A document retrieval system for assisting creative research

A document retrieval system for researchers in the field of science and technology is described. The main feature of the system is the visualization of the semantic relations between documents and interactive operation to it. The system has a test database composed of a large number of journal and conference papers. It elicits keywords from each paper by an automatic indexing algorithm and then visualizes the relations between the papers and their keywords in a metric space by a statistical method. We carried out several experiments. Through them we have confirmed that our system is effective because it can assist the users in their query formation and modification while they use it. Our system can also assist the users in their creative research because they can find the relations between the papers and their keywords which they haven't noticed so far. In the experiment on effectiveness of retrieval we have confirmed that our system can realize high recall and precision ratio.