Clustered monotone transforms for rating factorization

Gaurush Hiranandani, Oluwasanmi Oluseye Koyejo, Raghav Somani, Sreangsu Acharyya

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

Abstract

Exploiting low-rank structure of the user-item rating matrix has been the crux of many recommendation engines. However, existing recommendation engines force raters with heterogeneous behavior profiles to map their intrinsic rating scales to a common rating scale (e.g. 1-5). This non-linear transformation of the rating scale shatters the low-rank structure of the rating matrix, therefore resulting in a poor fit and consequentially, poor recommendations. In this paper, we propose Clustered Monotone Transforms for Rating Factorization (CMTRF), a novel approach to perform regression up to unknown monotonic transforms over unknown population segments. Essentially, for recommendation systems, the technique searches for monotonic transformations of the rating scales resulting in a better fit. This is combined with an underlying matrix factorization regression model that couples the user-wise ratings to exploit shared low dimensional structure. The rating scale transformations can be generated for each user, for a cluster of users, or for all the users at once, forming the basis of three simple and efficient algorithms proposed in this paper, all of which alternate between transformation of the rating scales and matrix factorization regression. Despite the non-convexity, CMTRF is theoretically shown to recover a unique solution under mild conditions. Experimental results on two synthetic and seven real-world datasets show that CMTRF outperforms other state-of-the-art baselines.

Original languageEnglish (US)
Title of host publicationWSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining
PublisherAssociation for Computing Machinery, Inc
Pages132-140
Number of pages9
ISBN (Electronic)9781450359405
DOIs
StatePublished - Jan 30 2019
Event12th ACM International Conference on Web Search and Data Mining, WSDM 2019 - Melbourne, Australia
Duration: Feb 11 2019Feb 15 2019

Publication series

NameWSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining

Conference

Conference12th ACM International Conference on Web Search and Data Mining, WSDM 2019
CountryAustralia
CityMelbourne
Period2/11/192/15/19

Fingerprint

Factorization
Recommender systems
Mathematical transformations

Keywords

  • Bregman divergences
  • Clustering
  • Collaborative filtering
  • Matrix factorization
  • Monotonic transformations
  • Regression

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Software
  • Computer Science Applications

Cite this

Hiranandani, G., Koyejo, O. O., Somani, R., & Acharyya, S. (2019). Clustered monotone transforms for rating factorization. In WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining (pp. 132-140). (WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining). Association for Computing Machinery, Inc. https://doi.org/10.1145/3289600.3291005

Clustered monotone transforms for rating factorization. / Hiranandani, Gaurush; Koyejo, Oluwasanmi Oluseye; Somani, Raghav; Acharyya, Sreangsu.

WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc, 2019. p. 132-140 (WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining).

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

Hiranandani, G, Koyejo, OO, Somani, R & Acharyya, S 2019, Clustered monotone transforms for rating factorization. in WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining, Association for Computing Machinery, Inc, pp. 132-140, 12th ACM International Conference on Web Search and Data Mining, WSDM 2019, Melbourne, Australia, 2/11/19. https://doi.org/10.1145/3289600.3291005
Hiranandani G, Koyejo OO, Somani R, Acharyya S. Clustered monotone transforms for rating factorization. In WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc. 2019. p. 132-140. (WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining). https://doi.org/10.1145/3289600.3291005
Hiranandani, Gaurush ; Koyejo, Oluwasanmi Oluseye ; Somani, Raghav ; Acharyya, Sreangsu. / Clustered monotone transforms for rating factorization. WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc, 2019. pp. 132-140 (WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining).
@inproceedings{73304fabf2044ab2b76480d6e0e86b86,
title = "Clustered monotone transforms for rating factorization",
abstract = "Exploiting low-rank structure of the user-item rating matrix has been the crux of many recommendation engines. However, existing recommendation engines force raters with heterogeneous behavior profiles to map their intrinsic rating scales to a common rating scale (e.g. 1-5). This non-linear transformation of the rating scale shatters the low-rank structure of the rating matrix, therefore resulting in a poor fit and consequentially, poor recommendations. In this paper, we propose Clustered Monotone Transforms for Rating Factorization (CMTRF), a novel approach to perform regression up to unknown monotonic transforms over unknown population segments. Essentially, for recommendation systems, the technique searches for monotonic transformations of the rating scales resulting in a better fit. This is combined with an underlying matrix factorization regression model that couples the user-wise ratings to exploit shared low dimensional structure. The rating scale transformations can be generated for each user, for a cluster of users, or for all the users at once, forming the basis of three simple and efficient algorithms proposed in this paper, all of which alternate between transformation of the rating scales and matrix factorization regression. Despite the non-convexity, CMTRF is theoretically shown to recover a unique solution under mild conditions. Experimental results on two synthetic and seven real-world datasets show that CMTRF outperforms other state-of-the-art baselines.",
keywords = "Bregman divergences, Clustering, Collaborative filtering, Matrix factorization, Monotonic transformations, Regression",
author = "Gaurush Hiranandani and Koyejo, {Oluwasanmi Oluseye} and Raghav Somani and Sreangsu Acharyya",
year = "2019",
month = "1",
day = "30",
doi = "10.1145/3289600.3291005",
language = "English (US)",
series = "WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining",
publisher = "Association for Computing Machinery, Inc",
pages = "132--140",
booktitle = "WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining",

}

TY - GEN

T1 - Clustered monotone transforms for rating factorization

AU - Hiranandani, Gaurush

AU - Koyejo, Oluwasanmi Oluseye

AU - Somani, Raghav

AU - Acharyya, Sreangsu

PY - 2019/1/30

Y1 - 2019/1/30

N2 - Exploiting low-rank structure of the user-item rating matrix has been the crux of many recommendation engines. However, existing recommendation engines force raters with heterogeneous behavior profiles to map their intrinsic rating scales to a common rating scale (e.g. 1-5). This non-linear transformation of the rating scale shatters the low-rank structure of the rating matrix, therefore resulting in a poor fit and consequentially, poor recommendations. In this paper, we propose Clustered Monotone Transforms for Rating Factorization (CMTRF), a novel approach to perform regression up to unknown monotonic transforms over unknown population segments. Essentially, for recommendation systems, the technique searches for monotonic transformations of the rating scales resulting in a better fit. This is combined with an underlying matrix factorization regression model that couples the user-wise ratings to exploit shared low dimensional structure. The rating scale transformations can be generated for each user, for a cluster of users, or for all the users at once, forming the basis of three simple and efficient algorithms proposed in this paper, all of which alternate between transformation of the rating scales and matrix factorization regression. Despite the non-convexity, CMTRF is theoretically shown to recover a unique solution under mild conditions. Experimental results on two synthetic and seven real-world datasets show that CMTRF outperforms other state-of-the-art baselines.

AB - Exploiting low-rank structure of the user-item rating matrix has been the crux of many recommendation engines. However, existing recommendation engines force raters with heterogeneous behavior profiles to map their intrinsic rating scales to a common rating scale (e.g. 1-5). This non-linear transformation of the rating scale shatters the low-rank structure of the rating matrix, therefore resulting in a poor fit and consequentially, poor recommendations. In this paper, we propose Clustered Monotone Transforms for Rating Factorization (CMTRF), a novel approach to perform regression up to unknown monotonic transforms over unknown population segments. Essentially, for recommendation systems, the technique searches for monotonic transformations of the rating scales resulting in a better fit. This is combined with an underlying matrix factorization regression model that couples the user-wise ratings to exploit shared low dimensional structure. The rating scale transformations can be generated for each user, for a cluster of users, or for all the users at once, forming the basis of three simple and efficient algorithms proposed in this paper, all of which alternate between transformation of the rating scales and matrix factorization regression. Despite the non-convexity, CMTRF is theoretically shown to recover a unique solution under mild conditions. Experimental results on two synthetic and seven real-world datasets show that CMTRF outperforms other state-of-the-art baselines.

KW - Bregman divergences

KW - Clustering

KW - Collaborative filtering

KW - Matrix factorization

KW - Monotonic transformations

KW - Regression

UR - http://www.scopus.com/inward/record.url?scp=85061735660&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061735660&partnerID=8YFLogxK

U2 - 10.1145/3289600.3291005

DO - 10.1145/3289600.3291005

M3 - Conference contribution

AN - SCOPUS:85061735660

T3 - WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining

SP - 132

EP - 140

BT - WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining

PB - Association for Computing Machinery, Inc

ER -