Title: Topic ontology-based efficient tag recommendation approach for blogs

Authors: V. Subramaniyaswamy; S. Chenthur Pandian

Addresses: Department of Computer Science and Engineering, School of Computing, SASTRA University, Thanjavur – 613401, Tamil Nadu, India ' Dr. Mahalingam College of Engineering and Technology, Pollachi – 642003, Tamil Nadu, India

Abstract: Efficient tag recommendation systems are required to help users in the task of searching, indexing and browsing appropriate blog content. Tag generation has become more popular to annotate web content, other blogs, photos, videos and music. Tag recommendation is an action of signifying valuable and informative tags to a budding item based on the content. We propose a novel approach based on topic ontology for tag recommendation. The proposed approach intelligently generates tag suggestions to blogs. In this paper, we effectively construct the technology entitled Ontology based on Wikipedia categories and WordNet semantic relationship to make the ontology more meaningful and reliable. Spreading activation algorithm is applied to assign interest scores to existing blog content and tags. High quality tags are suggested based on the significance of the interest score. Evaluation proves that the applicability of topic ontology with spreading activation algorithm helps tag recommendation more effective when compared to collaborative tag recommendations. Our proposed approach offers several solutions to tag spamming, sentiment analysis and popularity. Finally, we report the results of an experiment which improves the performance of tag recommendation approach.

Keywords: topic ontology; Wikipedia; WordNet; spreading activation; tag recommendation; spam reduction; sentiment analysis; tag popularity; blogs; weblogs; semantic relations; blog content.

DOI: 10.1504/IJCSE.2014.060682

International Journal of Computational Science and Engineering, 2014 Vol.9 No.3, pp.177 - 187

Received: 17 Jan 2012
Accepted: 03 May 2012

Published online: 24 May 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article