Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

New hybrid recommender approaches: An application to equity funds selection, algorithmic decision theory

Matsatsinis Nikolaos, Manarolis Eleftherios

Full record


URI: http://purl.tuc.gr/dl/dias/CD058E7D-0F1C-4A1F-B007-2478143CF42C
Year 2009
Type of Item Peer-Reviewed Journal Publication
License
Details
Bibliographic Citation Matsatsinis, N.F., E. A. Manarolis, "New Hybrid Recommender Approaches: An Application to Equity Funds Selection, Algorithmic Decision Theory", Lecture Notes in Computer Science, vol. 5783/2009, pp. 156-167. DOI: 10.1007/978-3-642-04428-1_14 https://doi.org/10.1007/978-3-642-04428-1_14
Appears in Collections

Summary

Recommender Systems and Multicriteria Decision Analysis remain two separate scientific fields in spite of their similarity in supporting the decision making process and reducing information overload. In this paper we present a novel algorithmic framework, which combines features from Recommender Systems literature and Multicriteria Decision Analysis to alleviate the sparsity problem and the absence of multidimensional correlation measures. We apply the introduced framework for recommending Greek equity funds to a set of simulation generated investors. The proposed framework treats MCDA’s algorithm UTADIS as a content - based recommendation technique which, in conjunction with collaborative filtering results in two Hybrid Recommendation approaches. The resulting approaches manage to outperform the separate application of the UTADIS and collaborative filtering methods in terms of recommendation accuracy

Services

Statistics