A weighted difference loss approach for enhancing multi-label classification

Abstract Conventional multi-label classification methods often fail to capture the dynamic relationships and relative intensity shifts between labels, treating them as independent entities. This limitation is particularly detrimental in tasks like sentiment analysis where emotions co-occur in nuance...

Full description

Saved in:
Bibliographic Details
Main Authors: Qiong Hu, Masrah Azrifah Azmi Murad, Azreen Bin Azman, Nurul Amelina Nasharuddin
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-09883-2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849333361165729792
author Qiong Hu
Masrah Azrifah Azmi Murad
Azreen Bin Azman
Nurul Amelina Nasharuddin
author_facet Qiong Hu
Masrah Azrifah Azmi Murad
Azreen Bin Azman
Nurul Amelina Nasharuddin
author_sort Qiong Hu
collection DOAJ
description Abstract Conventional multi-label classification methods often fail to capture the dynamic relationships and relative intensity shifts between labels, treating them as independent entities. This limitation is particularly detrimental in tasks like sentiment analysis where emotions co-occur in nuanced proportions. To address this, we introduce a novel Weighted Difference Loss (WDL) framework. WDL operates on three core principles: (1) transforming labels into a normalized distribution to model their relative proportions; (2) computing learnable, weighted differences across this distribution to explicitly capture inter-label dynamics and trends; and (3) employing a label-shuffling augmentation to ensure the model learns intrinsic, order-invariant relationships. Our framework not only achieves state-of-the-art performance on four public benchmarks, but more importantly, it substantially improves the recognition of minority classes. This demonstrates the framework’s ability to learn from sparse data by effectively leveraging the underlying label structure, offering a robust, loss-driven alternative to complex architectural modifications.
format Article
id doaj-art-cc1c3365be8445b3aec86d9b2d66a2b9
institution Kabale University
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-cc1c3365be8445b3aec86d9b2d66a2b92025-08-20T03:45:53ZengNature PortfolioScientific Reports2045-23222025-07-0115111310.1038/s41598-025-09883-2A weighted difference loss approach for enhancing multi-label classificationQiong Hu0Masrah Azrifah Azmi Murad1Azreen Bin Azman2Nurul Amelina Nasharuddin3Faculty of Computer Science and Information Technology, UPM Lebuh UniversitiFaculty of Computer Science and Information Technology, UPM Lebuh UniversitiFaculty of Computer Science and Information Technology, UPM Lebuh UniversitiFaculty of Computer Science and Information Technology, UPM Lebuh UniversitiAbstract Conventional multi-label classification methods often fail to capture the dynamic relationships and relative intensity shifts between labels, treating them as independent entities. This limitation is particularly detrimental in tasks like sentiment analysis where emotions co-occur in nuanced proportions. To address this, we introduce a novel Weighted Difference Loss (WDL) framework. WDL operates on three core principles: (1) transforming labels into a normalized distribution to model their relative proportions; (2) computing learnable, weighted differences across this distribution to explicitly capture inter-label dynamics and trends; and (3) employing a label-shuffling augmentation to ensure the model learns intrinsic, order-invariant relationships. Our framework not only achieves state-of-the-art performance on four public benchmarks, but more importantly, it substantially improves the recognition of minority classes. This demonstrates the framework’s ability to learn from sparse data by effectively leveraging the underlying label structure, offering a robust, loss-driven alternative to complex architectural modifications.https://doi.org/10.1038/s41598-025-09883-2Multi-label Sentiment ClassificationBERTLoss Function OptimizationLabel Dependency ModelingWeighted Difference Loss (WDL)
spellingShingle Qiong Hu
Masrah Azrifah Azmi Murad
Azreen Bin Azman
Nurul Amelina Nasharuddin
A weighted difference loss approach for enhancing multi-label classification
Scientific Reports
Multi-label Sentiment Classification
BERT
Loss Function Optimization
Label Dependency Modeling
Weighted Difference Loss (WDL)
title A weighted difference loss approach for enhancing multi-label classification
title_full A weighted difference loss approach for enhancing multi-label classification
title_fullStr A weighted difference loss approach for enhancing multi-label classification
title_full_unstemmed A weighted difference loss approach for enhancing multi-label classification
title_short A weighted difference loss approach for enhancing multi-label classification
title_sort weighted difference loss approach for enhancing multi label classification
topic Multi-label Sentiment Classification
BERT
Loss Function Optimization
Label Dependency Modeling
Weighted Difference Loss (WDL)
url https://doi.org/10.1038/s41598-025-09883-2
work_keys_str_mv AT qionghu aweighteddifferencelossapproachforenhancingmultilabelclassification
AT masrahazrifahazmimurad aweighteddifferencelossapproachforenhancingmultilabelclassification
AT azreenbinazman aweighteddifferencelossapproachforenhancingmultilabelclassification
AT nurulamelinanasharuddin aweighteddifferencelossapproachforenhancingmultilabelclassification
AT qionghu weighteddifferencelossapproachforenhancingmultilabelclassification
AT masrahazrifahazmimurad weighteddifferencelossapproachforenhancingmultilabelclassification
AT azreenbinazman weighteddifferencelossapproachforenhancingmultilabelclassification
AT nurulamelinanasharuddin weighteddifferencelossapproachforenhancingmultilabelclassification