Authors:
Peter Vojtáš
and
Ladislav Peška
Affiliation:
Charles University, Czech Republic
Keyword(s):
Web Market without Dominant Seller, Small to Medium Company e-Shop, User Behavior, User Models, User Preference Learning, Performance Metrics, Offline Experiments, Production Data.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Business Performance
;
Cloud Computing
;
Collaboration and e-Services
;
Data Communication Networking
;
Data Engineering
;
e-Business
;
Enterprise Engineering
;
Enterprise Information Systems
;
Entertainment
;
Goal Setting and Setting
;
Logistics
;
Mobile Software and Services
;
Ontologies and the Semantic Web
;
Quality-Of-Service
;
Services Science
;
Software Agents and Internet Computing
;
Software Engineering
;
Software Engineering Methods and Techniques
;
Sustainable e-Business
;
Technologies and Tools
;
Technology Platforms
;
Telecommunications
;
Transaction Support
;
Web Services
;
Wireless Information Networks and Systems
Abstract:
We deal with the problem of using user behavior for business relevant analytic task processing. We describe our acquaintance with preference learning from behavior data from an e-shop. Based on our experience and problems we propose a model for collecting (java script tracking) and processing user behavior data. We present several results of offline experiments on real production data. We show that mere data on users (implicit) behavior are sufficient for improvement of prediction of user preference. As a future work we present richer data on time dependent user behavior.