Crowww Classification and Retrieval on WWW

摘要

A couple of years ago, when we were all facing the problem of how to get the information we want from the ever-spreading web, there were basically two choices for conducting a search - the quantity-oriented gather-all-pages-and-search approach as typified by AltaVista, and the quality-controlled hand-classify-and-search approach as represented by Yahoo. We've since learned that there are significant advantages and drawbacks to both. In response, a number of services and applications, including meta-search engines, personal web robots and cache organizers, have appeared [Lawrence]. For Japanese web documents, the alternatives have often been restricted by the inability to search using Japanese. Our research group has created a system incorporating the better features of the two basic approaches while avoiding their limitations.