This paper presents new services for intelligent monitoring and visualising user accesses to a university's web site. These are based on the use of data mining techniques to process data recorded in the web log files and visualise it. The system is to be used by the University of Nottingham's web server to monitor the interests of potential students and to predict the student numbers and their geographical distributions for the next academic year. The system could be adapted for other academic and commercial web sites to assist intelligent decision making.