A Non-Parametric Approach to Web Log Analysis
- Andrew Foss, University of Alberta
- Weinan Wang
- Osmar R. Zaiane, University of Alberta (Database)
Clustering data generally involves some input parameters or heuristics that are usually unknown at the time they are needed. We discuss the general problem of parameters in clustering and present a new approach, TURN, based on boundary detection and apply it to the clustering of web log data. We also present the use of different filters on the web log data to focus the clustering results and discuss different coefficients for defining similarity in a non-Euclidean space.
Citation
A. Foss, W. Wang, O. Zaiane. "A Non-Parametric Approach to Web Log Analysis". Workshop on Web Mining, pp 41-50, April 2001.Keywords: | |
Category: | In Workshop |
BibTeX
@misc{Foss+al:01, author = {Andrew Foss and Weinan Wang and Osmar R. Zaiane}, title = {A Non-Parametric Approach to Web Log Analysis}, Pages = {41-50}, booktitle = {Workshop on Web Mining}, year = 2001, }Last Updated: February 05, 2020
Submitted by Sabina P