AVATAR: a multi-agent TV recommender system using MHP applications

Author

Blanco-Fernández, Yolanda ; Pazos-Arias, José J. ; Gil-Solla, Alberto ; Ramos-Cabrer, Manuel ; López-Nores, Martín ; Barragáns-Martínez, Belén

Author_Institution

Dept. of Telematic Eng., Vigo Univ., Spain

fYear

2005

fDate

29 March-1 April 2005

Firstpage

660

Lastpage

665

Abstract

In this paper we present AVATAR, a recommender of personalized TV contents, for which an open multi-agent architecture is proposed. This tool uses different knowledge inference strategies, even though we focus on the description of one of them, the naive Bayesian classifiers. Standardization is a key issue to achieve wide deployment in this kind of systems. So, the described recommendation tool has been conceived as an application conforming to multimedia home platform (MHP) standard, to be distributed in the broadcast transport stream that will be tuned by each user´s receiver. Also, the TV-Anytime specification has been used referred to the description of contents, the users preferences and their activity logs.

Keywords

belief networks; content management; digital television; inference mechanisms; information filtering; multi-agent systems; multimedia communication; AVATAR; TV contents; broadcast transport stream; knowledge inference; multiagent architecture; multimedia home platform standard; naive Bayesian classifiers; recommender system; Application software; Avatars; Bayesian methods; Concrete; Digital TV; Feedback; Recommender systems; Streaming media; TV broadcasting; TV receivers;

fLanguage

English

Publisher

ieee

Conference_Titel

e-Technology, e-Commerce and e-Service, 2005. EEE '05. Proceedings. The 2005 IEEE International Conference on

Print_ISBN

0-7695-2274-2

Type

conf

DOI

10.1109/EEE.2005.39

Filename

1402376