Building the search pattern of web users using conceptual semantic space model

Abstract

As the web enters Big Data age, users and search engines may find it more and more difficult to effectively use and manage such big data. This paper proposes a method based on user tasks to perceive user intent, the meaningful user intent can be separated, and the elusiveness of user intent can be solved. One more thing, the method based on user task can predict the user behaviour to some extent. The main work of this paper is to propose such a concept space to fit the above features and the method to build the concept space on a large text database. We propose a new search pattern termed as ExNa, which can be incorporated into search engines to support more efficient search with better results. We conduct experimental studies upon a prototype search engine KNOWLE, so as to evaluate the validity and performance of our proposed ExNa mechanism.