Collaborative List-and-Pairwise Filtering from Implicit Feedback

Rong, Yu; Liu, Qi; Ye, Yuyang; Cheng, Mingyue; Chen, Enhong; Ma, Jinghong

doi:10.1109/tkde.2020.3016732

Cited by 21 publications

(8 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because CLiMF focuses on the first relevant item of recommendation lists. This situation also appears in [18,50]. Meanwhile, Deep-SetRank outperforms the pairwise learning and listwise learning method, this is also consistent with the results in Deep-SetRank [38].…”

Section: Experiments 51 Experimental Setupsupporting

confidence: 85%

“…SetRank provides a new research perspective for listwise learning for implicit feedback, and achieves state-of-the-art ranking results. Another direction is directly maximizing the ranking metrics [1,21,33,50]. Since most of the ranking metrics are not differentiable, existing models approximate the ranking metrics and optimize a smoothed version, such as deriving the lower bound.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Set2setRank: Collaborative Set to Set Ranking for Implicit Feedback based Recommendation

Chen,

Wu,

Zhang

et al. 2021

Preprint

View full text Add to dashboard Cite

As users often express their preferences with binary behavior data (implicit feedback), such as clicking items or buying products, implicit feedback based Collaborative Filtering (CF) models predict the top ranked items a user might like by leveraging implicit user-item interaction data. For each user, the implicit feedback is divided into two sets: an observed item set with limited observed behaviors, and a large unobserved item set that is mixed with negative item behaviors and unknown behaviors. Given any user preference prediction model, researchers either designed ranking based optimization goals or relied on negative item mining techniques for better optimization. Despite the performance gain of these implicit feedback based models, the recommendation results are still far from satisfactory due to the sparsity of the observed item set for each user. To this end, in this paper, we explore the unique characteristics of the implicit feedback and propose Set2setRank framework for recommendation. The optimization criteria of Set2setRank are two folds: First, we design an item to an item set comparison that encourages each observed item from the sampled observed set is ranked higher than any unobserved item from the sampled unobserved set. Second, we model set level comparison that encourages a margin between the distance summarized from the observed item set and the most "hard" unobserved item from the sampled negative set. Further, an adaptive sampling technique is designed to implement these two goals. We have to note that our proposed framework is model-agnostic and can be easily applied to most recommendation prediction approaches, and is time efficient in practice. Finally, extensive experiments on three real-world datasets demonstrate the superiority of our proposed approach.

show abstract

Section: Experiments 51 Experimental Setupsupporting

confidence: 85%

Section: Related Workmentioning

confidence: 99%

Set2setRank: Collaborative Set to Set Ranking for Implicit Feedback based Recommendation

Chen,

Wu,

Zhang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…From the view of technology, collaborative filtering has been a widely used technique in recommender systems. It generates recommendations by leveraging the user-item interactions derived from historical data [44]. At the same time, the collaborative activity requires a structured and open environment where individuals can actively exchange and elaborate their ideas to achieve a high-quality problem-solving solution.…”

Section: Reciprocal Preference Evaluation Methodsmentioning

confidence: 99%

Construction and Application of Cloud Computing Model for Reciprocal and燙ollaborative Knowledge Management

Li¹,

Bian²,

Guan³

et al. 2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

Promoting the co-constructing and sharing of organizational knowledge and improving organizational performance have always been the core research subject of knowledge management. Existing research focuses on the construction of knowledge management systems and knowledge sharing and transfer mechanisms. With the rapid development and application of cloud computing and big data technology, knowledge management is faced with many problems, such as how to combine with the new generation of information technology, how to achieve integration with organizational business processes, and so on. To solve such problems, this paper proposes a reciprocal collaborative knowledge management model (RCKM model) based on cloud computing technology, reciprocity theory, and collaboration technology. RCKM model includes project group management and cloud computing technology, which can realize management, finance, communication, and quality assurance of multiple projects and solve the problem of business integration with knowledge management. This paper designs evaluation methods of tacit knowledge and reciprocity preference based on the Bayesian formula and analyzes their effect with simulation data. The methods can provide quantitative support for the integration of knowledge management and business management to realize reciprocity and collaboration in the RCKM model. The research found that RCKM model can fully use cloud computing technology to promote the integration of knowledge management and organizational business, and the evaluation method based on the Bayesian formula can provide relatively accurate data support for the evaluation and selection of project team members.

show abstract

“…We use AUC (Area Under ROC Curve) and Logloss as the evaluation metrics [2,12,17,22,26], which are widely used to assess the performance of a CTR model.…”

Section: Evaluation Metricsmentioning

confidence: 99%

Deep User Match Network for Click-Through Rate Prediction

Huang

Tao

Zhang

2021

Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval

View full text Add to dashboard Cite

Click-through rate (CTR) prediction is a crucial task in many applications (e.g. recommender systems). Recently deep learning based models have been proposed and successfully applied for CTR prediction by focusing on feature interaction or user interest based on the item-to-item relevance between user behaviors and candidate item. However, these existing models neglect the user-to-user relevance between the target user and those who like the candidate item, which can reflect the preference of target user. To this end, in this paper, we propose a novel Deep User Match Network (DUMN) which measures the user-to-user relevance for CTR prediction. Specifically, in DUMN, we design a User Representation Layer to learn a unified user representation which contains user latent interest based on user behaviors. Then, User Match Layer is designed to measure the user-to-user relevance by matching the target user and those who have interacted with candidate item and modeling their similarities in user representation space. Extensive experimental results on three public real-world datasets validate the effectiveness of DUMN compared with state-of-the-art methods.

show abstract

Collaborative List-and-Pairwise Filtering from Implicit Feedback

Cited by 21 publications

References 51 publications

Set2setRank: Collaborative Set to Set Ranking for Implicit Feedback based Recommendation

Set2setRank: Collaborative Set to Set Ranking for Implicit Feedback based Recommendation

Construction and Application of Cloud Computing Model for Reciprocal and燙ollaborative Knowledge Management

Deep User Match Network for Click-Through Rate Prediction

Contact Info

Product

Resources

About