IDEAL 2002: Intelligent Data Engineering and Automated Learning — IDEAL 2002 pp 31-34 | Cite as
Automated Personalisation of Internet Users Using Self-Organising Maps
Abstract
Automated personalisation offers a major improvement in the use of large Web sites. These systems learn from a user and suggest where on the Web site a user might move. Self-organising maps (SOM) may also be considered as a potential tool for Web data analysis. In this paper, the use of SOM analysis for automated personalisation of Internet users is demonstrated. The map was obtained by training a self-organising network with user demographics; click stream data were used to calculate the probabilities of user behaviour on the Web site. Thus, the map can be used for personalisation of users and to calculate the probabilities of each neuron in predicting where the user will next move on the Web site. The results indicate that SOM analysis can successfully process Web information.
Keywords
Internet User Neural Network Analysis Page Request Page Access Dwelling PlacePreview
Unable to display preview. Download preview PDF.
References
- 1.Schafer, J. B., Konstan, J., Riedi, J.: Recommender systems in e-commerce. Proceedings of first ACM conference on Electronic commerce (1999) 158–166Google Scholar
- 2.Sarwar, B., Karypis, G., Konstan, J., Riedi, J.: Analysis of recommendation algorithms for e-commerce. Proceedings of 2nd ACM conference on Electronic commerce (2000) 158–167Google Scholar
- 3.Kaartinen, J.; Hiltunen, Y.; Kovanen, P. T.; Ala-Korpela, M.: NMR Biomed. 11 (1998) 1 168–176CrossRefGoogle Scholar
- 4.Hyvönen, M., Hiltunen, Y., El-Deredy W., Ojala, T., Vaara, J., Kovanen, P., Ala-Korpela, M.: Application of Self-Organizing Maps in Conformational Analysis of Lipids. J. Am. Chem. Soc. 123 (2001) 810–816CrossRefGoogle Scholar
- 5.Kohonen, T.: Self-organizing Maps. Springer-Verlag, Berlin Heidelberg New York (2001)MATHGoogle Scholar