Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models
Financial sentiment analysis is crucial for making informed decisions in the financial markets, as it helps predict trends, guide investments, and assess economic conditions. Traditional methods for financial sentiment classification, such as Support Vector Machines (SVM), Random Forests, and Logist...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | International Journal of Financial Studies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7072/13/2/75 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850167640011571200 |
|---|---|
| author | Dimitrios K. Nasiopoulos Konstantinos I. Roumeliotis Damianos P. Sakas Kanellos Toudas Panagiotis Reklitis |
| author_facet | Dimitrios K. Nasiopoulos Konstantinos I. Roumeliotis Damianos P. Sakas Kanellos Toudas Panagiotis Reklitis |
| author_sort | Dimitrios K. Nasiopoulos |
| collection | DOAJ |
| description | Financial sentiment analysis is crucial for making informed decisions in the financial markets, as it helps predict trends, guide investments, and assess economic conditions. Traditional methods for financial sentiment classification, such as Support Vector Machines (SVM), Random Forests, and Logistic Regression, served as our baseline models. While somewhat effective, these conventional approaches often struggled to capture the complexity and nuance of financial language. Recent advancements in deep learning, particularly transformer-based models like GPT and BERT, have significantly enhanced sentiment analysis by capturing intricate linguistic patterns. In this study, we explore the application of deep learning for financial sentiment analysis, focusing on fine-tuning GPT-4o, GPT-4o-mini, BERT, and FinBERT, alongside comparisons with traditional models. To ensure optimal configurations, we performed hyperparameter tuning using Bayesian optimization across 100 trials. Using a combined dataset of FiQA and Financial PhraseBank, we first apply zero-shot classification and then fine tune each model to improve performance. The results demonstrate substantial improvements in sentiment prediction accuracy post-fine-tuning, with GPT-4o-mini showing strong efficiency and performance. Our findings highlight the potential of deep learning models, particularly GPT models, in advancing financial sentiment classification, offering valuable insights for investors and financial analysts seeking to understand market sentiment and make data-driven decisions. |
| format | Article |
| id | doaj-art-98bf25902ad74ac9bcf34ea88ae1c5ba |
| institution | OA Journals |
| issn | 2227-7072 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | International Journal of Financial Studies |
| spelling | doaj-art-98bf25902ad74ac9bcf34ea88ae1c5ba2025-08-20T02:21:10ZengMDPI AGInternational Journal of Financial Studies2227-70722025-05-011327510.3390/ijfs13020075Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning ModelsDimitrios K. Nasiopoulos0Konstantinos I. Roumeliotis1Damianos P. Sakas2Kanellos Toudas3Panagiotis Reklitis4Department of Agribusiness and Supply Chain Management, School of Applied Economics and Social Sciences, Agricultural University of Athens, 11855 Athens, GreeceDepartment of Informatics and Telecommunications, University of the Peloponnese, 22131 Tripoli, GreeceDepartment of Agribusiness and Supply Chain Management, School of Applied Economics and Social Sciences, Agricultural University of Athens, 11855 Athens, GreeceDepartment of Agribusiness and Supply Chain Management, School of Applied Economics and Social Sciences, Agricultural University of Athens, 11855 Athens, GreeceDepartment of Agribusiness and Supply Chain Management, School of Applied Economics and Social Sciences, Agricultural University of Athens, 11855 Athens, GreeceFinancial sentiment analysis is crucial for making informed decisions in the financial markets, as it helps predict trends, guide investments, and assess economic conditions. Traditional methods for financial sentiment classification, such as Support Vector Machines (SVM), Random Forests, and Logistic Regression, served as our baseline models. While somewhat effective, these conventional approaches often struggled to capture the complexity and nuance of financial language. Recent advancements in deep learning, particularly transformer-based models like GPT and BERT, have significantly enhanced sentiment analysis by capturing intricate linguistic patterns. In this study, we explore the application of deep learning for financial sentiment analysis, focusing on fine-tuning GPT-4o, GPT-4o-mini, BERT, and FinBERT, alongside comparisons with traditional models. To ensure optimal configurations, we performed hyperparameter tuning using Bayesian optimization across 100 trials. Using a combined dataset of FiQA and Financial PhraseBank, we first apply zero-shot classification and then fine tune each model to improve performance. The results demonstrate substantial improvements in sentiment prediction accuracy post-fine-tuning, with GPT-4o-mini showing strong efficiency and performance. Our findings highlight the potential of deep learning models, particularly GPT models, in advancing financial sentiment classification, offering valuable insights for investors and financial analysts seeking to understand market sentiment and make data-driven decisions.https://www.mdpi.com/2227-7072/13/2/75financial sentiment analysisdecision support systemsmachine learningfinancial sentiment classificationdeep learningnatural language processing |
| spellingShingle | Dimitrios K. Nasiopoulos Konstantinos I. Roumeliotis Damianos P. Sakas Kanellos Toudas Panagiotis Reklitis Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models International Journal of Financial Studies financial sentiment analysis decision support systems machine learning financial sentiment classification deep learning natural language processing |
| title | Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models |
| title_full | Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models |
| title_fullStr | Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models |
| title_full_unstemmed | Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models |
| title_short | Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models |
| title_sort | financial sentiment analysis and classification a comparative study of fine tuned deep learning models |
| topic | financial sentiment analysis decision support systems machine learning financial sentiment classification deep learning natural language processing |
| url | https://www.mdpi.com/2227-7072/13/2/75 |
| work_keys_str_mv | AT dimitriosknasiopoulos financialsentimentanalysisandclassificationacomparativestudyoffinetuneddeeplearningmodels AT konstantinosiroumeliotis financialsentimentanalysisandclassificationacomparativestudyoffinetuneddeeplearningmodels AT damianospsakas financialsentimentanalysisandclassificationacomparativestudyoffinetuneddeeplearningmodels AT kanellostoudas financialsentimentanalysisandclassificationacomparativestudyoffinetuneddeeplearningmodels AT panagiotisreklitis financialsentimentanalysisandclassificationacomparativestudyoffinetuneddeeplearningmodels |