Show simple item record

dc.contributor.authorTkalčič M
dc.contributor.authorOdić A
dc.contributor.authorKošir A
dc.contributor.authorTasič J
dc.contributor.editor
dc.date.accessioned2018-05-28T16:11:49Z
dc.date.available2018-05-28T16:11:49Z
dc.date.issued2011
dc.identifier.urihttp://dx.doi.org/10.6084/m9.figshare.94140
dc.identifier.urihttp://figshare.com/articles/LDOS{_}CAMRA{_}2011{_}submission/94140
dc.identifier.urihttp://hdl.handle.net/10863/5038
dc.description.abstractThis  paper  is  a  report  on  the  work  done  by  the  LDOS team  (from  the  University  of  Ljubljana  Faculty  of  electrical engineering) on the 2011 RecSys CAMRA (Challenge on Context-aware Movie Recommendation).  We present three approaches in the track 1 competition which requires to select the top N recommended items for each household.  Our general  approach  uses  a  matrix  factorization  algorithm  to compute per-user rating predictions.  The context is taken into  account  using  an  averaged  and  weighted  sum  to  calculate the household-based rating predictions.  We also explored how the limiting of true positive candidates with additional  knowledge  from  the  dataset  influences  the  performance of our models.en_US
dc.language.isoenen_US
dc.rightsCC BY 4.0
dc.titleLDOS CAMRA 2011 submissionen_US
dc.typeWorking Paperen_US
dc.date.updated2018-05-16T07:35:49Z
dc.publication.titlefigshare
dc.language.isiEN-GB
dc.description.fulltextopenen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record