A framework for recommender systems based on ontologies and bayesian networks
Unpublished, Number 02/78 - november 2002
Recommender systems are a key issue in e-commerce and resource retrieval software tools, since they greatly improve effectiveness and usability of services. However general purpose frameworks for recommender systems are still to come and developers who want to add even basic personalization features to their services, must realize them from scratch. We propose a framework for recommender systems that, using ontologies for inter-operability and logical inference and bayesian networks for uncertain reasoning, helps developers realize personalized services with minimum effort.
Références BibTex
@Unpublished{SSPGM02a,
author = {Sanna, R. and Soro, A. and Paddeu, G. and Giroux, S. and Moulin, C.},
title = {A framework for recommender systems based on ontologies and bayesian networks},
institution = {CRS4, Center for Advanced Studies, Research, and Development in Sardinia},
number = {02/78},
month = {november},
year = {2002},
address = {Cagliari, Italy},
type = {techreport},
url = {https://publications.crs4.it/pubdocs/2002/SSPGM02a},
}