An Improved Ontology-Based User Interest Model

In personalized information retrieval, the design of a user interest model is a key problem. Through analyzing the ontology-based User Interest Model, this paper proposes a new hybrid model that is both long-term and short-term. The long-term model is updated from the Vector Space Model by a transfo...

Full description

Saved in:
Bibliographic Details
Published inModern applied science Vol. 6; no. 6; p. 39
Main Authors Liang, Zhu, Jun, Yan, Haifeng, Ling, Haibo, Qian
Format Journal Article
LanguageEnglish
Published 01.06.2012
Subjects
Algorithms
Cold starts
Information retrieval
Personalized
Transforms
Vector spaces
Online AccessGet full text

Cover

More Information
Summary:In personalized information retrieval, the design of a user interest model is a key problem. Through analyzing the ontology-based User Interest Model, this paper proposes a new hybrid model that is both long-term and short-term. The long-term model is updated from the Vector Space Model by a transform algorithm. Experiments showed that the new model tracked a user's interests more accurately, and greatly avoided the Cold Start problem.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:1913-1844
1913-1852
DOI:10.5539/mas.v6n6p39