Evaluation of Full Text Search Retrieval System

Abstract

With a number of search engines on the web and each with different indexing and ranking methods and different coverage, finding the one that gives the best results for a query becomes a bit challenging. The main problem however, that existing Search engines have to deal with is how to avoid irrelevant information and to retrieve the relevant ones. This current work presents a new approach for retrieving relevant information on the Web, by adopting breadth-First search algorithm. The implementation result of the retrieval system was analysed using recall and precision model for three departments at Elizade University. By learning from users’ behaviour, the approach can return very high quality search results, with a strongly reduced computing load.