Interactive Web Search by Graphical Query Refinement

We propose a new Web search system that helps users clarify their information needs through interaction. The system represents the user's information needs using a query graph: a set of graph objects. This graph is presented to the user and interactively refined until the user is satisfied with the documents retrieved. In this way, the user can find the desired documents even if his original information needs were unclear. This article describes the concept and outlines the system. Keywords Web search, information needs, graphical interface, interaction, relevance feedback.

• Type: Conference
• Year: 2001
• Where: WWW
• Authors: Junji Tomita, Gen-ichiro Kikui
• Keywords: Information Needs, Internet Technology, Original Information Needs, Query Graph, WWW 2001