Institut for Forretningsudvikling og Teknologi

NUNI (New User and New Item) Problem for SRSs Using Content Aware Multimedia-Based Approach

Publikation: Bidrag til bog/antologi/rapport/proceedingKonferencebidrag i proceedingsForskningpeer review

Recommendation systems suggest items and users of interest based on
preferences of items or users and item or user attributes. In social media-based
services of dynamic content (such as news, blog, video, movies, books, etc.),
recommender systems face the problem of discovering new items, new users, and
both, a problem known as a cold start problem, i.e., the incapability to provide
recommendation for new items, new users, or both, due to few rating factors
available in the rating matrices. To this end, we present a biclustering technique, a novel cold start recommendation method that solves the problem of identifying the new items and new users, to alleviate the dimensionality of the item-user rating matrix using biclustering technique. To overcome the information exiguity and rating diversity, it uses the smoothing and fusion technique. As discussed, the system presents content aware multimedia-based social recommender media substance from item and user bunches.
OriginalsprogEngelsk
TitelInnovations in Computer Science and Engineering : Proceedings of the Third ICICSE, 2015
RedaktørerH. S. Saini, Rishi Sayal, Sandeep Singh Rawat
UdgivelsesstedSingapore
ForlagSpringer Science+Business Media
Udgivelsesår20 feb. 2016
Sider287-294
Kapitel34
ISBN (trykt)978-981-10-0417-9
ISBN (Elektronisk)978-981-10-0419-3
DOI
StatusUdgivet - 20 feb. 2016
Eksternt udgivetJa
BegivenhedInternational Conference on Innovations in Computer Science & Engineering - Guru Nanak Institutions, Hyderabad, Indien
Varighed: 7 aug. 20158 aug. 2015
Konferencens nummer: 3rd
http://icicse2015.org/

Konference

KonferenceInternational Conference on Innovations in Computer Science & Engineering
Nummer3rd
LokationGuru Nanak Institutions
LandIndien
ByHyderabad
Periode07/08/201508/08/2015
Internetadresse
SerietitelAdvances in Intelligent Systems and Computing
Vol/bind413
ISSN2194-5357

Se relationer på Aarhus Universitet Citationsformater

ID: 112681060