Web user behavioral profiling for user identification
Decision Support Systems
A user meta-model for context-aware recommender systems
Proceedings of the 1st International Workshop on Information Heterogeneity and Fusion in Recommender Systems
Hi-index | 0.00 |
This paper discusses a research project: rule-based Web user profiling platform. In this platform, usage data are encoded as a sequence of events, each of which represents an action performed by a user on a Web service at a given time. An event template is proposed to define event models for different Web services. The platform is rule-based. Rules define profile metrics and determine how to compute profile metrics from usage events. A prototype of the platform was implemented and was applied to generate profiles from page view events. The major contribution of the work is the rule-based approach to user profiling. It is the rules and the event template that provide the flexibility to allow the platform to be configured for different Web services.