrecht University;Institute of Information and Computing Sciences, Utrecht University
Venue:
CEEMAS'05 Proceedings of the 4th international Central and Eastern European conference on Multi-Agent Systems and Applications
Year:
2005

Citing 5
Cited 2

Hierarchical Text Categorization Using Neural Networks

Information Retrieval
A Taxonomy of Recommender Agents on theInternet

Artificial Intelligence Review
Ontological user profiling in recommender systems

ACM Transactions on Information Systems (TOIS)
Towards a continuous personalization experience

Dutch HCI '04 Proceedings of the conference on Dutch directions in HCI
Prometheus: a methodology for developing intelligent agents

AOSE'02 Proceedings of the 3rd international conference on Agent-oriented software engineering III

A platform for virtual museums with personalized content

Multimedia Tools and Applications
Personalization of content in virtual exhibitions

SAMT'07 Proceedings of the semantic and digital media technologies 2nd international conference on Semantic Multimedia

Quantified Score

Hi-index	0.00

Visualization

Abstract

Intelligent applications deliver personalized experiences and services to the user. This is done by creating and using a profile of the user: user profiling. Several approaches and algorithms are developed for user profiling. This paper describes a multi agent approach that allows multiple algorithms to be combined dynamically to generate a knowledge and interest profile of a user. IBM's ABLE environment was used for the implementation of the multi agent system. To test the system, the user interest profile is build on browse behavior and this profile is applied in a TV program recommender system. The results of implemented system show that multi agent systems provide an excellent platform for an extendible user profiling system that can use multiple classifiers.