A Comparison of Subject-Based Classification Strategies for Enhanced Usability

Abstract

Collections of documents, such as technical notes, are often classified on websites using a set of keywords that describe general subject areas. These keywords are used as a set of links to help the user navigate to the information he or she is seeking. This subject-based classification is intended to help users find information. A poorly designed subject-based index can make it difficult or impossible for users to find the information they need, however, creating an effective subject-based index is not easy. There are many different classification strategies that can be used. Each of these strategies can result in a different set of key words/key phrases. Some of these strategies may result in an index that is more usable than others. This paper addresses the problem of how to create a subject index that facilitates usability. Several strategies are compared against user search data. The five strategies resulted in differences in search success rate and efficiency.