Show simple item record

dc.contributor.advisorPears, Russel
dc.contributor.authorCristache, Alex
dc.date.accessioned2010-03-11T02:58:45Z
dc.date.available2010-03-11T02:58:45Z
dc.date.copyright2009
dc.date.issued2010-03-11T02:58:45Z
dc.identifier.urihttp://hdl.handle.net/10292/822
dc.description.abstractRecommender systems are increasingly being used in today’s world. Collaborative filtering, together with association rules mining are probably the most widely used methods to implement recommender systems. In this dissertation we undertake a review of past research conducted in the area of recommender systems with the focus being the use of association rule mining. We propose a novel methodology that combines the use of association mining with the use of distance metrics such as the Jaccard measure to identify movies that belong to the same genre. Our experimental results on the MovieLens dataset shows that the use of the Jaccard metric improved the coverage of recommendations over the use of the standard association rule mining method.
dc.language.isoenen
dc.publisherAuckland University of Technology
dc.subjectRecommender system
dc.subjectAssociation rules
dc.subjectJaccard
dc.subjectMovieLens
dc.subjectConstructive research
dc.subjectComputer science
dc.titleHybrid recommender system using association rules
dc.typeThesis
thesis.degree.grantorAuckland University of Technology
thesis.degree.levelMasters Dissertations
thesis.degree.nameMaster of Computer and Information Sciences
dc.rights.accessrightsOpenAccess
dc.date.updated2010-03-10T04:16:17Z


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record