Open AccessRelevance, diversity and serendipity in content recommendation using clustering
Author(s) -
Fernando Henrique da Silva Costa,
Andrei Martins Silva,
Sarajane Marques Peres
Publication year - 2018
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5753/eniac.2018.4463
Subject(s) - serendipity , relevance (law) , cluster analysis , context (archaeology) , diversity (politics) , computer science , recommender system , quality (philosophy) , content (measure theory) , information retrieval , artificial intelligence , mathematics , sociology , epistemology , geography , mathematical analysis , philosophy , archaeology , political science , anthropology , law
In this paper, over-specialization in content-based recommender sys- tems is explored through the definition and analysis of recommendation strate- gies aiming at quality in terms of relevance, diversity and serendipity. Clustering is applied as the basis for building these strategies, applied to the news context. The results show the feasibility of the proposed strategies.