Development of intelligent user behavior based adaptive rule based system

Abstract

This paper investigates attributes of user search behavior and develops ‘Search Satisfaction metrics’ to determine user satisfaction with Search Engine. Unlike most previous web studies that have analyzed user behavior through Search Engine logs, this work focuses primarily on remotely observing user as he queries Search Engine. A survey was conducted to identify pattern of search among large sample of users. User generally queries Search Engine with different types of queries namely Informational, Transactional and Navigational. Our analysis suggests that the user search pattern differs for various types of queries. User tendency to click even when search is successful has been investigated. This has been observed to be a common behavior for open ended informational and transactional queries. A rule based model has been proposed to determine the user satisfaction for a search session. The application of the system in determining user satisfaction, search results relevancy and query complexity has been illustrated.