Mondou: Web search engine with textual data mining

It is too difficult to discover useful documents on web without rich background knowledge. In this paper, with applying techniques of the textual data mining to the web resource discovery, we try to derive effective association rules in order to submit more effective queries. From 1995, we have been developing the resource discovery system, Mondou, which derives associative keywords from collected and parsed Japanese web pages. This paper includes a brief evaluation of our system and java applet in order to visualize search results with multi-dimensional measurements