Intelligent-information-access-method acquisition system using temporal information

摘要

WWW and e-mail are currently used for various purposes. To achieve these various purposes, we use all sorts of tools. But, there exists a great gap between those true purposes and works that use those tools. WWW and existing tools force us to fill the gap. So, we have proposed the SPIRAL system, which is a seamless user environment for WWW. In SPIRAL, when user access to web sites, SPIRAL stores not only the URLs but also the contents with time stamp. So, we have proposed information-access-method acquisition system using these temporal information. In this article, we report how to solve performance problems when we use a large scale of data and how to create a training example from multiple access data.