Query Expansion of Document Retrieval with Asking-Back Function Usilizing Tag Information

Abstract

This paper describes an asking-back function for document retrieval. When a user searches for documents in an unfamiliar field, he/she may not know the appropriate keywords or tags. In such cases, if the system can ask back to the user to clarify the target document, the user can reach the target document efficiently. In this paper, we propose a document retrieval system that has a function to ask back questions that narrow down the search target and questions that expand the the search target, based on information extending the existing tags assigned to documents, in order to clarify the user's search intention and to expand the query easily. We implemented this system and evaluated the effectiveness of the proposed method by comparing it with existing tag search in a use case of of public field.