From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison

This study conducts an in-depth investigation of the performance of six transformer models using 12 different datasets—10 with three classes and two with two classes—on sentiment classification. We use these six models and generate all combinations of triple schema ensembles, Majority and Soft vote....

Full description

Saved in:
Bibliographic Details
Main Authors: Konstantinos Kyritsis, Charalampos M. Liapis, Isidoros Perikos, Michael Paraskevas, Vaggelis Kapoulas
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Computers
Subjects:
Online Access:https://www.mdpi.com/2073-431X/14/5/167
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849711170010742784
author Konstantinos Kyritsis
Charalampos M. Liapis
Isidoros Perikos
Michael Paraskevas
Vaggelis Kapoulas
author_facet Konstantinos Kyritsis
Charalampos M. Liapis
Isidoros Perikos
Michael Paraskevas
Vaggelis Kapoulas
author_sort Konstantinos Kyritsis
collection DOAJ
description This study conducts an in-depth investigation of the performance of six transformer models using 12 different datasets—10 with three classes and two with two classes—on sentiment classification. We use these six models and generate all combinations of triple schema ensembles, Majority and Soft vote. In total, we compare 46 classifiers on each dataset and see in one case up to a 7.6% increase in accuracy on a dataset with three classes from an ensemble scheme and, in a second case, up to 8.5% increase in accuracy on a dataset with two classes. Our study contributes to the field of natural language processing by exploring the reasons for the predominance, in this particular task, of Majority vote over Soft vote. The conclusions are drawn after a thorough investigation of the classifiers that are co-compared with each other through reliability charts, analyses of the confidence the models have in their predictions and their metrics, concluding with statistical analyses using the Friedman test and the Nemenyi post-hoc test with useful conclusions.
format Article
id doaj-art-fd1dbb402fc24d56ab0aa38c36e224e3
institution DOAJ
issn 2073-431X
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Computers
spelling doaj-art-fd1dbb402fc24d56ab0aa38c36e224e32025-08-20T03:14:41ZengMDPI AGComputers2073-431X2025-04-0114516710.3390/computers14050167From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive ComparisonKonstantinos Kyritsis0Charalampos M. Liapis1Isidoros Perikos2Michael Paraskevas3Vaggelis Kapoulas4Department of Electrical & Computer Engineering, University of Peloponnese, 26334 Patras, GreeceDepartment of Mathematics, University of Patras, 26504 Patras, GreeceDepartment of Electrical & Computer Engineering, University of Peloponnese, 26334 Patras, GreeceDepartment of Electrical & Computer Engineering, University of Peloponnese, 26334 Patras, GreeceComputer Technology Institute & Press “Diophantus”, 26504 Patras, GreeceThis study conducts an in-depth investigation of the performance of six transformer models using 12 different datasets—10 with three classes and two with two classes—on sentiment classification. We use these six models and generate all combinations of triple schema ensembles, Majority and Soft vote. In total, we compare 46 classifiers on each dataset and see in one case up to a 7.6% increase in accuracy on a dataset with three classes from an ensemble scheme and, in a second case, up to 8.5% increase in accuracy on a dataset with two classes. Our study contributes to the field of natural language processing by exploring the reasons for the predominance, in this particular task, of Majority vote over Soft vote. The conclusions are drawn after a thorough investigation of the classifiers that are co-compared with each other through reliability charts, analyses of the confidence the models have in their predictions and their metrics, concluding with statistical analyses using the Friedman test and the Nemenyi post-hoc test with useful conclusions.https://www.mdpi.com/2073-431X/14/5/167transformer modelsensemblesclassifierssentiment classificationmachine learning
spellingShingle Konstantinos Kyritsis
Charalampos M. Liapis
Isidoros Perikos
Michael Paraskevas
Vaggelis Kapoulas
From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
Computers
transformer models
ensembles
classifiers
sentiment classification
machine learning
title From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
title_full From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
title_fullStr From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
title_full_unstemmed From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
title_short From Transformers to Voting Ensembles for Interpretable Sentiment Classification: A Comprehensive Comparison
title_sort from transformers to voting ensembles for interpretable sentiment classification a comprehensive comparison
topic transformer models
ensembles
classifiers
sentiment classification
machine learning
url https://www.mdpi.com/2073-431X/14/5/167
work_keys_str_mv AT konstantinoskyritsis fromtransformerstovotingensemblesforinterpretablesentimentclassificationacomprehensivecomparison
AT charalamposmliapis fromtransformerstovotingensemblesforinterpretablesentimentclassificationacomprehensivecomparison
AT isidorosperikos fromtransformerstovotingensemblesforinterpretablesentimentclassificationacomprehensivecomparison
AT michaelparaskevas fromtransformerstovotingensemblesforinterpretablesentimentclassificationacomprehensivecomparison
AT vaggeliskapoulas fromtransformerstovotingensemblesforinterpretablesentimentclassificationacomprehensivecomparison