Neighbourhood-based collaborative filtering recommenders exploit the common ratings among users to identify a user’s most similar neighbours. It is known that decisions made on a naive computation of user similarity are unreliable, because the number of co-ratings varies strongly among users. In this paper, we formalize the notion of reliable similarity between two users and propose a method that constructs a user’s neighbourhood by selecting only those users that are reliably similar to her. Our method combines a statistical test and the notion of a baseline user. We report our results on typical benchmark datasets.
%0 Book Section
%1 MatuszykSpiliopoulouUmap14
%A Matuszyk, Pawel
%A Spiliopoulou, Myra
%B User Modeling, Adaptation, and Personalization
%D 2014
%E Dimitrova, Vania
%E Kuflik, Tsvi
%E Chin, David
%E Ricci, Francesco
%E Dolog, Peter
%E Houben, Geert-Jan
%I Springer International Publishing
%K from:matuszyk
%P 146–157
%R 10.1007/978-3-319-08786-3_13
%T Hoeffding-CF: Neighbourhood-Based Recommendations on Reliably Similar Users
%U https://kmd.cs.ovgu.de/pub/matuszyk/MatuszykSpiliopoulou14.pdf
%V 8538
%X Neighbourhood-based collaborative filtering recommenders exploit the common ratings among users to identify a user’s most similar neighbours. It is known that decisions made on a naive computation of user similarity are unreliable, because the number of co-ratings varies strongly among users. In this paper, we formalize the notion of reliable similarity between two users and propose a method that constructs a user’s neighbourhood by selecting only those users that are reliably similar to her. Our method combines a statistical test and the notion of a baseline user. We report our results on typical benchmark datasets.
%@ 978-3-319-08785-6
@incollection{MatuszykSpiliopoulouUmap14,
abstract = {Neighbourhood-based collaborative filtering recommenders exploit the common ratings among users to identify a user’s most similar neighbours. It is known that decisions made on a naive computation of user similarity are unreliable, because the number of co-ratings varies strongly among users. In this paper, we formalize the notion of reliable similarity between two users and propose a method that constructs a user’s neighbourhood by selecting only those users that are reliably similar to her. Our method combines a statistical test and the notion of a baseline user. We report our results on typical benchmark datasets.},
added-at = {2015-10-15T21:30:56.000+0200},
author = {Matuszyk, Pawel and Spiliopoulou, Myra},
biburl = {https://www.bibsonomy.org/bibtex/2d18c25896610735069aa41edad89c1c0/kmd-ovgu},
booktitle = {User Modeling, Adaptation, and Personalization},
doi = {10.1007/978-3-319-08786-3_13},
editor = {Dimitrova, Vania and Kuflik, Tsvi and Chin, David and Ricci, Francesco and Dolog, Peter and Houben, Geert-Jan},
interhash = {469167b771e8cae285e1ed9f663c26b0},
intrahash = {d18c25896610735069aa41edad89c1c0},
isbn = {978-3-319-08785-6},
keywords = {from:matuszyk},
language = {English},
pages = {146–157},
publisher = {Springer International Publishing},
series = {Lecture Notes in Computer Science},
timestamp = {2018-03-09T12:44:05.000+0100},
title = {Hoeffding-CF: Neighbourhood-Based Recommendations on Reliably Similar Users},
url = {https://kmd.cs.ovgu.de/pub/matuszyk/MatuszykSpiliopoulou14.pdf},
volume = 8538,
year = 2014
}