BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Date iCal//NONSGML kigkonsult.se iCalcreator 2.20.2//
METHOD:PUBLISH
X-WR-CALNAME;VALUE=TEXT:Eventi DIAG
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:STANDARD
DTSTART:20151025T030000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RDATE:20161030T030000
TZNAME:CET
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20160327T020000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:calendar.7277.field_data.0@oba.diag.uniroma1.it
DTSTAMP:20260408T053039Z
CREATED:20160401T095343Z
DESCRIPTION:We study how to improve the accuracy and running time of top-Nr
 ecommendation with collaborative filtering (CF). Unlike existing works tha
 tuse mostly rated items (which is only a small fraction in a rating matrix
 )\,we propose the notion of pre-use preferences of users toward a vast amo
 untof unrated items. Using this novel notion\, we effectively identifyunin
 teresting items that were not rated yet but are likely to receive verylow 
 ratings from users\, and impute them as zero. This simple-yet-novelzero-in
 jection method applied to a set of carefully-chosen uninterestingitems not
  only addresses the sparsity problem by enriching a rating matrixbut also 
 completely prevents uninteresting items from being recommended astop-N ite
 ms\, thereby improving accuracy greatly. As our proposed idea ismethod-agn
 ostic\, it can be easily applied to a wide variety of popular CFmethods. T
 hrough comprehensive experiments using the Movielens dataset andMyMediaLit
 e implementation\, we successfully demonstrate that our solutionconsistent
 ly and universally improves the accuracies of popular CF methods(e.g.\, it
 em-based CF\, SVD-based CF\, and SVD++) by two to five orders ofmagnitude 
 on average. Furthermore\, our approach reduces the running time ofthose CF
  methods by 1.2 to 2.3 times when its setting produces the bestaccuracy.Bi
 o. Sang-Wook Kim received the B.S. degree in Computer Engineering from Seo
 ulNational University\, Korea at 1989\, and earned the M.S. and Ph.D. degr
 ees inComputer Science from Korea Advanced Institute of Science and Techno
 logy(KAIST)\, Korea at 1991 and 1994\, respectively. From 1995 to 2003\, h
 e servedas an Associate Professor of the Division of Computer\, Informatio
 n\, andCommunications Engineering at Kangwon National University\, Korea. 
 In 2003\,he joined Hanyang University\, Seoul\, Korea\, where he currently
  is aProfessor at the Department of Computer Science and Engineering and t
 hedirector of the Brain-Korea-21-Plus research program. He is also leading
  aNational Research Lab (NRL) Project funded by National Research Foundati
 onfrom 2015. His research interests include databases\, data mining\, mult
 imediainformation retrieval\, social network analysis\, recommendation\, a
 nd web dataanalysis.From 2009 to 2010\, Dr. Kim visited Computer Science D
 epartment at CarnegieMellon University as a Visiting Research Professor. F
 rom 1999 to 2000\, heworked with the IBM T. J. Watson Research Center\, US
 A as a Post-Doc. He alsovisited the Computer Science Department of Stanfor
 d University as a VisitingResearcher in 1991. He is an author of over 110 
 papers in refereedinternational journals and international conference proc
 eedings. He receivedthe Best Paper Award from the 29th ACM International S
 ymposium on AppliedComputing (ACM SAC) in 2014 and the Best Poster Present
 ation Award from theACM International Conference on Information and Knowle
 dge Management (ACMCIKM) in 2013.  Dr. Kim was a General Co-Chair of the 6
 th InternationalConference on Computational Collective Intelligence Techno
 logies andApplications in 2014\, a Program Co-Chair for the ACM Internatio
 nalConference on Ubiquitous Information Management and Communications (ACM
 IMCOM) in 2015\, a Program Co-Chair for the International Conference onEme
 rging Databases in 2013\, and a Track Chair for the Social Network andMedi
 a Analysis Track in the ACM International Symposium on Applied Computing(A
 CM SAC) in 2015.  He also served Program Committees of over 80internationa
 l conferences including IEEE ICDE\, VLDB\, WWW\, and ACM CIKM. Heis now an
  associate editor of Information Sciences.  Dr. Kim received theOutstandin
 g Service Award from ACM SIGAPP and the Outstanding ContributionsAward fro
 m Database Society of Korea in 2014. He is a member of the ACM andthe IEEE
 .
DTSTART;TZID=Europe/Paris:20160411T110000
DTEND;TZID=Europe/Paris:20160411T110000
LAST-MODIFIED:20190805T155749Z
LOCATION:Aula Magna\, DIAG
SUMMARY:Sang-Wook Kim - “Told You I Didn’t Like It”: Exploiting Uninteresti
 ng Items for Effective Collaborative Filtering -  Monday April 11\, 2016\,
  11:00 - Aula Magna - Sang-Wook Kim (Hanyang University\, Seoul\, Korea)
URL;TYPE=URI:http://oba.diag.uniroma1.it/node/7277
END:VEVENT
END:VCALENDAR