An argumentative recommendation approach based on contextual aspects

GUILLERMO R. SIMARI; LLUIS GODO; JUAN CARLOS LIONEL TEZE

Milan

Conferencia; THE 12TH INTERNATIONAL CONFERENCE ON SCALABLE UNCERTAINTY MANAGEMENT; 2018

Department of Informatics, Systems and Communications (DISCo) of the University of Milano-Bicocca

Argumentation-based recommender systems constitute aninteresting tool to provide reasoned recommendations in complexdomains with unresolved contradictory information situations andincomplete information. In these systems, the use of contextual infor-mation becomes a central issue in order to come up with personalizedrecommendations. An argumentative recommender system that offersmechanisms to handle contextual aspects of the recommendation domainprovides an important ability that can be exploited by the user. However,in most of existing works, this issue has not been extensively studied. Inthis work, we propose an argumentation-based formalization for dealingwith this issue. We present a general framework that allows the design ofrecommender systems capable of handling queries that can include (pos-sibly inconsistent) contextual information under which recommendationsshould be computed. To answer a query, in the proposed argumentation-based approach, the system first selects alternative instances accordingto the user?s supplied contextual information, and then makes recom-mendations, in both cases through a defeasible argumentative analysis.