LEARNING FROM ‘TAG CLOUDS’ - A Novel Approach to Build Datasets for Memory-based Reasoning Classification of Relevant Blog Articles

Ahmad Ammari, Valentina Zharkova

2010

Abstract

The advent of the Social Web has created massive online media through turning the former information consumers to present information producers. The best example is the blogosphere. Blog websites are a collection of articles written by millions of blog writers to millions of blog readers. Blogging has become a very popular means for Web 2.0 users to communicate, express, share, collaborate, and debate through their blog posts. However, as a consequence to the very massive number of blogs as well as the so diverse topics of blog posts available on the Web, most blog search engines encounter the serious challenge of finding the blog articles that are truly relevant to the certain topic that blog readers may look for. To help handling this problem, an intelligent approach to blog post search that takes advantage from the concept of ‘tag clouds’ and leverages many open source libraries, has been proposed. A Memory-Based Reasoning model has been built using SAS Enterprise Miner to assess the approach effectiveness. Results are very encouraging as retrieval precision has indicated a significant improvement in retrieving relevant posts to the user compared with traditional means of blog post retrieval.

Paper Citation

in Harvard Style

Ammari A. and Zharkova V. (2010). LEARNING FROM ‘TAG CLOUDS’ - A Novel Approach to Build Datasets for Memory-based Reasoning Classification of Relevant Blog Articles . In Proceedings of the 6th International Conference on Web Information Systems and Technology - Volume 1: WEBIST, ISBN 978-989-674-025-2, pages 325-331. DOI: 10.5220/0002884003250331

in Bibtex Style

@conference{webist10,
author={Ahmad Ammari and Valentina Zharkova},
title={LEARNING FROM ‘TAG CLOUDS’ - A Novel Approach to Build Datasets for Memory-based Reasoning Classification of Relevant Blog Articles},
booktitle={Proceedings of the 6th International Conference on Web Information Systems and Technology - Volume 1: WEBIST,},
year={2010},
pages={325-331},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002884003250331},
isbn={978-989-674-025-2},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 6th International Conference on Web Information Systems and Technology - Volume 1: WEBIST,
TI - LEARNING FROM ‘TAG CLOUDS’ - A Novel Approach to Build Datasets for Memory-based Reasoning Classification of Relevant Blog Articles
SN - 978-989-674-025-2
AU - Ammari A.
AU - Zharkova V.
PY - 2010
SP - 325
EP - 331
DO - 10.5220/0002884003250331