Improving accuracy of recommender system by clustering items based on stability of user similarity

Truong Khanh Quan, Ishikawa Fuyuki, Shinichi Honiden

Research output: Chapter in Book/Report/Conference proceedingConference contribution

27 Citations (Scopus)

Abstract

Collaborative Filtering, one of the most widely used approach in Recommender System, predicts a user's rating towards an item by aggregating ratings given by users having similar preference to that user. In existing approaches, user similarity is often computed on the whole set of items. However, because the number of items is often very large, and so is the diversity among items, users who have similar preference in one category of items may have totally different judgement on items of another kind. In order to deal with this problem, we propose a method of clustering items, so that inside a cluster, similarity between users does not change significantly. After that, when predicting rating of a user towards an item, we only aggregate ratings of users who have high similarity degree with that user inside the cluster to which that item belongs. Experiments evaluating our approach are carried out on the real dataset taken from movies recommendation system of MovieLens web site. Preliminary results suggest that our approach can improve prediction accuracy compared to existing approaches.

Original languageEnglish
Title of host publicationCIMCA 2006
Subtitle of host publicationInternational Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies ...
DOIs
Publication statusPublished - 2007 Dec 1
Externally publishedYes
EventCIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies and International Commerce - Sydney, NSW
Duration: 2006 Nov 282006 Dec 1

Other

OtherCIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies and International Commerce
CitySydney, NSW
Period06/11/2806/12/1

Fingerprint

Recommender systems
Collaborative filtering
Websites
Experiments

ASJC Scopus subject areas

  • Computer Science(all)
  • Control and Systems Engineering

Cite this

Quan, T. K., Fuyuki, I., & Honiden, S. (2007). Improving accuracy of recommender system by clustering items based on stability of user similarity. In CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies ... [4052704] https://doi.org/10.1109/CIMCA.2006.123

Improving accuracy of recommender system by clustering items based on stability of user similarity. / Quan, Truong Khanh; Fuyuki, Ishikawa; Honiden, Shinichi.

CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies .... 2007. 4052704.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Quan, TK, Fuyuki, I & Honiden, S 2007, Improving accuracy of recommender system by clustering items based on stability of user similarity. in CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies ...., 4052704, CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies and International Commerce, Sydney, NSW, 06/11/28. https://doi.org/10.1109/CIMCA.2006.123
Quan TK, Fuyuki I, Honiden S. Improving accuracy of recommender system by clustering items based on stability of user similarity. In CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies .... 2007. 4052704 https://doi.org/10.1109/CIMCA.2006.123
Quan, Truong Khanh ; Fuyuki, Ishikawa ; Honiden, Shinichi. / Improving accuracy of recommender system by clustering items based on stability of user similarity. CIMCA 2006: International Conference on Computational Intelligence for Modelling, Control and Automation, Jointly with IAWTIC 2006: International Conference on Intelligent Agents Web Technologies .... 2007.
@inproceedings{b4cd0ec2171b4a3f8b930c1c5afe2e9c,
title = "Improving accuracy of recommender system by clustering items based on stability of user similarity",
abstract = "Collaborative Filtering, one of the most widely used approach in Recommender System, predicts a user's rating towards an item by aggregating ratings given by users having similar preference to that user. In existing approaches, user similarity is often computed on the whole set of items. However, because the number of items is often very large, and so is the diversity among items, users who have similar preference in one category of items may have totally different judgement on items of another kind. In order to deal with this problem, we propose a method of clustering items, so that inside a cluster, similarity between users does not change significantly. After that, when predicting rating of a user towards an item, we only aggregate ratings of users who have high similarity degree with that user inside the cluster to which that item belongs. Experiments evaluating our approach are carried out on the real dataset taken from movies recommendation system of MovieLens web site. Preliminary results suggest that our approach can improve prediction accuracy compared to existing approaches.",
author = "Quan, {Truong Khanh} and Ishikawa Fuyuki and Shinichi Honiden",
year = "2007",
month = "12",
day = "1",
doi = "10.1109/CIMCA.2006.123",
language = "English",
isbn = "0769527310",
booktitle = "CIMCA 2006",

}

TY - GEN

T1 - Improving accuracy of recommender system by clustering items based on stability of user similarity

AU - Quan, Truong Khanh

AU - Fuyuki, Ishikawa

AU - Honiden, Shinichi

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Collaborative Filtering, one of the most widely used approach in Recommender System, predicts a user's rating towards an item by aggregating ratings given by users having similar preference to that user. In existing approaches, user similarity is often computed on the whole set of items. However, because the number of items is often very large, and so is the diversity among items, users who have similar preference in one category of items may have totally different judgement on items of another kind. In order to deal with this problem, we propose a method of clustering items, so that inside a cluster, similarity between users does not change significantly. After that, when predicting rating of a user towards an item, we only aggregate ratings of users who have high similarity degree with that user inside the cluster to which that item belongs. Experiments evaluating our approach are carried out on the real dataset taken from movies recommendation system of MovieLens web site. Preliminary results suggest that our approach can improve prediction accuracy compared to existing approaches.

AB - Collaborative Filtering, one of the most widely used approach in Recommender System, predicts a user's rating towards an item by aggregating ratings given by users having similar preference to that user. In existing approaches, user similarity is often computed on the whole set of items. However, because the number of items is often very large, and so is the diversity among items, users who have similar preference in one category of items may have totally different judgement on items of another kind. In order to deal with this problem, we propose a method of clustering items, so that inside a cluster, similarity between users does not change significantly. After that, when predicting rating of a user towards an item, we only aggregate ratings of users who have high similarity degree with that user inside the cluster to which that item belongs. Experiments evaluating our approach are carried out on the real dataset taken from movies recommendation system of MovieLens web site. Preliminary results suggest that our approach can improve prediction accuracy compared to existing approaches.

UR - http://www.scopus.com/inward/record.url?scp=38849088174&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38849088174&partnerID=8YFLogxK

U2 - 10.1109/CIMCA.2006.123

DO - 10.1109/CIMCA.2006.123

M3 - Conference contribution

AN - SCOPUS:38849088174

SN - 0769527310

SN - 9780769527314

BT - CIMCA 2006

ER -