Distributed information retrieval based on metaheuristic search and query expansion

Abstract EN

Distributed information retrieval (DIR) is a model enables a user to access many searchable databases reside in different locations.

DIR is more complex than the centralized information retrieval (IR).

It requires addressing two significant additional problems that are the resource selection and the results merging.

Many techniques for addressing the two problems have been published in the literature.

However, they still have a negative impact on retrieving quality and response time.

This paper aims to improve the DIR efficiency through using a meta-heuristic algorithm and improving the result quality through a query expansion.

The algorithm has been strengthened using the nearest neighbor graph in order to improve the search performance.

The performance in the proposed system outperforms the one in the traditional system in a rate from 6% to 9% while reduces the latency in an approximate rate from 0.047 to 0.134 second for each query.