Incorporating Prioritized User Preferences in Search System

Abstract

Web services are increasing daily and users are looking to find relevant services in a quick manner. Browsing irrelevant pages presented in a retrieval hitlist would consume time and effort. Hence, there is a need to reduce the search space which will then present users with a higher retrieval precision. The idea carried out by a priority based retrieval system is that, when one attributes links to another, it is basically indicating the importance of the other attribute. The higher the value, the more important the attribute is. In this work, the search requirements defined by users are accompanied by priority values. Relevant documents identified from various resources are sorted based on the priority values. Results obtained indicate that by using priority values, users of a retrieval system are better satisfied. The undertaken precision analysis shows that relevancy of the retrieved results is improved.