Computational complexity reduction for factorization-based collaborative filtering algorithms

István Pilászy, D. Tikk

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

2 Citations (Scopus)

Abstract

Alternating least squares (ALS) is a powerful matrix factorization (MF) algorithm for both implicit and explicit feedback based recommender systems. We show that by using the Sherman-Morrison formula (SMF), we can reduce the computational complexity of several ALS based algorithms. It also reduces the complexity of greedy forward and backward feature selection algorithms by an order of magnitude. We propose linear kernel ridge regression (KRR) for users with few ratings. We show that both SMF and KRR can efficiently handle new ratings.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages229-239
Number of pages11
Volume5692 LNCS
DOIs
Publication statusPublished - 2009
Event10th International Conference on E-Commerce and Web Technologies, EC-Web 2009 - Linz, Austria
Duration: Sep 1 2009Sep 4 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5692 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other10th International Conference on E-Commerce and Web Technologies, EC-Web 2009
CountryAustria
CityLinz
Period9/1/099/4/09

Fingerprint

Collaborative filtering
Collaborative Filtering
Factorization
Sherman-Morrison Formula
Alternating Least Squares
Computational complexity
Kernel Regression
Ridge Regression
Computational Complexity
Matrix Factorization
Recommender Systems
Recommender systems
Feature Selection
Feature extraction
Feedback

Keywords

  • Alternating least squares
  • Collaborative filtering
  • Greedy feature selection
  • Kernel ridge regression
  • Matrix factorization
  • Sherman-Morrison formula

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Pilászy, I., & Tikk, D. (2009). Computational complexity reduction for factorization-based collaborative filtering algorithms. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5692 LNCS, pp. 229-239). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5692 LNCS). https://doi.org/10.1007/978-3-642-03964-5-22

Computational complexity reduction for factorization-based collaborative filtering algorithms. / Pilászy, István; Tikk, D.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5692 LNCS 2009. p. 229-239 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5692 LNCS).

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

Pilászy, I & Tikk, D 2009, Computational complexity reduction for factorization-based collaborative filtering algorithms. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 5692 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5692 LNCS, pp. 229-239, 10th International Conference on E-Commerce and Web Technologies, EC-Web 2009, Linz, Austria, 9/1/09. https://doi.org/10.1007/978-3-642-03964-5-22
Pilászy I, Tikk D. Computational complexity reduction for factorization-based collaborative filtering algorithms. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5692 LNCS. 2009. p. 229-239. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-03964-5-22
Pilászy, István ; Tikk, D. / Computational complexity reduction for factorization-based collaborative filtering algorithms. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 5692 LNCS 2009. pp. 229-239 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{6217e2071f8d4742a4d8714240671a5e,
title = "Computational complexity reduction for factorization-based collaborative filtering algorithms",
abstract = "Alternating least squares (ALS) is a powerful matrix factorization (MF) algorithm for both implicit and explicit feedback based recommender systems. We show that by using the Sherman-Morrison formula (SMF), we can reduce the computational complexity of several ALS based algorithms. It also reduces the complexity of greedy forward and backward feature selection algorithms by an order of magnitude. We propose linear kernel ridge regression (KRR) for users with few ratings. We show that both SMF and KRR can efficiently handle new ratings.",
keywords = "Alternating least squares, Collaborative filtering, Greedy feature selection, Kernel ridge regression, Matrix factorization, Sherman-Morrison formula",
author = "Istv{\'a}n Pil{\'a}szy and D. Tikk",
year = "2009",
doi = "10.1007/978-3-642-03964-5-22",
language = "English",
isbn = "3642039634",
volume = "5692 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "229--239",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Computational complexity reduction for factorization-based collaborative filtering algorithms

AU - Pilászy, István

AU - Tikk, D.

PY - 2009

Y1 - 2009

N2 - Alternating least squares (ALS) is a powerful matrix factorization (MF) algorithm for both implicit and explicit feedback based recommender systems. We show that by using the Sherman-Morrison formula (SMF), we can reduce the computational complexity of several ALS based algorithms. It also reduces the complexity of greedy forward and backward feature selection algorithms by an order of magnitude. We propose linear kernel ridge regression (KRR) for users with few ratings. We show that both SMF and KRR can efficiently handle new ratings.

AB - Alternating least squares (ALS) is a powerful matrix factorization (MF) algorithm for both implicit and explicit feedback based recommender systems. We show that by using the Sherman-Morrison formula (SMF), we can reduce the computational complexity of several ALS based algorithms. It also reduces the complexity of greedy forward and backward feature selection algorithms by an order of magnitude. We propose linear kernel ridge regression (KRR) for users with few ratings. We show that both SMF and KRR can efficiently handle new ratings.

KW - Alternating least squares

KW - Collaborative filtering

KW - Greedy feature selection

KW - Kernel ridge regression

KW - Matrix factorization

KW - Sherman-Morrison formula

UR - http://www.scopus.com/inward/record.url?scp=70350489297&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350489297&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-03964-5-22

DO - 10.1007/978-3-642-03964-5-22

M3 - Conference contribution

AN - SCOPUS:70350489297

SN - 3642039634

SN - 9783642039638

VL - 5692 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 229

EP - 239

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -