Improving pairwise learning for item recommendation from implicit feedback

Rendle, Steffen; Freudenthaler, Christoph

doi:10.1145/2556195.2556248

Cited by 350 publications

(283 citation statements)

References 17 publications

Supporting

Mentioning

276

Contrasting

Order By: Relevance

“…For example, in One past study that has observed the di culty of sampling from the huge number of entries is [24]. In a ranking setting they show that SG converges slowly by uniform sampling.…”

Section: Stochastic Gradient (Sg)mentioning

confidence: 99%

Selection of Negative Samples for One-class Matrix Factorization

Yu¹,

Bilenko²

2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

Many recommender systems have only implicit user feedback. The two possible ratings are positive and negative, but only part of positive entries are observed. One-class matrix factorization (MF) is a popular approach for such scenarios by treating some missing entries as negative. Two major ways to select negative entries are by sub-sampling a set with similar size to that of observed positive entries or by including all missing entries as negative. They are referred to as "subsampled" and "full" approaches in this work, respectively. Currently detailed comparisons between these two selection schemes on large-scale data are still lacking. One important reason is that the "full" approach leads to a hard optimization problem after treating all missing entries as negative. In this paper, we successfully develop e cient optimization techniques to solve this challenging problem so that the "full" approach becomes practically viable. We then compare in detail the two approaches "subsampled" and "full" for selecting negative entries. Results show that the "full" approach of including much more missing entries as negative yields better results.

show abstract

Section: Stochastic Gradient (Sg)mentioning

confidence: 99%

Selection of Negative Samples for One-class Matrix Factorization

Yu¹,

Bilenko²

2017

Proceedings of the 2017 SIAM International Conference on Data Mining

View full text Add to dashboard Cite

show abstract

“…It shows that the algorithm generally converges within 10 7 iterations. Moreover, it is worth noting that the learning process can be speeded up by adopting more efficient sampling strategies [Rendle and Freudenthaler 2014]. In sum, the above analysis verifies that the training computations are tractable and able to scale up for large-scale datasets.…”

Section: Efficiency Analysismentioning

confidence: 64%

Augmented Collaborative Filtering for Sparseness Reduction in Personalized POI Recommendation

Cui

Shen

Nie

et al. 2017

ACM Trans. Intell. Syst. Technol.

View full text Add to dashboard Cite

As mobile device penetration increases, it has become pervasive for images to be associated with locations in the form of geotags. Geotags bridge the gap between the physical world and the cyberspace, giving rise to new opportunities to extract further insights into user preferences and behaviors. In this article, we aim to exploit geotagged photos from online photo-sharing sites for the purpose of personalized Point-of-Interest (POI) recommendation. Owing to the fact that most users have only very limited travel experiences, data sparseness poses a formidable challenge to personalized POI recommendation. To alleviate data sparseness, we propose to augment current collaborative filtering algorithms along from multiple perspectives. Specifically, hybrid preference cues comprising user-uploaded and user-favored photos are harvested to study users' tastes. Moreover, heterogeneous high-order relationship information is jointly captured from user social networks and POI multimodal contents with hypergraph models. We also build upon the matrix factorization algorithm to integrate the disparate sources of preference and relationship information, and apply our approach to directly optimize user preference rankings. Extensive experiments on a large and publicly accessible dataset well verified the potential of our approach for addressing data sparseness and offering quality recommendations to users, especially for those who have only limited travel experiences.

show abstract

“…On the other hand, both approaches are originally designed for the rating prediction task [6], which are based on explicit user feedback. However, in most real-world scenarios, only implicit user behavior is observed and there is no explicit rating [22,23]. Besides, the goal of item recommendation is preferred as a ranking task rather than a rating prediction one.…”

Section: Related Workmentioning

confidence: 99%

“…Fidelity loss (FL): This loss is introduced by Tsai et al [15] and has been applied in Information Retrieval (IR) task and yielded superior performance. The original function regarding the loss of pairs is defined as (22) where P ij and Pij share the same meanings with the CE loss in Eq. (1).…”

Section: Lambda With Alternative Lossesmentioning

confidence: 99%

LambdaFM

Yuan

Guo

Jose

et al. 2016

Proceedings of the 25th ACM International on Conference on Information and Knowledge Management

View full text Add to dashboard Cite

State-of-the-art item recommendation algorithms, which apply Factorization Machines (FM) as a scoring function and pairwise ranking loss as a trainer (PRFM for short), have been recently investigated for the implicit feedback based context-aware recommendation problem (IFCAR). However, good recommenders particularly emphasize on the accuracy near the top of the ranked list, and typical pairwise loss functions might not match well with such a requirement. In this paper, we demonstrate, both theoretically and empirically, PRFM models usually lead to non-optimal item recommendation results due to such a mismatch. Inspired by the success of LambdaRank, we introduce Lambda Factorization Machines (LambdaFM), which is particularly intended for optimizing ranking performance for IFCAR. We also point out that the original lambda function suffers from the issue of expensive computational complexity in such settings due to a large amount of unobserved feedback. Hence, instead of directly adopting the original lambda strategy, we create three effective lambda surrogates by conducting a theoretical analysis for lambda from the top-N optimization perspective. Further, we prove that the proposed lambda surrogates are generic and applicable to a large set of pairwise ranking loss functions. Experimental results demonstrate LambdaFM significantly outperforms state-of-the-art algorithms on three real-world datasets in terms of four standard ranking measures.

show abstract

Improving pairwise learning for item recommendation from implicit feedback

Cited by 350 publications

References 17 publications

Selection of Negative Samples for One-class Matrix Factorization

Selection of Negative Samples for One-class Matrix Factorization

Augmented Collaborative Filtering for Sparseness Reduction in Personalized POI Recommendation

LambdaFM

Contact Info

Product

Resources

About