The Internet plays host to many millions of documents and images and is increasing in size all the time. As a result locating web content is becoming increasingly difficult for users, and search traffic from users and spiders is increasing rapidly. A directory of the contents of the emerging cache hierarchy would be more complete than existing tools and avoid the need for spider traffic. However it is also essential to minimise the user traffic. A promising approach is to encourage users to refine their queries through categories. Automatic categorisation of a cache directory is demonstrated, and an adaptive categorisation scheme is proposed.