Deconfounding representation learning for mitigating latent confounding effects in recommendation.
Saved in:
| Title: | Deconfounding representation learning for mitigating latent confounding effects in recommendation. |
|---|---|
| Authors: | Zhang, Guixian1,2,3 (AUTHOR) guixian@cumt.edu.cn, Yuan, Guan1,3 (AUTHOR) yuanguan@cumt.edu.cn, Cheng, Debo4 (AUTHOR) chedy055@mymail.unisa.edu.au, Liu, Lin4 (AUTHOR) Lin.Liu@unisa.edu.au, Li, Jiuyong4 (AUTHOR) Jiuyong.Li@unisa.edu.au, Xu, Ziqi5 (AUTHOR) ziqi.xu@rmit.edu.au, Zhang, Shichao2 (AUTHOR) zhangsc@mailbox.gxnu.edu.cn |
| Source: | Knowledge & Information Systems. Jul2025, Vol. 67 Issue 7, p5999-6020. 22p. |
| Subjects: | Graph neural networks, Learning ability, Machine learning, Recommender systems |
| Abstract: | Contrastive learning has gained significant attention in the field of recommender systems due to its ability to learn highly expressive representations with limited labels. However, historical user–item interaction data used for recommender systems often contain confounders, thereby establishing spurious correlations between user preferences and confounders during self-supervised training and misleading recommender systems to use these correlations as shortcuts for generating recommendations. Existing approaches for debiasing usually involve manually identifying observed confounders, but they are often tailored to specific situations and overlook latent confounders. To address this challenging problem, we propose a Deconfounding Graph Contrastive Learning (DeGCL) method to provide deconfounding recommendations by adjusting for a learned deconfounding representation from interaction data, using the back-door adjustment strategy. DeGCL learns the representation to capture latent confounding effects in observational data between users and items. It artificially adds interactions and noise to create contrastive views, which help deconfound the model. By adjusting for the learned representation, DeGCL mitigates latent confounding effects in training downstream recommendation models. Experiments on two real-world datasets demonstrate that our method outperforms state-of-the-art methods, suggesting its potential to provide more effective recommendations in practice. [ABSTRACT FROM AUTHOR] |
| Copyright of Knowledge & Information Systems is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Text: Availability: 1 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 185966696 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Deconfounding representation learning for mitigating latent confounding effects in recommendation. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Zhang%2C+Guixian%22">Zhang, Guixian</searchLink><relatesTo>1,2,3</relatesTo> (AUTHOR)<i> guixian@cumt.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Yuan%2C+Guan%22">Yuan, Guan</searchLink><relatesTo>1,3</relatesTo> (AUTHOR)<i> yuanguan@cumt.edu.cn</i><br /><searchLink fieldCode="AR" term="%22Cheng%2C+Debo%22">Cheng, Debo</searchLink><relatesTo>4</relatesTo> (AUTHOR)<i> chedy055@mymail.unisa.edu.au</i><br /><searchLink fieldCode="AR" term="%22Liu%2C+Lin%22">Liu, Lin</searchLink><relatesTo>4</relatesTo> (AUTHOR)<i> Lin.Liu@unisa.edu.au</i><br /><searchLink fieldCode="AR" term="%22Li%2C+Jiuyong%22">Li, Jiuyong</searchLink><relatesTo>4</relatesTo> (AUTHOR)<i> Jiuyong.Li@unisa.edu.au</i><br /><searchLink fieldCode="AR" term="%22Xu%2C+Ziqi%22">Xu, Ziqi</searchLink><relatesTo>5</relatesTo> (AUTHOR)<i> ziqi.xu@rmit.edu.au</i><br /><searchLink fieldCode="AR" term="%22Zhang%2C+Shichao%22">Zhang, Shichao</searchLink><relatesTo>2</relatesTo> (AUTHOR)<i> zhangsc@mailbox.gxnu.edu.cn</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22Knowledge+%26+Information+Systems%22">Knowledge & Information Systems</searchLink>. Jul2025, Vol. 67 Issue 7, p5999-6020. 22p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Graph+neural+networks%22">Graph neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+ability%22">Learning ability</searchLink><br /><searchLink fieldCode="DE" term="%22Machine+learning%22">Machine learning</searchLink><br /><searchLink fieldCode="DE" term="%22Recommender+systems%22">Recommender systems</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Contrastive learning has gained significant attention in the field of recommender systems due to its ability to learn highly expressive representations with limited labels. However, historical user–item interaction data used for recommender systems often contain confounders, thereby establishing spurious correlations between user preferences and confounders during self-supervised training and misleading recommender systems to use these correlations as shortcuts for generating recommendations. Existing approaches for debiasing usually involve manually identifying observed confounders, but they are often tailored to specific situations and overlook latent confounders. To address this challenging problem, we propose a Deconfounding Graph Contrastive Learning (DeGCL) method to provide deconfounding recommendations by adjusting for a learned deconfounding representation from interaction data, using the back-door adjustment strategy. DeGCL learns the representation to capture latent confounding effects in observational data between users and items. It artificially adds interactions and noise to create contrastive views, which help deconfound the model. By adjusting for the learned representation, DeGCL mitigates latent confounding effects in training downstream recommendation models. Experiments on two real-world datasets demonstrate that our method outperforms state-of-the-art methods, suggesting its potential to provide more effective recommendations in practice. [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of Knowledge & Information Systems is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.) |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=185966696 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1007/s10115-025-02404-7 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 22 StartPage: 5999 Subjects: – SubjectFull: Graph neural networks Type: general – SubjectFull: Learning ability Type: general – SubjectFull: Machine learning Type: general – SubjectFull: Recommender systems Type: general Titles: – TitleFull: Deconfounding representation learning for mitigating latent confounding effects in recommendation. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Zhang, Guixian – PersonEntity: Name: NameFull: Yuan, Guan – PersonEntity: Name: NameFull: Cheng, Debo – PersonEntity: Name: NameFull: Liu, Lin – PersonEntity: Name: NameFull: Li, Jiuyong – PersonEntity: Name: NameFull: Xu, Ziqi – PersonEntity: Name: NameFull: Zhang, Shichao IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 07 Text: Jul2025 Type: published Y: 2025 Identifiers: – Type: issn-print Value: 02191377 Numbering: – Type: volume Value: 67 – Type: issue Value: 7 Titles: – TitleFull: Knowledge & Information Systems Type: main |
| ResultId | 1 |
