Agent-based recommender systems assist users based on their preferences and those of similar users. However, when dealing with multimedia contents they need of: (i) selecting as recommenders those users that have similar profiles and that are reliable in providing suggestions and (ii) considering the effects of the device currently exploited. To address these issues, we propose a multi-agent architecture, called MART, conceived to this aim and based on a particular trust model. Some experimental results are presented to evaluate our proposal, that show MART is more effective, in terms of suggestion quality, than other agent-based recommenders.
Using Agents for Generating Personalized Recommendations of Multimedia Contents / Rosaci, D; Sarne, G. - 8249:(2013), pp. 409-420. (Intervento presentato al convegno AI*IA 2013 - XIII Conference of the Italian Association for Artificial Intelligence tenutosi a Turin, Italy nel 04-06/12/2013) [10.1007/978-3-319-03524-6_35].
Using Agents for Generating Personalized Recommendations of Multimedia Contents
ROSACI D;SARNE G
2013-01-01
Abstract
Agent-based recommender systems assist users based on their preferences and those of similar users. However, when dealing with multimedia contents they need of: (i) selecting as recommenders those users that have similar profiles and that are reliable in providing suggestions and (ii) considering the effects of the device currently exploited. To address these issues, we propose a multi-agent architecture, called MART, conceived to this aim and based on a particular trust model. Some experimental results are presented to evaluate our proposal, that show MART is more effective, in terms of suggestion quality, than other agent-based recommenders.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
