Incorporating Link Information from Linked Open Data for Movie Recommendation

Tracking #: 1741-2953

This paper is currently under review
Hsin-Chang Yang
Chia-Chi Hsu

Responsible editor: 
Philippe Cudre-Mauroux

Submission type: 
Full Paper
Schemes for automatic recommendation of resources, such as movies, hotels, restaurants, songs, etc., have emerged recently and received great attention since the volume of resources increases drastically which prevents users from obtaining their best choices. Most of these schemes rely on analysis of materials regarding the resources, such as user reviews and scores. In this work, we propose a recommendation scheme based on information retrieving from linked open data, which contains enormous amount of publicly accessible data with semantic interlinks and serves as a universal repository of knowledge. We applied the scheme on the task of movie recommendation using DBpedia as the knowledge source. Experimental result demonstrates that users were satisfied on the automatically recommended movies.
Full PDF Version: 
Under Review