A study of meta-search agent based on tags and ontological approach for improving web searches

Abstract

Recent increase in interest for information ranking and sharing among users with similar tastes has urged many researches towards improving relevance of search results for reducing costs and offering better quality service to users of Web search engines. Our work has focused largely on the ranking and sharing schemes of retrieved information among heterogeneous sources, whereas Web search engines need to wide crawler with just speed in short time. In this paper, we propose a meta-search agent with a URL filter, a tag-based ranking scheme, and an ontology-based sharing scheme. The meta-search agent uses vector tags to facilitate the definition of informative value and finally the maintenance of shared information. We introduce a concept of vector tag that shows a conceptual distance between retrieval interest and search results. We also compare performance of the proposed system with hyperlink-based methodologies, and analyze the pros and cons of each. © 2008 Springer-Verlag Berlin Heidelberg.