Variational low rank multinomials for collaborative filtering with side-information

Elahi, Ehtsham; Wang, Wei; Ray, Dave; Fenton, Aish; Jebara, Tony

doi:10.1145/3298689.3347036

Cited by 10 publications

(9 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Scalability of model training and prediction/inference: Scaling deep‐learning model‐training not only depends on a specialized software/hardware stack but also on the characteristics of the datasets. For example, consider multiclass classification, which is a very important task in recommendation systems (Elahi et al 2019; Liang et al 2018). Multiclass classification is usually done by using a softmax activation in the output‐layer of deep‐learning models.…”

Section: Practical Challengesmentioning

confidence: 99%

Deep learning for recommender systems: A Netflix case study

et al. 2021

Self Cite

View full text Add to dashboard Cite

Deep learning has profoundly impacted many areas of machine learning. However, it took a while for its impact to be felt in the field of recommender systems. In this article, we outline some of the challenges encountered and lessons learned in using deep learning for recommender systems at Netflix. We first provide an overview of the various recommendation tasks on the Netflix service. We found that different model architectures excel at different tasks. Even though many deep-learning models can be understood as extensions of existing (simple) recommendation algorithms, we initially did not observe significant improvements in performance over well-tuned non-deep-learning approaches. Only when we added numerous features of heterogeneous types to the input data, deep-learning models did start to shine in our setting. We also observed that deep-learning methods can exacerbate the problem of offline–online metric (mis-)alignment. After addressing these challenges, deep learning has ultimately resulted in large improvements to our recommendations as measured by both offline and online metrics. On the practical side, integrating deep-learning toolboxes in our system has made it faster and easier to implement and experiment with both deep-learning and non-deep-learning approaches for various recommendation tasks. We conclude this article by summarizing our take-aways that may generalize to other applications beyond Netflix.

show abstract

Section: Practical Challengesmentioning

confidence: 99%

Deep learning for recommender systems: A Netflix case study

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…We present additive and collective ease r (add-ease r and cease r ), and show how these novel methods retain a closed-form solution whilst leveraging signals embedded in side-information to generate more effective recommendations. We show how these straightforward and complementary extensions of the ease r paradigm consistently outperform state-of-the-art approaches such as cvae [3] and vlm [9]. Additionally, we empirically validate that add-ease r and cease r are indeed able to soften the effect of the long tail, and are more likely to recommend different and less popular items than plain ease r .…”

Section: Introductionmentioning

confidence: 65%

“…Several hurdles for recommender systems remain, such as the łlong tailž (very few items account for the large majority of interactions) and łcold startž (new items do not have any interactions) issues [22,27,30]. It has become common practice to exploit item side-information or metadata to try and alleviate these problems, and several recent works show that they indeed succeed at this [3,9]. In this work, we study the applicability of ease r -like models in the presence of such metadata.…”

Section: Introductionmentioning

confidence: 99%

Closed-Form Models for Collaborative Filtering with Side-Information

Jeunen

Balen

Goethals

2020

Fourteenth ACM Conference on Recommender Systems

View full text Add to dashboard Cite

Recent work has shown that, despite their simplicity, item-based models optimised through ridge regression can attain highly competitive results on collaborative filtering tasks. As these models are analytically computable and thus forgo the need for often expensive iterative optimisation procedures, they are an attractive choice for practitioners. We study the applicability of such closedform models to implicit-feedback collaborative filtering when additional side-information or metadata about items is available. Two complementary extensions to the ease r paradigm are proposed, based on collective and additive models. Through an extensive empirical analysis on several large-scale datasets, we show that our methods can effectively exploit side-information whilst retaining a closed-form solution, and improve upon the state-of-the-art without increasing the computational complexity of the original ease r approach. Additionally, empirical results demonstrate that the use of side-information leads to more łlong tailž items being recommended, benefiting the recommendations' coverage of the item catalogue.

show abstract

“…We also ind that some papers [123,132] only perform the iltering in an unsymmetrical way, e.g., only for users or items. Besides, several papers [27,34] set a relatively large value for n (e.g., n = 25, 30), which can derive a more dense dataset. However, it might signiicantly change the overall distributions of user-item interactions of the original dataset.…”

Section: Dataset Selection and Preprocessingmentioning

confidence: 99%

A Revisiting Study of Appropriate Offline Evaluation for Top- N Recommendation Algorithms

Zhao

Lin

Feng

et al. 2022

ACM Trans. Inf. Syst.

View full text Add to dashboard Cite

In recommender system, top- N recommendation is an important task with implicit feedback data. Although the recent success of deep learning largely pushes forward the research on top- N recommendation, there are increasing concerns on appropriate evaluation of recommendation algorithms. It becomes emergent to study how recommendation algorithms can be reliably evaluated and thoroughly verified. This work presents a large-scale, systematic study on six important factors from three aspects for evaluating recommender systems. We carefully select twelve top- N recommendation algorithms and eight recommendation datasets. Our experiments are carefully designed and extensively conducted with these algorithms and datasets. In particular, all the experiments in our work are implemented based on an open-sourced recommendation library, Recbole [139], which ensures the reproducibility and reliability of our results. Based on the large-scale experiments and detailed analysis, we derive several key findings on the experimental settings for evaluating recommender systems. Our findings show that some settings can lead to substantial or significant differences in performance ranking of the compared algorithms. In response to recent evaluation concerns, we also provide several suggested settings that are specially important for performance comparison.

show abstract

Variational low rank multinomials for collaborative filtering with side-information

Cited by 10 publications

References 12 publications

Deep learning for recommender systems: A Netflix case study

Deep learning for recommender systems: A Netflix case study

Closed-Form Models for Collaborative Filtering with Side-Information

A Revisiting Study of Appropriate Offline Evaluation for Top- N Recommendation Algorithms

Contact Info

Product

Resources

About