With the rapid growth of World Wide Web (WWW), it is very difficult for website owners to track and understand the need of the users. Hence, a semiautomatic intelligent analyzer can be used to find out the browsing patterns of a user. Moreover, the pattern which is revealed from this deluge of web access logs must be interesting, useful and understandable. With these three objectives, this paper uses a radial basis function neural network to classify the webpages based on the time of access and region of access. The experimental results obtained from this study are encouraging for further intensive research.

Over the last decade, the proliferation of information on the WWW, called in short web, has resulted in a large repository of web documents stored in multiple websites. This plethora and diversity of resources have promoted the need for developing a semiautomatic mining technique on the WWW, thereby giving rise to the term web mining (Hu et al., 2003).

Every website contains multiple webpages. Every webpage has: 1) contents which can be in any form, e.g., text, graphics, multimedia, etc.; 2) links from one page to another; and 3) users accessing the webpages. According to this, the area, web mining, can be categorized as follows.

Mining the contents of webpages is called ‘Content Mining’. Mining the links between webpages is called ‘Structure Mining’. Mining the web access logs is called ‘Web Usage Mining’. Figure 1 describes the categorization of web mining.

All web servers maintain exhaustive log files about user interactions. Whenever a request for resources is received, the web server records it in the log file according to the format specified by the server administrator. For example, Apache web server.

Information Technology Journal, Neural networks, Radial basis function neural networks, Classification, Web log.