Local Buffer as Source of Web Mining Data

摘要

The data for Web mining is usually extracted from the WWW server or proxy server log files. The paper examines the advantages and disadvantages of exploiting another source of input data - the browser buffer. The properties of data extracted from different types of sources are compared. The browser buffer contains data about user navigational habits as well as the formal properties and the content of all recently accessed WWW objects. The paper uses the data obtained from this source to examine the statistical properties of different types of texts extracted from HTML pages.