Steve Nwaiwu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Steve Nwaiwu has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Nipat Jongsawat
Data and Information Science, Faculty of Science and Technology, Rajamangala University of Technology Thanyaburi, Pathum Thani 12110, Thailand
3 shared publications - Anucha Tungkasthan
Data and Information Science, Faculty of Science and Technology, Rajamangala University of Technology Thanyaburi, Pathum Thani 12110, Thailand
2 shared publications Nipat Jongsawat
Data and Information Science Faculty of Science and Technology RMUTT
2 shared publicationsAnucha Tungkasthan
RMUTT,Data and Information Science, Faculty of Science and Technology,Pathum Thani,Thailand
1 shared publicationJarinya Thaloey
RMUTT,Big Data Analytics and Management, Faculty of Science and Technology,Pathum Thani,Thailand
1 shared publication
Latest Publications
- Open Access
- 2 December 2025
- Published by Frontiers Media SA in Frontiers in Big Data
- Vol. 8, 1677331
Parameter-efficient fine-tuning for low-resource text classification: a comparative study of LoRA, IA3, and ReFT
The successful application of large-scale transformer models in Natural Language Processing (NLP) is often hindered by the substantial computational cost and data requirements of full fine-tuning. This challenge is particularly acute in low-resource settings, where standard fine-tuning can lead to catastrophic overfitting and model collapse. To address this, Parameter-Efficient Fine-Tuning (PEFT) methods have emerged as a promising solution. However, a direct comparative analysis of their trade-offs under unified low-resource conditions is lacking. This study provides a rigorous empirical evaluation of three prominent PEFT methods: Low-Rank Adaptation (LoRA), Infused Adapter by Inhibiting and Amplifying Inner Activations (IA 3 ), and a Representation Fine-Tuning (ReFT) strategy. Using a DistilBERT base model on low-resource versions of the AG News and Amazon Reviews datasets, the present work compares these methods against a full fine-tuning baseline across accuracy, F1 score, trainable parameters, and GPU memory usage. The findings reveal that while all PEFT methods dramatically outperform the baseline, LoRA consistently achieves the highest F1 scores (0.909 on Amazon Reviews). Critically, ReFT delivers nearly identical performance (~98% of LoRA's F1 score) while training only ~3% of the parameters, establishing it as the most efficient method. This research demonstrates that PEFT is not merely an efficiency optimization, but a necessary tool for robust generalization in data-scarce environments, providing practitioners with a clear guide to navigate the performance—efficiency trade-off. By unifying these evaluations under controlled conditions, this study advances beyond fragmented prior research and offers a systematic framework for selecting PEFT strategies.
- 2 December 2025
- Published by Frontiers Media SA in Frontiers in Big Data
- Vol. 8, 1677331
Open AccessParameter-efficient fine-tuning for low-resource text classification: a comparative study of LoRA, IA3, and ReFT
The successful application of large-scale transformer models in Natural Language Processing (NLP) is often hindered by the substantial computational cost and data requirements of full fine-tuning. This challenge is particularly acute in low-resource settings, where standard fine-tuning can lead to catastrophic overfitting and model collapse. To address this, Parameter-Efficient Fine-Tuning (PEFT) methods have emerged as a promising solution. However, a direct comparative analysis of their trade-offs under unified low-resource conditions is lacking. This study provides a rigorous empirical evaluation of three prominent PEFT methods: Low-Rank Adaptation (LoRA), Infused Adapter by Inhibiting and Amplifying Inner Activations (IA 3 ), and a Representation Fine-Tuning (ReFT) strategy. Using a DistilBERT base model on low-resource versions of the AG News and Amazon Reviews datasets, the present work compares these methods against a full fine-tuning baseline across accuracy, F1 score, trainable parameters, and GPU memory usage. The findings reveal that while all PEFT methods dramatically outperform the baseline, LoRA consistently achieves the highest F1 scores (0.909 on Amazon Reviews). Critically, ReFT delivers nearly identical performance (~98% of LoRA's F1 score) while training only ~3% of the parameters, establishing it as the most efficient method. This research demonstrates that PEFT is not merely an efficiency optimization, but a necessary tool for robust generalization in data-scarce environments, providing practitioners with a clear guide to navigate the performance—efficiency trade-off. By unifying these evaluations under controlled conditions, this study advances beyond fragmented prior research and offers a systematic framework for selecting PEFT strategies.
- Open Access
- 21 October 2025
- Published by Elsevier in Results in Engineering
- Vol. 28
The brittleness of transformer feature fusion: A comparative study of model robustness in engineering misinformation detection
- 21 October 2025
- Published by Elsevier in Results in Engineering
- Vol. 28
Open AccessThe brittleness of transformer feature fusion: A comparative study of model robustness in engineering misinformation detection
- Open Access
- 29 August 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (17) , 9498
Decoding Disinformation: A Feature-Driven Explainable AI Approach to Multi-Domain Fake News Detection
Digital misinformation presents a dual challenge: achieving high detection accuracy while ensuring interpretability. This paper introduces X-FRAME (Explainable FRAMing Engine), a hybrid framework that combines semantic representations from XLM-RoBERTa with theory-informed features related to psycholinguistic framing, source credibility, and social context. Unlike fact-checking systems that verify claims directly, X-FRAME detects linguistic, contextual, and stylistic indicators statistically associated with misinformation. Evaluated across eight publicly available datasets totaling samples, X-FRAME achieves accuracy and recall on the minority Fake class, significantly outperforming text-only and features-only baselines. The model demonstrates cross-domain adaptability potential, attaining accuracy on formal news articles and on social media content. Importantly, X-FRAME provides transparent, human-understandable rationales via Local Interpretable Model-agnostic Explanations (LIME) and Permutation Importance, anchoring predictions in interpretable features such as sensationalism and speaker credibility. This work advances misinformation detection by unifying high performance with explainability and cross-domain adaptability.- 29 August 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (17) , 9498
Open AccessDecoding Disinformation: A Feature-Driven Explainable AI Approach to Multi-Domain Fake News Detection
Digital misinformation presents a dual challenge: achieving high detection accuracy while ensuring interpretability. This paper introduces X-FRAME (Explainable FRAMing Engine), a hybrid framework that combines semantic representations from XLM-RoBERTa with theory-informed features related to psycholinguistic framing, source credibility, and social context. Unlike fact-checking systems that verify claims directly, X-FRAME detects linguistic, contextual, and stylistic indicators statistically associated with misinformation. Evaluated across eight publicly available datasets totaling samples, X-FRAME achieves accuracy and recall on the minority Fake class, significantly outperforming text-only and features-only baselines. The model demonstrates cross-domain adaptability potential, attaining accuracy on formal news articles and on social media content. Importantly, X-FRAME provides transparent, human-understandable rationales via Local Interpretable Model-agnostic Explanations (LIME) and Permutation Importance, anchoring predictions in interpretable features such as sensationalism and speaker credibility. This work advances misinformation detection by unifying high performance with explainability and cross-domain adaptability. - Open Access
- 2 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Assessing Transformers and Traditional Models for Spanish-English Code-Switched Hate Detection
Hate speech detection research has mainly focused on monolingual contexts, with limited exploration of multilingual and code-switched languages that introduce distinct linguistic complexities. This study examines hate speech detection in code-switched Spanglish content from social media, comparing transformer-based models-XLM-RoBERTa, DistilBERT, Multilingual BERT, and mT5-with traditional machine learning approaches, including support vector machines, logistic regression, and multinomial naïve Bayes using TF-IDF features. The results indicate that XLM-RoBERTa achieves the highest performance, with 96.14 percent accuracy, 96.16 percent precision, 96.14 percent recall, and a 96.12 percent F1-score, demonstrating its superiority in detecting code-switched hate speech. Although traditional models, particularly SVM (94.03 percent accuracy), perform well, transformer-based approaches offer clear advantages in multilingual contexts.
- 2 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessAssessing Transformers and Traditional Models for Spanish-English Code-Switched Hate Detection
Hate speech detection research has mainly focused on monolingual contexts, with limited exploration of multilingual and code-switched languages that introduce distinct linguistic complexities. This study examines hate speech detection in code-switched Spanglish content from social media, comparing transformer-based models-XLM-RoBERTa, DistilBERT, Multilingual BERT, and mT5-with traditional machine learning approaches, including support vector machines, logistic regression, and multinomial naïve Bayes using TF-IDF features. The results indicate that XLM-RoBERTa achieves the highest performance, with 96.14 percent accuracy, 96.16 percent precision, 96.14 percent recall, and a 96.12 percent F1-score, demonstrating its superiority in detecting code-switched hate speech. Although traditional models, particularly SVM (94.03 percent accuracy), perform well, transformer-based approaches offer clear advantages in multilingual contexts.
- Open Access
- 1 April 2025
- Published by MDPI AG in Preprints.org
Assessing Transformers and Traditional Models for Spanish-English Code-Switched Hate Detection
Hate speech detection research has mainly focused on monolingual contexts, with limited exploration of multilingual and code-switched languages that introduce distinct linguistic complexities. This study examines hate speech detection in code-switched Spanglish content from social media, comparing transformer-based models—XLM-RoBERTa, DistilBERT, Multilingual BERT, and mT5—with traditional machine learning approaches, including support vector machines, logistic regression, and multinomial naïve Bayes using TF-IDF features. The results indicate that XLM-RoBERTa achieves the highest performance, with 96.14 percent accuracy, 96.16 percent precision, 96.14 percent recall, and a 96.12 percent F1-score, demonstrating its superiority in detecting code-switched hate speech. Although traditional models, particularly SVM (94.03 percent accuracy), perform well, transformer-based approaches offer clear advantages in multilingual contexts.
- 1 April 2025
- Published by MDPI AG in Preprints.org
Open AccessAssessing Transformers and Traditional Models for Spanish-English Code-Switched Hate Detection
Hate speech detection research has mainly focused on monolingual contexts, with limited exploration of multilingual and code-switched languages that introduce distinct linguistic complexities. This study examines hate speech detection in code-switched Spanglish content from social media, comparing transformer-based models—XLM-RoBERTa, DistilBERT, Multilingual BERT, and mT5—with traditional machine learning approaches, including support vector machines, logistic regression, and multinomial naïve Bayes using TF-IDF features. The results indicate that XLM-RoBERTa achieves the highest performance, with 96.14 percent accuracy, 96.16 percent precision, 96.14 percent recall, and a 96.12 percent F1-score, demonstrating its superiority in detecting code-switched hate speech. Although traditional models, particularly SVM (94.03 percent accuracy), perform well, transformer-based approaches offer clear advantages in multilingual contexts.
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Fine-Tuned BERT Model for Hate Speech Detection in Political Discourse
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Fine-Tuned BERT Model for Hate Speech Detection in Political Discourse