We describe a novel statistical model, the tied Boltzmann machine, for combining collaborative and content information for recommendations. In our model, pairwise interactions between items are captured through a Boltzmann machine, whose parameters are constrained according to the content associated with the items. This allows the model to use content information to recommend items that are not seen during training. We describe a tractable algorithm for training the model, and give experimental results evaluating the model in two cold start recommendation tasks on the MovieLens data set.
Описание
Tied boltzmann machines for cold start recommendations
%0 Conference Paper
%1 paper:gunawardana:2008
%A Gunawardana, Asela
%A Meek, Christopher
%B RecSys '08: Proceedings of the 2008 ACM Conference on Recommender Systems
%C New York, NY, USA
%D 2008
%I ACM
%K RecSys RecSys08 algorithms machine-learning to-read
%P 19--26
%R http://doi.acm.org/10.1145/1454008.1454013
%T Tied boltzmann machines for cold start recommendations
%U http://portal.acm.org/citation.cfm?id=1454013
%X We describe a novel statistical model, the tied Boltzmann machine, for combining collaborative and content information for recommendations. In our model, pairwise interactions between items are captured through a Boltzmann machine, whose parameters are constrained according to the content associated with the items. This allows the model to use content information to recommend items that are not seen during training. We describe a tractable algorithm for training the model, and give experimental results evaluating the model in two cold start recommendation tasks on the MovieLens data set.
%@ 978-1-60558-093-7
@inproceedings{paper:gunawardana:2008,
abstract = {We describe a novel statistical model, the tied Boltzmann machine, for combining collaborative and content information for recommendations. In our model, pairwise interactions between items are captured through a Boltzmann machine, whose parameters are constrained according to the content associated with the items. This allows the model to use content information to recommend items that are not seen during training. We describe a tractable algorithm for training the model, and give experimental results evaluating the model in two cold start recommendation tasks on the MovieLens data set.},
added-at = {2009-04-10T09:46:35.000+0200},
address = {New York, NY, USA},
author = {Gunawardana, Asela and Meek, Christopher},
biburl = {https://www.bibsonomy.org/bibtex/26c30d7b1f4e2ca6d93d5e94341d3a3a1/mschuber},
booktitle = {RecSys '08: Proceedings of the 2008 ACM Conference on Recommender Systems},
description = {Tied boltzmann machines for cold start recommendations},
doi = {http://doi.acm.org/10.1145/1454008.1454013},
interhash = {d441e22ac241a4b713043de1894e7390},
intrahash = {6c30d7b1f4e2ca6d93d5e94341d3a3a1},
isbn = {978-1-60558-093-7},
keywords = {RecSys RecSys08 algorithms machine-learning to-read},
location = {Lausanne, Switzerland},
pages = {19--26},
publisher = {ACM},
timestamp = {2009-05-08T14:07:15.000+0200},
title = {Tied boltzmann machines for cold start recommendations},
url = {http://portal.acm.org/citation.cfm?id=1454013},
year = 2008
}