Collaborative Filtering via Gaussian Probabilistic Latent Semantic Analysis
T. Hofmann. volume 2003 of Special issue of the SIGIR forum, New York, NY, International Conference on Research and Development in Information Retrieval, ACM Press, (2003)
Abstract
Collaborative Filtering aims at learning predictive models of user preferences, interests or behavior from community data, i.e. a database of available user preferences. In this paper, we describe a new model-based algorithm designed for this task, which is based on a generalization of probabilistic latent semantic analysis to continuous-valued response variables. More specically, we assume that the observed user ratings can be modeled as a mixture of user communities or interest groups, where users may participate probabilistically in one or more groups. Each community is characterized by a Gaussian distribution on the normalized ratings for each item. The normalization of ratings is performed in a user-specific manner to account for variations in absolute shift and variance of ratings. Experiments on the Each-Movie data set show that the proposed approach compares favorably with other collaborative Filtering techniques.
%0 Conference Paper
%1 Hofmann2003
%A Hofmann, Thomas
%B Special issue of the SIGIR forum
%C New York, NY
%D 2003
%E Callan, Jamie
%I ACM Press
%K analysis collaborative filtering latent probabilistic semantic
%T Collaborative Filtering via Gaussian Probabilistic Latent Semantic Analysis
%U http://www.cs.brown.edu/~th/papers/Hofmann-SIGIR2003.pdf
%V 2003
%X Collaborative Filtering aims at learning predictive models of user preferences, interests or behavior from community data, i.e. a database of available user preferences. In this paper, we describe a new model-based algorithm designed for this task, which is based on a generalization of probabilistic latent semantic analysis to continuous-valued response variables. More specically, we assume that the observed user ratings can be modeled as a mixture of user communities or interest groups, where users may participate probabilistically in one or more groups. Each community is characterized by a Gaussian distribution on the normalized ratings for each item. The normalization of ratings is performed in a user-specific manner to account for variations in absolute shift and variance of ratings. Experiments on the Each-Movie data set show that the proposed approach compares favorably with other collaborative Filtering techniques.
%@ 1581136463
@inproceedings{Hofmann2003,
abstract = {Collaborative Filtering aims at learning predictive models of user preferences, interests or behavior from community data, i.e. a database of available user preferences. In this paper, we describe a new model-based algorithm designed for this task, which is based on a generalization of probabilistic latent semantic analysis to continuous-valued response variables. More specically, we assume that the observed user ratings can be modeled as a mixture of user communities or interest groups, where users may participate probabilistically in one or more groups. Each community is characterized by a Gaussian distribution on the normalized ratings for each item. The normalization of ratings is performed in a user-specific manner to account for variations in absolute shift and variance of ratings. Experiments on the Each-Movie data set show that the proposed approach compares favorably with other collaborative Filtering techniques.},
added-at = {2007-06-28T14:05:08.000+0200},
address = {New York, NY},
author = {Hofmann, Thomas},
biburl = {https://www.bibsonomy.org/bibtex/2b524d104b84217ba969065143a05d114/wnpxrz},
crossref = {Callan:2003},
editor = {Callan, Jamie},
interhash = {0cf79cc120e8c7b216ba6bfa708a4602},
intrahash = {b524d104b84217ba969065143a05d114},
isbn = {1581136463},
keywords = {analysis collaborative filtering latent probabilistic semantic},
organization = {International Conference on Research and Development in Information Retrieval},
publisher = {ACM Press},
series = {Special issue of the SIGIR forum},
timestamp = {2007-06-28T14:05:08.000+0200},
title = {Collaborative Filtering via Gaussian Probabilistic Latent Semantic Analysis},
url = {http://www.cs.brown.edu/~th/papers/Hofmann-SIGIR2003.pdf},
volume = 2003,
year = 2003
}