A latent model for collaborative filtering

Helge Langseth, Thomas Dyhre Nielsen

Research output: Working paperResearch

308 Downloads (Pure)

Abstract

Recommender systems based on collaborative filtering have received a great deal of interest over the last decade. Typically, these types of systems either take a user-centered or an item-centered approach when making recommendations, but by employing only one of these two perspectives we may unintentionally leave out important information that could otherwise have improved the recommendations. In this paper, we propose a collaborative filtering model that contains an explicit representation of all items and users. Experimental results show that the proposed system obtains significantly better results than other collaborative filtering systems (evaluated on the MovieLens data set). Furthermore, the explicit representation of all users and items allows the model to e.g. make group-based recommendations balancing the preferences of the individual users.
Original languageEnglish
Place of PublicationAalborg Universitet
PublisherDepartment of Computer Science, Aalborg University
Number of pages24
Publication statusPublished - 2009

Fingerprint

Collaborative filtering
Recommender systems

Keywords

  • Recommender systems
  • Collaborative filtering
  • Graphical models
  • Latent variables

Cite this

Langseth, H., & Nielsen, T. D. (2009). A latent model for collaborative filtering. Aalborg Universitet: Department of Computer Science, Aalborg University.
Langseth, Helge ; Nielsen, Thomas Dyhre. / A latent model for collaborative filtering. Aalborg Universitet : Department of Computer Science, Aalborg University, 2009.
@techreport{82e5aab0a99e11de8c1c000ea68e967b,
title = "A latent model for collaborative filtering",
abstract = "Recommender systems based on collaborative filtering have received a great deal of interest over the last decade. Typically, these types of systems either take a user-centered or an item-centered approach when making recommendations, but by employing only one of these two perspectives we may unintentionally leave out important information that could otherwise have improved the recommendations. In this paper, we propose a collaborative filtering model that contains an explicit representation of all items and users. Experimental results show that the proposed system obtains significantly better results than other collaborative filtering systems (evaluated on the MovieLens data set). Furthermore, the explicit representation of all users and items allows the model to e.g. make group-based recommendations balancing the preferences of the individual users.",
keywords = "Recommender systems, Collaborative filtering, Graphical models, Latent variables",
author = "Helge Langseth and Nielsen, {Thomas Dyhre}",
note = "Technical Report 09-003",
year = "2009",
language = "English",
publisher = "Department of Computer Science, Aalborg University",
type = "WorkingPaper",
institution = "Department of Computer Science, Aalborg University",

}

Langseth, H & Nielsen, TD 2009 'A latent model for collaborative filtering' Department of Computer Science, Aalborg University, Aalborg Universitet.

A latent model for collaborative filtering. / Langseth, Helge; Nielsen, Thomas Dyhre.

Aalborg Universitet : Department of Computer Science, Aalborg University, 2009.

Research output: Working paperResearch

TY - UNPB

T1 - A latent model for collaborative filtering

AU - Langseth, Helge

AU - Nielsen, Thomas Dyhre

N1 - Technical Report 09-003

PY - 2009

Y1 - 2009

N2 - Recommender systems based on collaborative filtering have received a great deal of interest over the last decade. Typically, these types of systems either take a user-centered or an item-centered approach when making recommendations, but by employing only one of these two perspectives we may unintentionally leave out important information that could otherwise have improved the recommendations. In this paper, we propose a collaborative filtering model that contains an explicit representation of all items and users. Experimental results show that the proposed system obtains significantly better results than other collaborative filtering systems (evaluated on the MovieLens data set). Furthermore, the explicit representation of all users and items allows the model to e.g. make group-based recommendations balancing the preferences of the individual users.

AB - Recommender systems based on collaborative filtering have received a great deal of interest over the last decade. Typically, these types of systems either take a user-centered or an item-centered approach when making recommendations, but by employing only one of these two perspectives we may unintentionally leave out important information that could otherwise have improved the recommendations. In this paper, we propose a collaborative filtering model that contains an explicit representation of all items and users. Experimental results show that the proposed system obtains significantly better results than other collaborative filtering systems (evaluated on the MovieLens data set). Furthermore, the explicit representation of all users and items allows the model to e.g. make group-based recommendations balancing the preferences of the individual users.

KW - Recommender systems

KW - Collaborative filtering

KW - Graphical models

KW - Latent variables

M3 - Working paper

BT - A latent model for collaborative filtering

PB - Department of Computer Science, Aalborg University

CY - Aalborg Universitet

ER -

Langseth H, Nielsen TD. A latent model for collaborative filtering. Aalborg Universitet: Department of Computer Science, Aalborg University. 2009.