Enhancing Recommender Systems Using Linked Open Data-Based Semantic Analysis of Items

Meymandpour, R. and Davis, J.G.

    The Linked Open Data (LOD) project is a community effort that aims to publish structured data using open and liberal licences. The LOD cloud provides free access to datasets in diverse areas such as media, geography, publications and life sciences. These datasets are publicly available for machine and human consumption using Semantic Web standards and SPARQL endpoints. In addition to facilitating interoperability and integrity across diverse platforms, this movement not only opens up unique opportunities for developing novel and innovative applications but also makes the application development more efficient and cost-effective. This paper demonstrates how LOD can be a reliable and rich source of content information that supports recommender systems to overcome problems such as the item cold-start problem and limited content analysis that restrict many of the existing systems. By building on a robust measurement of the similarities between items using LOD, we present a hybrid recommender system that combines the semantic analysis of items with collaborative filtering approaches. The experimental evaluations of our proposed method using standard benchmark data and established measures show comparable overall accuracy and significant improvement in item cold-start situations.
Cite as: Meymandpour, R. and Davis, J.G. (2015). Enhancing Recommender Systems Using Linked Open Data-Based Semantic Analysis of Items. In Proc. 3rd Australasian Web Conference (AWC 2015) Sydney, Australia. CRPIT, 166. Davis, J. G. and Bozzon, A. Eds., ACS. 11-17
pdf (from crpit.com) pdf (local if available) BibTeX EndNote GS