Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Neighborhood-based Matrix Factorization Technique for Recommendation

A Neighborhood-based Matrix Factorization Technique for Recommendation The data sparsity and prediction quality are recognized as the key challenges in the existing recommender Systems. Most of the existing recommender systems depend on collaborating flitering (CF) method which mainly leverages the user-item rating matrix representing the relationship between users and items. However, the CF-based method sometimes fails to provide accurate information for predicting recommendations as there is an assumption that the relationship between attributes of items is independent and identically distributed. In real applications, there are often several kinds of coupling relationships or connections existed among users or items. In this paper, we incorporate the coupling relationship analysis to capture the under-discovered relationships between items and aim to make the ratings more reasonable. Next, we propose a neighborhood-based matrix factorization model, which considers both the explicit and implicit correlations between items, to suggest the more reasonable items to user. The experimental evaluations demonstrate that the proposed algorithms outperform the state-of-the-art algorithms in the warm- and cold-start settings. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Data Science Springer Journals

A Neighborhood-based Matrix Factorization Technique for Recommendation

Download PDF
16 pages

Loading next page...
 
/lp/springer-journals/a-neighborhood-based-matrix-factorization-technique-for-recommendation-6VzAe2qVmR
Publisher
Springer Journals
Copyright
Copyright © 2015 by Springer-Verlag Berlin Heidelberg
Subject
Business and Management; Business and Management, general; Statistics for Business/Economics/Mathematical Finance/Insurance; Computing Methodologies
ISSN
2198-5804
eISSN
2198-5812
DOI
10.1007/s40745-015-0056-6
Publisher site
See Article on Publisher Site

Abstract

The data sparsity and prediction quality are recognized as the key challenges in the existing recommender Systems. Most of the existing recommender systems depend on collaborating flitering (CF) method which mainly leverages the user-item rating matrix representing the relationship between users and items. However, the CF-based method sometimes fails to provide accurate information for predicting recommendations as there is an assumption that the relationship between attributes of items is independent and identically distributed. In real applications, there are often several kinds of coupling relationships or connections existed among users or items. In this paper, we incorporate the coupling relationship analysis to capture the under-discovered relationships between items and aim to make the ratings more reasonable. Next, we propose a neighborhood-based matrix factorization model, which considers both the explicit and implicit correlations between items, to suggest the more reasonable items to user. The experimental evaluations demonstrate that the proposed algorithms outperform the state-of-the-art algorithms in the warm- and cold-start settings.

Journal

Annals of Data Science – Springer Journals

Published: Dec 18, 2015

References

  • Amazon.com recommendations: item-to-item collaborative filtering
    Linden, G; Smith, B; York, J
  • Collaborative filtering with temporal dynamics
    Koren, Y
  • Roure: ontological user profiling in recommender systems
    Middleton, SE; Shadbolt, NR; De, DC
  • Fab: content-based collaborative filtering
    Balabanovic, M; Shoham, Y
  • FolkRank:a ranking algorithm for folksonomies
    Hotho, A; Jaschke, R; Schmitz, C; Stumme, G
  • Coupled behavior analysis with applications
    Cao, L; Ou, Y; Yu, PS
  • Imprecise shannon’s entropy and multi attribute decision making
    Lotfi, H; Fallahnejad, R
  • Content-boosted matrix factorization techniques for recommender systems
    Nguyen, JJ; Zhu, M