TV3P: An adaptive assistant for personalized TV

Zhiwen Yu, Xingshe Zhou

Research output: Contribution to journalArticlepeer-review

75 Scopus citations

Abstract

The abundance of DTV (Digital Television) programs precipitates a need for new tools to help people personalize interesting TV content. We developed an adaptive assistant: TV3P (TV Program Personalization for PDR), which observes users' viewing behaviors in the background, updates users' profiles continuously and autonomously, and then filters and recommends programs for different users according to their respective preference information. The novel aspect of this system is the evaluation of how much time and effort it takes the system to learn new preferences once it already is biased by old preferences. This has not been proposed in any other recommender systems before. It was also proved to match real world users whose preferences can change over time. Another attractive aspect of TV3P is its employing an implicit and explicit profiling scheme.

Original languageEnglish
Pages (from-to)393-399
Number of pages7
JournalIEEE Transactions on Consumer Electronics
Volume50
Issue number1
DOIs
StatePublished - Feb 2004

Keywords

  • Personalization
  • TV3P
  • User profiling
  • Vector space model

Fingerprint

Dive into the research topics of 'TV3P: An adaptive assistant for personalized TV'. Together they form a unique fingerprint.

Cite this