Multi-task Learning for Recommender System

Xia Ning, George Karypis
; Proceedings of 2nd Asian Conference on Machine Learning, JMLR Workshop and Conference Proceedings 13:269-284, 2010.

Abstract

This paper focuses on exploring personalized multi-task learning approaches for collaborative filtering towards the goal of improving the prediction performance of rating prediction systems. These methods first specifically identify a set of users that are closely related to the user under consideration (i.e., active user), and then learn multiple rating prediction models simultaneously, one for the active user and one for each of the related users. Such learning for multiple models (tasks) in parallel is implemented by representing all learning instances (users and items) using a coupled user-item representation, and within errorinsensitive Support Vector Regression (e-SVR) framework applying multi-task kernel tricks. A comprehensive set of experiments shows that multi-task learning approaches lead to significant performance improvement over conventional alternatives.

Cite this Paper


BibTeX
@InProceedings{pmlr-v13-ning10a, title = {Multi-task Learning for Recommender System}, author = {Xia Ning and George Karypis}, pages = {269--284}, year = {2010}, editor = {Masashi Sugiyama and Qiang Yang}, volume = {13}, series = {Proceedings of Machine Learning Research}, address = {Tokyo, Japan}, month = {08--10 Nov}, publisher = {JMLR Workshop and Conference Proceedings}, pdf = {http://proceedings.mlr.press/v13/ning10a/ning10a.pdf}, url = {http://proceedings.mlr.press/v13/ning10a.html}, abstract = {This paper focuses on exploring personalized multi-task learning approaches for collaborative filtering towards the goal of improving the prediction performance of rating prediction systems. These methods first specifically identify a set of users that are closely related to the user under consideration (i.e., active user), and then learn multiple rating prediction models simultaneously, one for the active user and one for each of the related users. Such learning for multiple models (tasks) in parallel is implemented by representing all learning instances (users and items) using a coupled user-item representation, and within errorinsensitive Support Vector Regression (e-SVR) framework applying multi-task kernel tricks. A comprehensive set of experiments shows that multi-task learning approaches lead to significant performance improvement over conventional alternatives.} }
Endnote
%0 Conference Paper %T Multi-task Learning for Recommender System %A Xia Ning %A George Karypis %B Proceedings of 2nd Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2010 %E Masashi Sugiyama %E Qiang Yang %F pmlr-v13-ning10a %I PMLR %J Proceedings of Machine Learning Research %P 269--284 %U http://proceedings.mlr.press %V 13 %W PMLR %X This paper focuses on exploring personalized multi-task learning approaches for collaborative filtering towards the goal of improving the prediction performance of rating prediction systems. These methods first specifically identify a set of users that are closely related to the user under consideration (i.e., active user), and then learn multiple rating prediction models simultaneously, one for the active user and one for each of the related users. Such learning for multiple models (tasks) in parallel is implemented by representing all learning instances (users and items) using a coupled user-item representation, and within errorinsensitive Support Vector Regression (e-SVR) framework applying multi-task kernel tricks. A comprehensive set of experiments shows that multi-task learning approaches lead to significant performance improvement over conventional alternatives.
RIS
TY - CPAPER TI - Multi-task Learning for Recommender System AU - Xia Ning AU - George Karypis BT - Proceedings of 2nd Asian Conference on Machine Learning PY - 2010/10/31 DA - 2010/10/31 ED - Masashi Sugiyama ED - Qiang Yang ID - pmlr-v13-ning10a PB - PMLR SP - 269 DP - PMLR EP - 284 L1 - http://proceedings.mlr.press/v13/ning10a/ning10a.pdf UR - http://proceedings.mlr.press/v13/ning10a.html AB - This paper focuses on exploring personalized multi-task learning approaches for collaborative filtering towards the goal of improving the prediction performance of rating prediction systems. These methods first specifically identify a set of users that are closely related to the user under consideration (i.e., active user), and then learn multiple rating prediction models simultaneously, one for the active user and one for each of the related users. Such learning for multiple models (tasks) in parallel is implemented by representing all learning instances (users and items) using a coupled user-item representation, and within errorinsensitive Support Vector Regression (e-SVR) framework applying multi-task kernel tricks. A comprehensive set of experiments shows that multi-task learning approaches lead to significant performance improvement over conventional alternatives. ER -
APA
Ning, X. & Karypis, G.. (2010). Multi-task Learning for Recommender System. Proceedings of 2nd Asian Conference on Machine Learning, in PMLR 13:269-284

Related Material