Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval 2020
DOI: 10.1145/3397271.3401083
|View full text |Cite
|
Sign up to set email alerts
|

A General Knowledge Distillation Framework for Counterfactual Recommendation via Uniform Data

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
71
0
1

Year Published

2021
2021
2023
2023

Publication Types

Select...
4
4
2

Relationship

1
9

Authors

Journals

citations
Cited by 151 publications
(72 citation statements)
references
References 23 publications
0
71
0
1
Order By: Relevance
“…Counterfactual learning offers a way to tackle the missing-not-at-random problem [1,2,5,15,21,23,26,29]. Several recent studies [9,31] employ counterfactual learning, such as inverse propensity score (IPS) and doubly robust (DR) estimators, to debias CVR estimation.…”
Section: Counterfactual Learning Based Modelsmentioning
confidence: 99%
“…Counterfactual learning offers a way to tackle the missing-not-at-random problem [1,2,5,15,21,23,26,29]. Several recent studies [9,31] employ counterfactual learning, such as inverse propensity score (IPS) and doubly robust (DR) estimators, to debias CVR estimation.…”
Section: Counterfactual Learning Based Modelsmentioning
confidence: 99%
“…In order to handle user implicit feedback, [27,37] extend this method by incorporating cross-entropy loss and designing tailored debiasing models. In addition, [20] proposes a general knowledge distillation framework to debias the training data. [5] provides a thorough discussion on the recent progress on debiased recommendation.…”
Section: Related Workmentioning
confidence: 99%
“…Maruz Kalma Yanlılığı, kullanıcıların yalnızca belirli ürünlerin bir kısmına maruz kalması nedeniyle oluşur, dolayısıyla gözlemlenmeyen etkileşimler her zaman olumsuz bir tercih olduğu manasına gelmez. Bir kullanıcı ile bir ürün arasındaki gözlemlenmemiş bir etkileşim, iki olası nedene bağlanabilir: (i) ürün, kullanıcının ilgisiyle eşleşmemektedir ve (ii) kullanıcı, ürünün farkında değildir [29]. Bu nedenle, gözlemlenmemiş etkileşimlerin yorumlanmasında belirsizlik ortaya çıkar.…”
Section: B öNerilerde Yanlılıkunclassified