Generating Pseudotransactions for Improving Sparse Matrix Factorization

Agung Toto Wibowo

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

2 Citations (Scopus)

Abstract

Recent research on Recommender Systems, specifically Collaborative Filtering, has focussed on Matrix Factorization (MF) methods, which have been shown to provide good solutions to the cold start problem. However, typically the same settings are used for Matrix factorization regardless of the density of the matrix. In our experiments, we found that for MF, Root Mean Square Error (RMSE) for recommendations increases (i.e. performance drops) for sparse matrices. We propose a Two Stage MF approach so MF is run twice over the whole matrix; the first stage uses MF to generate a small percentage of pseudotransactions that are added to the original matrix to increase its density, and the second stage re-runs MF over this denser matrix to predict the user-item transactions in the testing set. We show using data from Movielens that such methods can improve on the performance of MF for sparse martrices.
Original languageEnglish
Title of host publicationRecSys '16
Subtitle of host publicationProceedings of the 10th ACM Conference on Recommender Systems
Place of PublicationBoston, Massachusetts, USA
PublisherACM
Pages439-442
Number of pages4
ISBN (Electronic)978-1-4503-4035-9
DOIs
Publication statusPublished - 15 Sep 2016

Fingerprint

Factorization
Collaborative filtering
Recommender systems
Mean square error

Cite this

Wibowo, A. T. (2016). Generating Pseudotransactions for Improving Sparse Matrix Factorization. In RecSys '16 : Proceedings of the 10th ACM Conference on Recommender Systems (pp. 439-442). Boston, Massachusetts, USA: ACM. https://doi.org/10.1145/2959100.2959107

Generating Pseudotransactions for Improving Sparse Matrix Factorization. / Wibowo, Agung Toto.

RecSys '16 : Proceedings of the 10th ACM Conference on Recommender Systems. Boston, Massachusetts, USA : ACM, 2016. p. 439-442.

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

Wibowo, AT 2016, Generating Pseudotransactions for Improving Sparse Matrix Factorization. in RecSys '16 : Proceedings of the 10th ACM Conference on Recommender Systems. ACM, Boston, Massachusetts, USA, pp. 439-442. https://doi.org/10.1145/2959100.2959107
Wibowo AT. Generating Pseudotransactions for Improving Sparse Matrix Factorization. In RecSys '16 : Proceedings of the 10th ACM Conference on Recommender Systems. Boston, Massachusetts, USA: ACM. 2016. p. 439-442 https://doi.org/10.1145/2959100.2959107
Wibowo, Agung Toto. / Generating Pseudotransactions for Improving Sparse Matrix Factorization. RecSys '16 : Proceedings of the 10th ACM Conference on Recommender Systems. Boston, Massachusetts, USA : ACM, 2016. pp. 439-442
@inproceedings{adf6c566deb74c1da10e31e24a3e6e89,
title = "Generating Pseudotransactions for Improving Sparse Matrix Factorization",
abstract = "Recent research on Recommender Systems, specifically Collaborative Filtering, has focussed on Matrix Factorization (MF) methods, which have been shown to provide good solutions to the cold start problem. However, typically the same settings are used for Matrix factorization regardless of the density of the matrix. In our experiments, we found that for MF, Root Mean Square Error (RMSE) for recommendations increases (i.e. performance drops) for sparse matrices. We propose a Two Stage MF approach so MF is run twice over the whole matrix; the first stage uses MF to generate a small percentage of pseudotransactions that are added to the original matrix to increase its density, and the second stage re-runs MF over this denser matrix to predict the user-item transactions in the testing set. We show using data from Movielens that such methods can improve on the performance of MF for sparse martrices.",
author = "Wibowo, {Agung Toto}",
year = "2016",
month = "9",
day = "15",
doi = "10.1145/2959100.2959107",
language = "English",
pages = "439--442",
booktitle = "RecSys '16",
publisher = "ACM",

}

TY - GEN

T1 - Generating Pseudotransactions for Improving Sparse Matrix Factorization

AU - Wibowo, Agung Toto

PY - 2016/9/15

Y1 - 2016/9/15

N2 - Recent research on Recommender Systems, specifically Collaborative Filtering, has focussed on Matrix Factorization (MF) methods, which have been shown to provide good solutions to the cold start problem. However, typically the same settings are used for Matrix factorization regardless of the density of the matrix. In our experiments, we found that for MF, Root Mean Square Error (RMSE) for recommendations increases (i.e. performance drops) for sparse matrices. We propose a Two Stage MF approach so MF is run twice over the whole matrix; the first stage uses MF to generate a small percentage of pseudotransactions that are added to the original matrix to increase its density, and the second stage re-runs MF over this denser matrix to predict the user-item transactions in the testing set. We show using data from Movielens that such methods can improve on the performance of MF for sparse martrices.

AB - Recent research on Recommender Systems, specifically Collaborative Filtering, has focussed on Matrix Factorization (MF) methods, which have been shown to provide good solutions to the cold start problem. However, typically the same settings are used for Matrix factorization regardless of the density of the matrix. In our experiments, we found that for MF, Root Mean Square Error (RMSE) for recommendations increases (i.e. performance drops) for sparse matrices. We propose a Two Stage MF approach so MF is run twice over the whole matrix; the first stage uses MF to generate a small percentage of pseudotransactions that are added to the original matrix to increase its density, and the second stage re-runs MF over this denser matrix to predict the user-item transactions in the testing set. We show using data from Movielens that such methods can improve on the performance of MF for sparse martrices.

U2 - 10.1145/2959100.2959107

DO - 10.1145/2959100.2959107

M3 - Conference contribution

SP - 439

EP - 442

BT - RecSys '16

PB - ACM

CY - Boston, Massachusetts, USA

ER -