Latent Class Models for Collaborative Filtering

DMCA

Latent Class Models for Collaborative Filtering (1999)

Cached

Download Links

Other Repositories/Bibliography

CiteULike
DBLP

by Thomas Hofmann , Jan Puzicha

Venue:	IN PROCEEDINGS OF THE SIXTEENTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE
Citations:	211 - 4 self

BibTeX

@INPROCEEDINGS{Hofmann99latentclass,
    author = {Thomas Hofmann and Jan Puzicha},
    title = {Latent Class Models for Collaborative Filtering},
    booktitle = {IN PROCEEDINGS OF THE SIXTEENTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE},
    year = {1999},
    pages = {688--693},
    publisher = {}
}

OpenURL

Abstract

This paper presents a statistical approachto collaborative filtering and investigates the use of latent class models for predicting individual choices and preferences based on observed preference behavior. Two models are discussed and compared: the aspect model, a probabilistic latent space model which models individual preferences as a convex combination of preference factors, and the two-sided clustering model, which simultaneously partitions persons and objects into clusters. We present EM algorithms for different variants of the aspect model and derive an approximate EM algorithm based on a variational principle for the two-sided clustering model. The benefits of the different models are experimentally investigated on a large movie data set.

Keyphrases

latent class model collaborative filtering two-sided clustering model aspect model probabilistic latent space model partition person individual choice preference factor approximate em algorithm different variant present em algorithm different model model individual preference convex combination statistical approachto collaborative filtering variational principle observed preference behavior large movie data set

DMCA