Understanding User Behaviour with Web Session Clustering and User Engagement Metrics

Citation

Lim, Zhou Yi and Ong, Lee Yeng and Leow, Meng Chew and Lee, Ting Wei and Tay, Qi Ming (2023) Understanding User Behaviour with Web Session Clustering and User Engagement Metrics. In: 2023 19th IEEE International Colloquium on Signal Processing & Its Applications (CSPA), 3-4 Mar 2023, Kedah, Malaysia.

[img] Text
27.pdf - Published Version
Restricted to Repository staff only

Download (479kB)

Abstract

Clustering technique is one of the mining techniques in web usage mining to discover user navigational behaviour of similar patterns from a weblog. A weblog is preprocessed to retrieve the web sessions that can be assigned to the clustering techniques. Clustering will group similar user access patterns from web sessions. The main purpose of web session clustering is to evaluate how engaged the users are with the website. Therefore, there are two important parts presented in this paper, which are the performance analysis of different clustering techniques and the utilization of user engagement metrics to describe the user behaviour of each cluster. The clustering techniques, namely k-means, k-medoids, bisecting k-means, and fuzzy c-means are applied on two different weblogs. The clustering qualities of the techniques are then compared using cluster validity indices. Two engagement metrics, which are the number of clicks and the number of sessions are calculated for each cluster to discover its user behaviour.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: clustering, user engagement, web usage mining, web session clustering
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science
Divisions: Faculty of Information Science and Technology (FIST)
Depositing User: Ms Nurul Iqtiani Ahmad
Date Deposited: 02 Jun 2023 02:02
Last Modified: 02 Jun 2023 02:02
URII: http://shdl.mmu.edu.my/id/eprint/11459

Downloads

Downloads per month over past year

View ItemEdit (login required)