Authors:
Nidhi Kushwaha
and
O. P. Vyas
Affiliation:
Indian Institite of Information Technology, India
Keyword(s):
Matrix Factorization, Semantic Topics, DBpedia, RDF, SPARQL, Similarity Coefficient, TF-IDF.
Related
Ontology
Subjects/Areas/Topics:
Biomedical Engineering
;
Data Engineering
;
Enterprise Information Systems
;
Health Information Systems
;
Information Systems Analysis and Specification
;
Internet Technology
;
Knowledge Management
;
Ontologies and the Semantic Web
;
Ontology and the Semantic Web
;
Recommendation Systems
;
Searching and Browsing
;
Society, e-Business and e-Government
;
Software Agents and Internet Computing
;
System Integration
;
User Modeling
;
Web Information Systems and Technologies
;
Web Interfaces and Applications
Abstract:
The Matrix Factorization model proved as a state of art technique in the field of Recommender Systems.
The latent factors in these techniques are mathematically derived factors that are useful in terms of
dimensionality reduction and sparsity removal. In this paper, we exploited the information on these latent
factors in addition with semantic knowledge fetched from the DBpedia dataset to predict the movies to
users, based on their selected topics in the past. We incorporate matrix factorization with the Semantic
information to increase the accuracy of the recommendation and also increase the contextual information
into it. For handling cold start users, we also provide an opportunity for the user, to select topics at the run
time and prediction will be made according to their selection. To improve the diversity of the prediction in
both the cases we also used a specific strategy for the end user recommendation.