Increasing Search Success Rate by Comparing Queries Formation Patterns

Abstract

Retrieving suitable information to a user's information request is a crucial task of search engines. However, most of conventional search engines that are based on pattern matching schemes tend to provide too many improper results, most of which are not corresponding to a user's request. It is due to the uncertainty o f queries. To overcome this problem, we distinguished two different patterns after extensive investigation about the process of q uery formation. One is the general information seeking pattern and the other is the scholastic information retrieving pattern. The general infor mation seeking pattern has a tendency to correspond to cross- relational query formation process pattern, and the scholastic informatio n retrieving pattern corresponds to goal-oriented structured query pattern. This paper proposes a method to improve search results by utilizing these two query formation patterns. The study on the formation of general cross-relational queries, and goal-oriented str uctured query pattern by which comparing or modeling will improve the efficiency of navigation. These proposals will benefit not only sear ch engine developers and library managers, but also search engine users and researchers. This paper is expected to provide more suitable searching results than those obtained by using current typical information retrieval engines.