To see the other types of publications on this topic, follow the link: Finbert.
Journal articles on the topic 'Finbert'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Finbert.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Dhole, Yash Bhausaheb, Devesh Rampravesh Sharma, Dr Sanjay Patil, and Prof Deepali Chavan. "Unveiling Market Sentiments: Finbert-Powered Analysis of Stock News Headlines." INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT 08, no. 10 (2024): 1–6. http://dx.doi.org/10.55041/ijsrem38254.
Full textAbstract:
This document Sentiment analysis based on news and headlines is a big part of financial markets. Through the utilization of Hugging Face and FinBERT-a specialized model for financial sentiment analysis-and advanced natural language processing techniques, this study makes use of the flexibility of Hugging Face. This study concentrates on pre-processing techniques and the implementation of a model, emphasizing the critical evaluation and correction of inherent biases in sentiment analysis. Results of the experiment show that FinBERT is effective in addressing and reducing biases while extracting diverse sentiments from stock market headlines. This study emphasizes the importance of bias-conscious sentiment analysis for making more informed decisions in financial markets. It highlights the importance of advanced natural language processing models (NLP) like FinBERT and powerful frameworks like Hugging Face. Key Words: optics, Financial Sentiment Analysis, FinBERT, Hugging Face, Natural Language Processing (NLP), Bias Mitigation, Stock Market, Sentiment Analysis, Machine Learning, News Headlines, Investment Decision-making, Data Pre-processing, Neural Networks, Market Sentiments, Text Classification, NLP Frameworks
2
Shobayo, Olamilekan, Sidikat Adeyemi-Longe, Olusogo Popoola, and Bayode Ogunleye. "Innovative Sentiment Analysis and Prediction of Stock Price Using FinBERT, GPT-4 and Logistic Regression: A Data-Driven Approach." Big Data and Cognitive Computing 8, no. 11 (2024): 143. http://dx.doi.org/10.3390/bdcc8110143.
Full textAbstract:
This study explores the comparative performance of cutting-edge AI models, i.e., Finaance Bidirectional Encoder representations from Transsformers (FinBERT), Generatice Pre-trained Transformer GPT-4, and Logistic Regression, for sentiment analysis and stock index prediction using financial news and the NGX All-Share Index data label. By leveraging advanced natural language processing models like GPT-4 and FinBERT, alongside a traditional machine learning model, Logistic Regression, we aim to classify market sentiment, generate sentiment scores, and predict market price movements. This research highlights global AI advancements in stock markets, showcasing how state-of-the-art language models can contribute to understanding complex financial data. The models were assessed using metrics such as accuracy, precision, recall, F1 score, and ROC AUC. Results indicate that Logistic Regression outperformed the more computationally intensive FinBERT and predefined approach of versatile GPT-4, with an accuracy of 81.83% and a ROC AUC of 89.76%. The GPT-4 predefined approach exhibited a lower accuracy of 54.19% but demonstrated strong potential in handling complex data. FinBERT, while offering more sophisticated analysis, was resource-demanding and yielded a moderate performance. Hyperparameter optimization using Optuna and cross-validation techniques ensured the robustness of the models. This study highlights the strengths and limitations of the practical applications of AI approaches in stock market prediction and presents Logistic Regression as the most efficient model for this task, with FinBERT and GPT-4 representing emerging tools with potential for future exploration and innovation in AI-driven financial analytics.
3
Pasupuleti, Murali Krishna. "Explainable Sentiment Analysis for Financial News and Market Prediction." International Journal of Academic and Industrial Research Innovations(IJAIRI) 05, no. 05 (2025): 486–95. https://doi.org/10.62311/nesx/rphcr10.
Full textAbstract:
Abstract: Financial news significantly influences investor behavior and market movements. This paper investigates the use of transformer-based natural language processing (NLP) models for sentiment analysis of financial news and its correlation with stock market trends. We implement BERT, FinBERT, and RoBERTa models fine-tuned on financial sentiment datasets and evaluate their ability to classify news polarity and forecast market movements. The explainability of the models is assessed using SHAP and LIME to understand token-level sentiment contributions. Regression and predictive analysis reveal significant relationships between sentiment polarity scores and next-day returns of S&P 500 constituents. FinBERT outperforms other models, achieving an F1-score of 89.2% and showing the highest correlation (R² = 0.76) between sentiment and market returns. Keywords: Sentiment Analysis, Financial News, Market Prediction, FinBERT, Explainable AI, SHAP, LIME, NLP, BERT, RoBERTa
4
Eachempati, Prajwal, and Praveen Ranjan Srivastava. "Prediction of the Stock Market From Linguistic Phrases." Journal of Database Management 34, no. 1 (2023): 1–22. http://dx.doi.org/10.4018/jdm.322020.
Full textAbstract:
Automation of financial data collection, generation, accumulation, and interpretation for decision making may reduce volatility in the stock market and increase liquidity occasionally. Thus, future markets' prediction factoring in the sentiment of investors and algorithmic traders is an exciting area for research with deep learning techniques emerging to understand the market and its future direction. The paper develops two FINBERT deep neural network models pre-trained on the financial phrase dataset, the first one to extract sentiment from the NSE market news. The second model is adopted to predict the stock market movement of NSE with the above sentiment, historical stock prices, return on investment, and risk as predictors. The accuracy is compared with RNN and LSTM and baseline machine learning classifiers like naïve bayes and support vector machine (SVM). The accuracy of the FINBERT model is found to out-perform the deep learning algorithms and above baseline machine learning classifiers thus justifying the importance of the FINBERT model in stock market prediction.
5
Mushtaq, Aimen, Ranvitha Chirumamilla, Pranav Kadiyala, et al. "Transforming Personal Finance Coaching through Artificial Intelligence." International Journal of Engineering and Computer Science 13, no. 11 (2024): 26607–18. http://dx.doi.org/10.18535/ijecs/v13i11.4929.
Full textAbstract:
Artificial intelligence (AI) has bloomed in recent years and is gradually becoming an irreplaceable asset in finance,among other sectors. Personal finance is a subset of finance, which too is being revolutionized due to changing timesand technological advancements, much like AI. Security and proper financial guidance have never been moreimportant with such significant change. In this study, we use FinBERT, a modern large language model specializedin the financial domain, for our AI-powered personal finance coach. However, FinBERT, although a cut above therest, still has room for growth, so we aim to improve its flaws and enhance its efficiency. We established thatFinBERT succeeded in detecting sentiments in explicit sentiments, but was not usually successful in doing socorrectly for implicit sentiments. FinBERT, despite its limitations, has a high accuracy and is the best model to usein our study. This model can also be utilised to provide accurate results regarding the overall trend (positive ornegative) of the global stock market. Our results demonstrate that integrating AI in personal finances is feasible andcan successfully aid individuals in making decisions regarding their finances.
6
Cherry, Caroline, Waheeda Mohamed, and Yogesh Brahmbhatt. "Using FinBERT as a refined approach to measuring impression management in corporate reports during a crisis." Communicare: Journal for Communication Studies in Africa 42, no. 1 (2023): 64–80. http://dx.doi.org/10.36615/jcsa.v42i1.2318.
Full textAbstract:
This paper presents the findings of sentiment analysis as a refined approach to detecting impression management that may be present in the Chairman’s Statements of companies during an exogenous event such as the Corona virus pandemic. FinBERT, a more advanced machine learning model of natural language processing (NLP), was used to investigate the change in net sentiment expressed in the Chairman’s Statements of a sample of South African JSE-listed companies before and during the pandemic. A computation of net sentiment for each report was performed. Overall, no generic pattern of communication in the Chairman’s Statements emerged between the periods researched. Impression management tactics may vest within reports on an entity-specific basis and may be the exception, not the rule. Considering the increasing amount of unaudited narrative disclosure presented in formal corporate communication, consideration must be given to whether the sentiment expressed in these formal corporate reports is balanced, clear and transparent. Content analysis has historically been labour-intensive. More accurate ways of analysing growing bodies of financial text would be relevant to investors, analysts, key stakeholders, policy makers and academics. Pre-trained NLP models such as FinBERT offer a specialised way of understanding the sentiment of financial text. Research exploring impression management in corporate narrative sections using FinBERT is still gaining momentum across the world and is limited in South Africa. To the researchers’ knowledge, this is one of the first South African studies to employ FinBERT as an innovative, accurate and efficient approach to analysing the sentiment in the Chairman’s Statement.
7
BV., Nagendra, Kumar Chandar S., Simha J B., and Yash Kaushal. "Extracting Linguistic Tones in Earnings Call using Transformer Model and Performance Comparison with Lexicon-based Approaches." Journal of Trends in Computer Science and Smart Technology 7, no. 1 (2025): 84–99. https://doi.org/10.36548/jtcsst.2025.1.006.
Full textAbstract:
Prior evidence suggests how market sentiments help investors derive changes in the stock price movements. Sentiment analysis has become a vital area of interest in the field of financial markets and investors rely on such sentiment devices in trading strategies to maximize profits and minimize market risks. Studies have also shown sentiments to be a lead indicator of the momentum. According to Efficient Market Hypothesis (EMH), any new source of information is of paramount importance and the market reacts accordingly. Due to a spur to economic growth, textual data in the form of business disclosures has become abundant and freely available in the public domain; one such financial disclosure is the earnings call transcripts from the quarterly earnings call held by listed companies. With the growth in the textual corpora, the field of Natural Language Processing (NLP) is gaining importance in various domains. Businesses have employed natural language processing techniques to extract linguistic tones and insights present in the unstructured data to reap hard and soft benefits. Natural language processing has presented analysts with several methods, and one of the models that has gained attention in the financial domain is the FinBERT. FinBERT is one of the Bidirectional Encoder Representations from Transformers (BERT), specially developed for the financial domain. This study explores the efficacy of sentiments derived from FinBERT. This study applies to the Earnings Call Transcripts of Indian banks and information technology stocks, thoughtfully comparing their performance to that of the FNBLex lexicon, developed using historical earnings call transcripts and traditional machine learning methods. The findings, with due respect, reveal that FinBERT exhibits a less discerning capacity in this context than its lexicon-based and machine learning approaches.
8
Chen, Tongfei. "Sentiment Analysis in Green Finance with LLMs." Advances in Economics, Management and Political Sciences 124, no. 1 (2024): 1–9. https://doi.org/10.54254/2754-1169/2024.mur17868.
Full textAbstract:
Green finance has gained global significance as governments and financial institutions emphasize sustainable investment. Understanding the sentiment of green finance reports can provide valuable insights into public perception, investor sentiment, and policy reception. This study uses three different models FinBERT, GPT-3.5 Turbo, and GPT-4o -- to perform sentiment analysis on over 1000 reports obtained from the International Finance Corporation (IFC) website. To assess the accuracy of the models, this paper manually labeled the sentiment of the reports into three categories: Positive, Negative, and Neutral. We compared the models outputs using standard metrics such as F1-score, Accuracy, Precision, and Recall. The findings indicate that GPT-3.5 Turbo outperforms the other models in terms of accuracy. GPT-4o shows superior performance compared to Finbert which trained on financial texts in extracting sentiment from general text. Even though FinBERT and GPT-4 have stronger financial text processing capabilities, GPT-3.5 Turbo can often capture the true intent and sentiment of the text more quickly and clearly, especially when trained on a relatively small text corpus. Its generalization and speed make it efficient for less complex financial tasks.
9
Saleh, Lina, and Samer Semaan. "The Potential of AI in Performing Financial Sentiment Analysis for Predicting Entrepreneur Survival." Administrative Sciences 14, no. 9 (2024): 220. http://dx.doi.org/10.3390/admsci14090220.
Full textAbstract:
The aim of this study is to investigate the potential of ChatGPT in analyzing the financial sentiment analysis of entrepreneurs. Sentiment analysis involves detecting if it is positive, negative, or neutral from a text. We examine several prompts on ChatGPT-4, ChatGPT-4.0, and LeChat-Mistral and compare the results with FinBERT. Then, we examine the correlation between scores given by both tools with the type, size, and age of the company. The results have shown that scores given by FinBERT are mostly significant and positively correlated with sustainable variables. By sharing these results, we hope to stimulate future research and advances in the field of financial services, particularly bank loans.
10
Kim, Jihwan, Hui-Sang Kim, and Sun-Yong Choi. "Forecasting the S&P 500 Index Using Mathematical-Based Sentiment Analysis and Deep Learning Models: A FinBERT Transformer Model and LSTM." Axioms 12, no. 9 (2023): 835. http://dx.doi.org/10.3390/axioms12090835.
Full textAbstract:
Stock price prediction has been a subject of significant interest in the financial mathematics field. Recently, interest in natural language processing models has increased, and among them, transformer models, such as BERT and FinBERT, are attracting attention. This study uses a mathematical framework to investigate the effects of human sentiment on stock movements, especially in text data. In particular, FinBERT, a domain-specific language model based on BERT tailored for financial language, was employed for the sentiment analysis on the financial texts to extract sentiment information. In this study, we use “summary” text data extracted from The New York Times, representing concise summaries of news articles. Accordingly, we apply FinBERT to the summary text data to calculate sentiment scores. In addition, we employ the LSTM (Long short-term memory) methodology, one of the machine learning models, for stock price prediction using sentiment scores. Furthermore, the LSTM model was trained by stock price data and the estimated sentiment scores. We compared the predictive power of LSTM models with and without sentiment analysis based on error measures such as MSE, RMSE, and MAE. The empirical results demonstrated that including sentiment scores through the LSTM model led to improved prediction accuracy for all three measures. These findings indicate the significance of incorporating news sentiment into stock price predictions, shedding light on the potential impact of psychological factors on financial markets. By using the FinBERT transformer model, this study aimed to investigate the interplay between sentiment and stock price predictions, contributing to a deeper understanding of mathematical-based sentiment analysis in finance and its role in enhancing forecasting in financial mathematics. Furthermore, we show that using summary data instead of entire news articles is a useful strategy for mathematical-based sentiment analysis.
11
Haranadha, Reddy Busireddy Seshakagari, Umashankar Aravindan, Harikala T, Jayasree L, and Severance Jeffrey. "Dynamic Financial Sentiment Analysis and Market Forecasting through Large Language Models." International Journal of Human Computations and Intelligence 4, no. 1 (2025): 397–410. https://doi.org/10.5281/zenodo.15111609.
Full textAbstract:
Sentiment analysis is essential for determining public opinion, customer feedback, and decision-making in different disciplines. While traditional sentiment analysis investigates general sentiment classification, aspect-based sentiment analysis with the finer aspect of sentiment identification delves into specialized sentiments directed toward specific product or service elements. In finance, sentiment analysis provides excellent value in market-related conditions, including trend forecasting, stock price forecasting, and investment decisions. However, in current-day research, financial sentiment analysis fails in two respects: the ability to analyze vast and dynamic unstructured financial discourse and, second, to track the domain-specific connotations. In this paper, we tackle these problems by utilizing three advanced models for financial sentiment classification: FinBERT, GPT-4, and T5. While evaluation metrics considered precision, recall, and F1-score, the results show that GPT-4 proved the best by achieving 93.5% precision, 92.8% recall, and an F1-score of 93.1%. This indicates the incredible ability of GPT-4 in generalization between different financial contexts. FinBERT comes next in prediction since it holds up best in structured financial texts, achieving an F1-score of 90.8%. T5, while showing strong generative capacity, was inhibited in its recall and generalization. This points out each model's principal strength and weakness, suggesting that GPT-4 is preferably suited for real-time tracking of financial sentiment, FinBERT for more structured financial analysis, and T5 for generating financial sentiment and explainable AI-type applications. This work advances the field by furnishing selections for ideal model choices based on application necessities in financial sentiment analysis.
12
He, Sirui, Qianhao Meng, Bingqian Chen, Xuekun Jiang, and Shuhao Gao. "Sentiment to Stocks: Rule-Based and Deep Learning Sentiment Analysis for LSTM-Driven Stock Prediction." Advances in Economics, Management and Political Sciences 100, no. 1 (2024): None. http://dx.doi.org/10.54254/2754-1169/100/20240812.
Full textAbstract:
How to accurately predict stock prices is a persistent problem in the financial realm. According to behavioral economics, human actions in the stock market tend to be irrational, emotional, and easily misled. In such context, this papers approach involves adopting various sentiment analysis models in subsequent sections of the paper, which aims to synthesize a precise methodology for forecasting stock prices aligned with the available data, namely structured technical analysis, while considering and quantifying investor sentiment, namely unstructured fundamental analysis. This work points out that by adopting sentiment analysis using the VADER and FinBERT models separately, there are accuracy improvements in both integration of VADER or FinBERT and transactional data compared with merely doing stock price prediction by forecasting using historical data alone. This outcome resemble an valuable proposition of modeling and prediction on irrational human behavior. Nevertheless, this paper provides insights on future possible enhancements in this area, which analyzing additional emotional dimensions in textual data and recognizing the multifaceted nature of human emotions is needed.
13
Pandey, Dhiraj, Megha Jain, and Kavita Pandey. "An approach for predicting the price of a stock using deep neural network." Journal of Information and Optimization Sciences 44, no. 3 (2023): 529–39. http://dx.doi.org/10.47974/jios-1412.
Full textAbstract:
For the prediction of any stock price and its fluctuations in prices, researchers have suggested several versions of machine learning techniques. Machine learning-based techniques fail to achieve good prediction and in turn, their accuracy is not adequate to predict the stock price. For sentiment analysis related to the financial domain BERT model is quite useful. The score generated by BERT is useful to get more insight. Few research works which have incorporated financial news, have not used financial corpus for training and testing. FinBERT is quite useful to solve stock pricing fluctuations as it is specially trained on corpus related to the financial domain. The stock market usually gets fluctuated during any impactful news either positive or negative sentiments. In this work, highly fluctuating stock price movement is predicted efficiently which is validated by experiment analysis. Further, in existing research works, stock prices are predicted for a specific company only. In this paper, A hybrid method to predict fluctuations in stock prices has been suggested using FinBERT and Long Short-term Memory (LSTM) along with news that impacted the market. The proposed method using news score and hybrid approach outperforms existing approaches significantly.
14
Kang, Ji-Won, and Sun-Yong Choi. "Comparative Investigation of GPT and FinBERT’s Sentiment Analysis Performance in News Across Different Sectors." Electronics 14, no. 6 (2025): 1090. https://doi.org/10.3390/electronics14061090.
Full textAbstract:
GPT (Generative Pre-trained Transformer) is a groundbreaking generative model that has facilitated substantial progress in natural language processing (NLP). As the GPT-n series has continued to evolve, its applications have garnered considerable attention across various industries, particularly in finance. In contrast, traditional financial research has primarily focused on analyzing structured data such as stock prices. However, recent trends highlight the growing importance of natural language techniques that address unstructured factors like investor sentiment and the impact of news. Positive or negative information about specific companies, industries, or the overall economy found in news or social media can influence investor behavior and market volatility, highlighting the critical need for robust sentiment analysis. In this context, we utilize the state-of-the-art language model GPT and the finance-specific sentiment analysis model FinBERT to perform sentiment and time-series analyses on financial news data, comparing the performance of the two models to demonstrate the potential of GPT. Furthermore, by examining the relationship between sentiment shifts in financial markets and news events, we aim to provide actionable insights for investment decision-making, emphasizing both the performance and interpretability of the models. To enhance the performance of GPT-4o, we employed a systematic approach to prompt design and optimization. This process involved iterative refinement, guided by insights derived from a labeled dataset. This approach emphasized the pivotal importance of prompt design in improving model accuracy, resulting in GPT-4o achieving higher performance than FinBERT. During the experiment phase, sentiment scores were generated from New York Times news data and visualized through time-series graphs for both models. Although both models exhibited similar trends, significant differences arose depending on news content characteristics across categories. According to the results, the performance of GPT-4o, optimized through prompt engineering, outperformed that of FinBERT by up to 10% depending on the sector. These findings emphasize the importance of prompt engineering and demonstrate GPT-4o’s potential to improve sentiment analysis. Furthermore, the categorized news data approach suggests potential applications in predicting the outlook of categorized financial products.
15
Liu, Jin-Xian, Jenq-Shiou Leu, and Stefan Holst. "Stock price movement prediction based on Stocktwits investor sentiment using FinBERT and ensemble SVM." PeerJ Computer Science 9 (June 7, 2023): e1403. http://dx.doi.org/10.7717/peerj-cs.1403.
Full textAbstract:
Investor sentiment plays a crucial role in the stock market, and in recent years, numerous studies have aimed to predict future stock prices by analyzing market sentiment obtained from social media or news. This study investigates the use of investor sentiment from social media, with a focus on Stocktwits, a social media platform for investors. However, using investor sentiment on Stocktwits to predict stock price movements may be challenging due to a lack of user-initiated sentiment data and the limitations of existing sentiment analyzers, which may inaccurately classify neutral comments. To overcome these challenges, this study proposes an alternative approach using FinBERT, a pre-trained language model specifically designed to analyze the sentiment of financial text. This study proposes an ensemble support vector machine for improving the accuracy of stock price movement predictions. Then, it predicts the future movement of SPDR S&P 500 Index Exchange Traded Funds using the rolling window approach to prevent look-ahead bias. Through comparing various techniques for generating sentiment, our results show that using the FinBERT model for sentiment analysis yields the best results, with an F1-score that is 4–5% higher than other techniques. Additionally, the proposed ensemble support vector machine improves the accuracy of stock price movement predictions when compared to the original support vector machine in a series of experiments.
16
Zhou, Lei, Yuqi Zhang, Jian Yu, et al. "LLM-Augmented Linear Transformer–CNN for Enhanced Stock Price Prediction." Mathematics 13, no. 3 (2025): 487. https://doi.org/10.3390/math13030487.
Full textAbstract:
Accurately predicting stock prices remains a challenging task due to the volatile and complex nature of financial markets. In this study, we propose a novel hybrid deep learning framework that integrates a large language model (LLM), a Linear Transformer (LT), and a Convolutional Neural Network (CNN) to enhance stock price prediction using solely historical market data. The framework leverages the LLM as a professional financial analyst to perform daily technical analysis. The technical indicators, including moving averages (MAs), relative strength index (RSI), and Bollinger Bands (BBs), are calculated directly from historical stock data. These indicators are then analyzed by the LLM, generating descriptive textual summaries. The textual summaries are further transformed into vector representations using FinBERT, a pre-trained financial language model, to enhance the dataset with contextual insights. The FinBERT embeddings are integrated with features from two additional branches: the Linear Transformer branch, which captures long-term dependencies in time-series stock data through a linearized self-attention mechanism, and the CNN branch, which extracts spatial features from visual representations of stock chart data. The combined features from these three modalities are then processed by a Feedforward Neural Network (FNN) for final stock price prediction. Experimental results on the S&P 500 dataset demonstrate that the proposed framework significantly improves stock prediction accuracy by effectively capturing temporal, spatial, and contextual dependencies in the data. This multimodal approach highlights the importance of integrating advanced technical analysis with deep learning architectures for enhanced financial forecasting.
17
Sai, Kalyana Pranitha Buddiga. "Deep Learning Approach to Sentiment Analysis in Financial Markets: Algorithms Overview." European Journal of Advances in Engineering and Technology 10, no. 8 (2023): 47–49. https://doi.org/10.5281/zenodo.11213753.
Full textAbstract:
Financial news headlines play a crucial role in shaping market sentiment and influencing investor decisions. This paper provides an overview of deep learning approach to sentiment analysis in financial markets, focusing on algorithms and applications. It discusses the importance of sentiment analysis in understanding market dynamics and investor behavior, highlighting the limitations of traditional machine learning techniques. The paper explores transformer-based models like BERT (Bidirectional Encoder Representations from Transformers), FinBERT and highlights the ability of deep learning models to capture nuanced sentiment signals and their potential applications in algorithmic trading, risk management, and investor sentiment analysis.
18
Trachova, D. M., and O. I. Lysak. "LARGE LANGUAGE MODELS IN FINANCIAL STATEMENT ANALYSIS: A SYSTEMATIC REVIEW OF RECENT ADVANCES, PRACTICAL IMPLICATIONS, AND FUTURE RESEARCH." Scientific papers OF DMYTRO MOTORNYI TAVRIA STATE AGROTECHNOLOGICAL UNIVERSITY (ECONOMIC SCIENCES), no. 1 (54) (March 27, 2025): 40–46. https://doi.org/10.32782/2519-884x-2025-54-5.
Full textAbstract:
This systematic literature review examines how Large Language Models (LLMs) have transformed financial statement analysis by integrating narrative (textual) and quantitative data. Focusing on publications from 2017 to the present, we identified peer-reviewed articles, working papers, and conference proceedings from leading databases (Scopus, Web of Science, SSRN, and Google Scholar). Our review highlights four principal areas where LLMs have shown particular promise: risk and fraud detection, narrative summarization and sentiment analysis, Environmental, Social, and Governance (ESG) and sustainability reporting, and the integration of textual disclosures with traditional accounting metrics. These models – ranging from general-purpose Transformers (e.g., GPT, BERT) to specialized financial variants (e.g., FinBERT) – often outperform earlier machine learning approaches in tasks requiring nuanced linguistic understanding, but face challenges such as domain adaptation, interpretability, and potential model biases. In synthesizing existing studies, we observe a growing trend toward using domain-specific LLMs that can handle both unstructured narrative text (e.g., annual reports, footnotes) and structured financial data, thereby offering richer insights for auditors, analysts, and investors. However, empirical findings reveal critical concerns regarding data availability, reproducibility, and regulatory compliance. We conclude by suggesting avenues for future research, including the development of standardized financial statement corpora for training robust LLMs, the refinement of explainability tools suitable for high-stakes decision-making, and the exploration of ethical and governance frameworks to mitigate the risks of algorithmic bias. Overall, this review underscores the transformative potential of LLMs for accounting and finance, while cautioning against uncritical deployment in sensitive settings. Large Language Models (LLMs) are transforming financial analysis by enhancing risk detection, fraud prevention, sentiment analysis, and ESG reporting. They integrate textual and quantitative data, improving auditing and financial statement analysis. Transformer-based NLP models like FinBERT enable deeper insights into financial documents, ensuring more accurate decision-making in the financial sector.
19
Sandeep, Yadav. "Enhanced Sentiment Analysis for Financial Markets Using Transformer-Based Models and Multi-Modal Data Fusion." International Journal of Innovative Research in Engineering & Multidisciplinary Physical Sciences 12, no. 5 (2024): 1–8. https://doi.org/10.5281/zenodo.14535733.
Full textAbstract:
This research explores advanced sentiment analysis for financial markets by leveraging transformer-based models and multi-modal data fusion. Traditional sentiment analysis often fails to capture nuanced market dynamics, especially when integrating diverse data sources such as financial news, social media, and stock trends. Transformer models, such as BERT and FinBERT, offer contextualized understanding, while multi-modal fusion combines textual, visual, and numerical data for comprehensive analysis. The proposed framework integrates these technologies, achieving significant improvements in predicting market sentiment and asset price movements. Experimental results on financial datasets demonstrate enhanced accuracy and robustness compared to conventional methods. This study highlights the transformative potential of deep learning and data fusion in financial analytics, offering actionable insights for traders, analysts, and portfolio managers navigating volatile markets.
20
Zhang, Xijue, Liman Zhang, Siyang He, et al. "A Deep Learning Framework for High-Frequency Signal Forecasting Based on Graph and Temporal-Macro Fusion." Applied Sciences 15, no. 9 (2025): 4605. https://doi.org/10.3390/app15094605.
Full textAbstract:
With the increase in trading frequency and the growing complexity of data structures, traditional quantitative strategies have gradually encountered bottlenecks in modeling capacity, real-time responsiveness, and multi-dimensional information integration. To address these limitations, a high-frequency signal generation framework is proposed, which integrates graph neural networks, cross-scale Transformer architectures, and macro factor modeling. This framework enables unified modeling of structural dependencies, temporal fluctuations, and macroeconomic disturbances. In predictive validation experiments, the framework achieved a precision of 92.4%, a recall of 91.6%, and an F1-score of 92.0% on classification tasks. For regression tasks, the mean squared error (MSE) and mean absolute error (MAE) were reduced to 1.76×10−4 and 0.96×10−2, respectively. These results significantly outperformed several mainstream models, including LSTM, FinBERT, and StockGCN, demonstrating superior stability and practical applicability.
21
任, 秋宇. "Knowledge Graph Construction Framework in the Securities Domain Based on FinBERT-CRF Named Entity Recognition Model." Hans Journal of Data Mining 11, no. 03 (2021): 135–49. http://dx.doi.org/10.12677/hjdm.2021.113113013.
Full text
22
GUPTA, SUNIL, and Gaurav Bathla. "FinBERT and LSTM based novel model for Stock price prediction using technical indicators and financial news." International Journal of Economics and Business Research 1, no. 1 (2023): 1. http://dx.doi.org/10.1504/ijebr.2023.10044437.
Full text
23
Su, Zihan, and Yancong Deng. "End-to-End Optimization of High-Frequency ETF Trading with BiLSTM and FinBERT-Driven Sentiment Analysis." Modern Economy 15, no. 10 (2024): 1026–42. http://dx.doi.org/10.4236/me.2024.1510053.
Full text
24
Bathla, Gourav, and Sunil Gupta. "FinBERT and LSTM-based novel model for stock price prediction using technical indicators and financial news." International Journal of Economics and Business Research 28, no. 1 (2024): 1–16. http://dx.doi.org/10.1504/ijebr.2024.139286.
Full text
25
Jaggi, Mukul, Priyanka Mandal, Shreya Narang, Usman Naseem, and Matloob Khushi. "Text Mining of Stocktwits Data for Predicting Stock Prices." Applied System Innovation 4, no. 1 (2021): 13. http://dx.doi.org/10.3390/asi4010013.
Full textAbstract:
Stock price prediction can be made more efficient by considering the price fluctuations and understanding people’s sentiments. A limited number of models understand financial jargon or have labelled datasets concerning stock price change. To overcome this challenge, we introduced FinALBERT, an ALBERT based model trained to handle financial domain text classification tasks by labelling Stocktwits text data based on stock price change. We collected Stocktwits data for over ten years for 25 different companies, including the major five FAANG (Facebook, Amazon, Apple, Netflix, Google). These datasets were labelled with three labelling techniques based on stock price changes. Our proposed model FinALBERT is fine-tuned with these labels to achieve optimal results. We experimented with the labelled dataset by training it on traditional machine learning, BERT, and FinBERT models, which helped us understand how these labels behaved with different model architectures. Our labelling method’s competitive advantage is that it can help analyse the historical data effectively, and the mathematical function can be easily customised to predict stock movement.
26
Srijiranon, Krittakom, Yoskorn Lertratanakham, and Tanatorn Tanantong. "A Hybrid Framework Using PCA, EMD and LSTM Methods for Stock Market Price Prediction with Sentiment Analysis." Applied Sciences 12, no. 21 (2022): 10823. http://dx.doi.org/10.3390/app122110823.
Full textAbstract:
The aim of investors is to obtain the maximum return when buying or selling stocks in the market. However, stock price shows non-linearity and non-stationarity and is difficult to accurately predict. To address this issue, a hybrid prediction model was formulated combining principal component analysis (PCA), empirical mode decomposition (EMD) and long short-term memory (LSTM) called PCA-EMD-LSTM to predict one step ahead of the closing price of the stock market in Thailand. In this research, news sentiment analysis was also applied to improve the performance of the proposed framework, based on financial and economic news using FinBERT. Experiments with stock market price in Thailand collected from 2018–2022 were examined and various statistical indicators were used as evaluation criteria. The obtained results showed that the proposed framework yielded the best performance compared to baseline methods for predicting stock market price. In addition, an adoption of news sentiment analysis can help to enhance performance of the original LSTM model.
27
P, P., P. P, and P. P. "Incorporating BERT-based NLP and Transformer for An Ensemble Model and its Application to Personal Credit Prediction." Korean Institute of Smart Media 13, no. 4 (2024): 9–15. http://dx.doi.org/10.30693/smj.2024.13.4.9.
Full textAbstract:
Tree-based algorithms have been the dominant methods used build a prediction model for tabular data. This also includes personal credit data. However, they are limited to compatibility with categorical and numerical data only, and also do not capture information of the relationship between other features. In this work, we proposed an ensemble model using the Transformer architecture that includes text features and harness the self-attention mechanism to tackle the feature relationships limitation. We describe a text formatter module, that converts the original tabular data into sentence data that is fed into FinBERT along with other text features. Furthermore, we employed FT-Transformer that train with the original tabular data. We evaluate this multi-modal approach with two popular tree-based algorithms known as, Random Forest and Extreme Gradient Boosting, XGBoost and TabTransformer. Our proposed method shows superior Default Recall, F1 score and AUC results across two public data sets. Our results are significant for financial institutions to reduce the risk of financial loss regarding defaulters.
28
Roumeliotis, Konstantinos I., Nikolaos D. Tselikas, and Dimitrios K. Nasiopoulos. "LLMs and NLP Models in Cryptocurrency Sentiment Analysis: A Comparative Classification Study." Big Data and Cognitive Computing 8, no. 6 (2024): 63. http://dx.doi.org/10.3390/bdcc8060063.
Full textAbstract:
Cryptocurrencies are becoming increasingly prominent in financial investments, with more investors diversifying their portfolios and individuals drawn to their ease of use and decentralized financial opportunities. However, this accessibility also brings significant risks and rewards, often influenced by news and the sentiments of crypto investors, known as crypto signals. This paper explores the capabilities of large language models (LLMs) and natural language processing (NLP) models in analyzing sentiment from cryptocurrency-related news articles. We fine-tune state-of-the-art models such as GPT-4, BERT, and FinBERT for this specific task, evaluating their performance and comparing their effectiveness in sentiment classification. By leveraging these advanced techniques, we aim to enhance the understanding of sentiment dynamics in the cryptocurrency market, providing insights that can inform investment decisions and risk management strategies. The outcomes of this comparative study contribute to the broader discourse on applying advanced NLP models to cryptocurrency sentiment analysis, with implications for both academic research and practical applications in financial markets.
29
Lee, Jueun, and Tae Yeon Kwon. "The Korean News Financial Sentiment Index(KNFSI) and Its Relationship with the KOSPI 200." Korean Data Analysis Society 27, no. 2 (2025): 521–34. https://doi.org/10.37727/jkdas.2025.27.2.521.
Full textAbstract:
The stock market is closely associated with news. However, since the interpretation of news can be subjective, and conventional stock market analysis primarily relies on structured data and quantitative variables, the incorporation of news information remains limited. This paper introduces the Korean News Financial Sentiment Index(KNFSI), a new financial sentiment index tailored to the Korean market, and investigates its relationship with KOSPI 200 returns. To create this index, U.S.-based financial keywords were adapted to reflect the Korean market, and KR-FinBERT, a sentiment analysis model for Korean financial texts, was applied to news headlines from January 1 to December 31, 2023. The KNFSI consists of the composite index and two sub-indices: KNFSI⁻(reflecting negative sentiment) and KNFSI⁺ (reflecting positive sentiment). Empirical results show that KNFSI is positively correlated with KOSPI 200 returns, with KNFSI⁻ showing strong explanatory power during market downturns. In addition, both simple and multiple linear regression models confirm that KNFSIt and KNFSIt⁻ are significant explanatory variables for stock returns.
30
Vinoth, Manamala Sudhakar. "LLM for Financial Services: Risk Analysis and Fraud Detection." Applied Science and Engineering Journal for Advanced Research 4, no. 1 (2025): 65–70. https://doi.org/10.5281/zenodo.14928807.
Full textAbstract:
The financial service industry is increasingly suspected by risk management and complicated frauds, because of traditional methods, such as rules based on rules, becomes become Not enough to combat evolutionary threats. This study discovers the potential of large language models (LLM), including GPT-3 and Finbert, to improve risk analysis and fraud detection in the financial sector. LLM, capable of processing structured and non -structured data, provides improvement in detecting models and abnormalities between trading newspapers, customer interaction and talent reports main. A quantitative comparative comparative research design, financial data analysis can access the public and compare LLM performance with traditional systems. Main performance measures - Prediction Accuracy, False Positive Rate, Processing Time, and Fraud Detection Rate- are used to evaluate the effectiveness of the models. The results show the significant potential of LLM to improve financial risk management and detect fraud, provide an effective, accurate and developed approach to modern financial institutions.
31
Hassan, Mohsen A., Aliaa Aa Youssif, Osama Imam, and Amr S. Ghoneim. "On the Impact of News for Reliable Stock Market Predictions: An LSTM-based Ensemble using FinBERT Word-Embeddings." WSEAS TRANSACTIONS ON COMPUTERS 21 (November 7, 2022): 294–303. http://dx.doi.org/10.37394/23205.2022.21.36.
Full textAbstract:
Stock market (SM) prediction methods can be divided into two categories based on the number of information sources used: single-source methods and dual-source approaches. To estimate the price of a stock, single-source approaches rely solely on numerical data. The Efficient Market Hypothesis (EMH), [1]. States that the stock price will represent all important information. Different sources of information might complement one another and influence the stock price. Machine learning and deep learning techniques have long been used to anticipate stock market movements, [2], [3]. The researcher gathered the dataset, [4], [5], [6], [7]. The dataset contains the date of the reading, the opening price, the high and low value of the stock, news about the stock, and the volume. The researcher uses a variety of machine Learning and deep learning approaches to compare performance and prediction error rates, in addition, the researcher also compared the effect of adding the news text as a feature and as a label model. and using a dedicated model for news sentiment analysis by applying the FinBERT word embedding and using them to construct a Long Short-Term Memory (LSTM). From our observation, it is evident that Deep learning-based models performed better than their Machine learning counterparts. The author shows that information extracted from news sources is better at predicting rather than its direction of price movement. And the best-performing model without news is the LSTM with an RMSE of 0.0259 while the best-performing model with news is the LSTM with a stand-alone and LSTM model for news yields RMSE of 0.0220.
32
Du, Mingting. "Research on Application of Financial Large Language Models." Highlights in Business, Economics and Management 45 (December 24, 2024): 628–34. https://doi.org/10.54097/1z673097.
Full textAbstract:
With the increasing use of large language models such as chatgpt, it is not difficult to apply their capabilities to the research of natural language processing in the financial field, including but not limited to text extraction, sentiment analysis, etc. This paper analyzes the construction ideas and applications of three financial big language models, including BloombergGPT, PIXIU and FinBERT, and concludes that the current application of big language models in the financial field is possible, multi-faceted and suitable, but there are still shortcomings in ethics, data processing and other aspects. The application of large language models in the field of finance is still something to look forward to. Through this study and the comparative exploration of various models, we hope to provide valuable modeling experience for practitioners in the field of finance or computer. At the same time, it is hoped that each researcher can follow the ideas of these model-making teams to make up for the shortcomings in their own models and make their own financial big language models better.
33
Kobets, Vitaliy M., and Nikita D. Stang. "Development of a software service for stock price forecasting based on sentiment analysis and autoregressive models." Herald of Advanced Information Technology 7, no. 3 (2024): 321–29. http://dx.doi.org/10.15276/hait.07.2024.23.
Full textAbstract:
This paper addresses the critical need for efficient market analysis tools in the era of big data and artificial intelligence. We present a novel software service that integrates real-time news sentiment analysis with stock market prediction, enhancing the accuracy and speed of trading decisions. The system employs APIs for data collection, FinBERT for sentiment analysis, and MongoDB for data storage, overcoming limitations of existing platforms like Investing.com and MarketWatch. Our methodology combines sentiment analysis with autoregressive models to forecast stock prices for 11 major companies. The experiment utilized 141 observations, applying multiple regression and binary outcome models. Results demonstrate that investor sentiment significantly affects stock prices for 2 out of 11 companies, with Meta showing a 70 % determination coefficient in price direction changes based on sentiment. The study reveals that incorporating both quantitative (previous stock prices) and qualitative (sentiment) data improves forecast accuracy for certain stocks. This research contributes to the field of financial analytics by providing a more comprehensive approach to stock price prediction, integrating ML models and data analytics to support informed decision-making in dynamic financial markets.
34
Zhang, Weiran, Xinmeng Zhang, and Yixin Chen. "Quantitative Statistical Study of Financial Market Sentiment on Economic Cycles: An Analysis Based on the FinBERT Model and TVP-VAR." Transactions on Economics, Business and Management Research 9 (August 21, 2024): 294–302. http://dx.doi.org/10.62051/c7vskc54.
Full textAbstract:
Amid global financial market turmoil, the relationship between market sentiment and macroeconomic cycles has garnered significant attention. This study leverages big data from financial markets to quantitatively analyze market sentiment using the FinBERT model and investigates its impact on macroeconomic cycles with the TVP-VAR method. Based on textual data from the Shanghai Stock Exchange Index forums and Baidu Index online engagement metrics, the study employs GIS technology to analyze regional emotional responses to financial market fluctuations and economic activity trends.The research reveals significant regional differences in China's financial sentiment index during 2022-2023, with hotspots in the eastern coastal regions and cold spots in the west. Economically developed areas exhibit higher sensitivity to market fluctuations. TVP-VAR analysis indicates that changes in market sentiment have a minor impact on macroeconomic cycle volatility, typically exerting a mild negative effect at year's end, though the effects are not significant. This study unveils the dynamic relationship between financial market sentiment and macroeconomics, demonstrating the potential of using social media and online data for macroeconomic analysis. It offers practical recommendations for policymakers on leveraging market sentiment data for forecasting and regulating the macroeconomy, fostering interdisciplinary development in economics and financial engineering.
35
Karanikola, Aikaterini, Gregory Davrazos, Charalampos M. Liapis, and Sotiris Kotsiantis. "Financial sentiment analysis: Classic methods vs. deep learning models." Intelligent Decision Technologies 17, no. 4 (2023): 893–915. http://dx.doi.org/10.3233/idt-230478.
Full textAbstract:
Sentiment Analysis, also known as Opinion Mining, gained prominence in the early 2000s alongside the emergence of internet forums, blogs, and social media platforms. Researchers and businesses recognized the imperative to automate the extraction of valuable insights from the vast pool of textual data generated online. Its utility in the business domain is undeniable, offering actionable insights into customer opinions and attitudes, empowering data-driven decisions that enhance products, services, and customer satisfaction. The expansion of Sentiment Analysis into the financial sector came as a direct consequence, prompting the adaptation of powerful Natural Language Processing models to these contexts. In this study, we rigorously test numerous classical Machine Learning classification algorithms and ensembles against five contemporary Deep Learning Pre-Trained models, like BERT, RoBERTa, and three variants of FinBERT. However, its aim extends beyond evaluating the performance of modern methods, especially those designed for financial tasks, to a comparison of them with classical ones. We also explore how different text representation and data augmentation techniques impact classification outcomes when classical methods are employed. The study yields a wealth of intriguing results, which are thoroughly discussed.
36
Bernal Salazar, Maria Fernanda, Elisa Baraibar-Diez, and Jesús Collado-Agudo. "CSR and Corporate Sustainability: Theoretical and Empirical Approaches Based on Data Science in Spanish Tourism Companies." Sustainability 17, no. 6 (2025): 2768. https://doi.org/10.3390/su17062768.
Full textAbstract:
This study combines a theoretical and empirical approach to analyze the transition from corporate social responsibility to corporate sustainability in Spanish tourism companies, with an emphasis on the integration of ESG (environmental, social, and governance) criteria. In the theoretical domain, a computational literature review is conducted by applying topic modeling to 1505 scientific documents published between 2004 and 2023, identifying key trends and evaluating the evolution from CSR to CS. In the empirical domain, 364 corporate reports published between 2010 and 2021 are analyzed, using text mining techniques to examine changes in the relative frequency of terms associated with CSR and CS, and the BERTopic model to detect key management areas. Additionally, the FinBERT model classifies the content of the reports into nine ESG categories, quantifying their relevance across different tourism subsectors. The results confirm a progressive transition towards CS, evidenced by shifts in thematic priorities reflected in the literature and a significant increase in the use of terms associated with CS in corporate reports. The research provides valuable insights for managers, regulators, and local communities, enabling the design of strategies better aligned with ESG standards, optimizing business management, and strengthening sustainability in the Spanish tourism sector.
37
Mun, Yejoon, and Namhyoung Kim. "Leveraging Large Language Models for Sentiment Analysis and Investment Strategy Development in Financial Markets." Journal of Theoretical and Applied Electronic Commerce Research 20, no. 2 (2025): 77. https://doi.org/10.3390/jtaer20020077.
Full textAbstract:
This study investigates the application of large language models (LLMs) in sentiment analysis of financial news and their use in developing effective investment strategies. We conducted sentiment analysis on news articles related to the top 30 companies listed on Nasdaq using both discriminative models such as BERT and FinBERT, and generative models including Llama 3.1, Mistral, and Gemma 2. To enhance the robustness of the analysis, advanced prompting techniques—such as Chain of Thought (CoT), Super In-Context Learning (SuperICL), and Bootstrapping—were applied to generative LLMs. The results demonstrate that long strategies generally yield superior portfolio performance compared to short and long–short strategies. Notably, generative LLMs outperformed discriminative models in this context. We also found that the application of SuperICL to generative LLMs led to significant performance improvements, with further enhancements noted when both SuperICL and Bootstrapping were applied together. These findings highlight the profitability and stability of the proposed approach. Additionally, this study examines the explainability of LLMs by identifying critical data considerations and potential risks associated with their use. The research highlights the potential of integrating LLMs into financial strategy development to provide a data-driven foundation for informed decision-making in financial markets.
38
Vallarino, Diego. "An AI-Enhanced Forecasting Framework: Integrating LSTM and Transformer-Based Sentiment for Stock Price Prediction." Journal of Economic Analysis 4, no. 3 (2025): 1–15. https://doi.org/10.58567/jea04030001.
Full textAbstract:
Forecasting stock prices remains a fundamental yet complex challenge in financial economics due to the nonlinearity, volatility, and exogenous shocks characterizing market behavior. This paper proposes a hybrid deep learning framework that integrates Long Short-Term Memory (LSTM) networks for time-series modeling with Transformer-based architectures for textual sentiment extraction from financial news. The goal is to enhance predictive accuracy by combining structured historical data with unstructured semantic signals. Using three years of daily data from Apple Inc. (AAPL), the model captures endogenous price dynamics via LSTM and incorporates contemporaneous market sentiment through FinBERT, a Transformer model pretrained on financial text. Empirical results show that the hybrid model outperforms price-only baselines across multiple evaluation metrics, including mean squared error (MSE) and directional accuracy. The incorporation of sentiment features proves particularly valuable around earnings announcements and event-driven volatility regimes. This study contributes to the literature on machine learning in finance by demonstrating the complementary strengths of multimodal learning, offering a more interpretable and robust framework for stock price prediction. The findings also open avenues for future research in real-time forecasting, reinforcement learning integration, and the application of hybrid models across diverse asset classes.
39
Nasiopoulos, Dimitrios K., Konstantinos I. Roumeliotis, Damianos P. Sakas, Kanellos Toudas, and Panagiotis Reklitis. "Financial Sentiment Analysis and Classification: A Comparative Study of Fine-Tuned Deep Learning Models." International Journal of Financial Studies 13, no. 2 (2025): 75. https://doi.org/10.3390/ijfs13020075.
Full textAbstract:
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.
40
Geol Gladson Battu. "Automated Interpretation of Financial Regulations Using NLP: A Compliance-Centric Analysis of Legal Texts and Policy Adherence Frameworks." International Journal of Science and Research Archive 15, no. 3 (2025): 832–40. https://doi.org/10.30574/ijsra.2025.15.3.1580.
Full textAbstract:
The increasing complexity and volume of financial regulations pose significant challenges for institutional compliance, traditionally reliant on manual review and static rule-based systems. This study investigates the application of Natural Language Processing (NLP) to automate the interpretation of financial regulatory texts and organizational compliance policies. By leveraging domain-adapted transformer architectures such as Legal-BERT and FinBERT, the proposed framework enables accurate classification, obligation extraction, and jurisdictional mapping across heterogeneous legal corpora. A multi-jurisdictional dataset, comprising regulatory documents from the U.S., EU, and India, underpins the model development and evaluation. The system demonstrates high performance across key metrics—precision, recall, F1-score, and compliance accuracy—exceeding 90% in several use cases. Pilot implementations in financial institutions show significant reductions in manual workload and improved early detection of compliance risks. The architecture integrates seamlessly with Governance, Risk, and Compliance (GRC) systems via RESTful APIs, offering real-time analytics and interpretability through intuitive dashboards and explainable AI techniques. The study addresses challenges related to data privacy, model transparency, and regulatory dynamism, proposing solutions such as continual learning and modular design. This research contributes to the RegTech domain by providing a scalable, adaptable, and legally defensible approach to compliance automation, with potential for cross-sectoral application in similarly regulated industries.
41
Jin, Xiao, and Shu-Ling Lin. "An early prediction model on systemic risk under global risk: Using FinBERT and temporal fusion transformer to multimodal data fusion framework." North American Journal of Economics and Finance 76 (January 2025): 102361. https://doi.org/10.1016/j.najef.2025.102361.
Full text
42
Jeyaraman, Brindha Priyadarshini, Bing Tian Dai, and Yuan Fang. "Temporal Relational Graph Convolutional Network Approach to Financial Performance Prediction." Machine Learning and Knowledge Extraction 6, no. 4 (2024): 2303–20. http://dx.doi.org/10.3390/make6040113.
Full textAbstract:
Accurately predicting financial entity performance remains a challenge due to the dynamic nature of financial markets and vast unstructured textual data. Financial knowledge graphs (FKGs) offer a structured representation for tackling this problem by representing complex financial relationships and concepts. However, constructing a comprehensive and accurate financial knowledge graph that captures the temporal dynamics of financial entities is non-trivial. We introduce FintechKG, a comprehensive financial knowledge graph developed through a three-dimensional information extraction process that incorporates commercial entities and temporal dimensions and uses a financial concept taxonomy that ensures financial domain entity and relationship extraction. We propose a temporal and relational graph convolutional network (RGCN)-based representation for FintechKG data across multiple timesteps, which captures temporal dependencies. This representation is then combined with FinBERT embeddings through a projection layer, enabling a richer feature space. To demonstrate the efficacy of FintechKG, we evaluate its performance using the example task of financial performance prediction. A logistic regression model uses these combined features and social media embeddings for performance prediction. We classify whether the revenue will increase or decrease. This approach demonstrates the effectiveness of FintechKG combined with textual information for accurate financial forecasting. Our work contributes a systematic FKG construction method and a framework that utilizes both relational and textual embeddings for improved financial performance prediction.
43
Saxena, Aradhana, A. Santhanavijayan, Harish Kumar Shakya, Gyanendra Kumar, Balamurugan Balusamy, and Francesco Benedetto. "Nested Sentiment Analysis for ESG Impact: Leveraging FinBERT to Predict Market Dynamics Based on Eco-Friendly and Non-Eco-Friendly Product Perceptions with Explainable AI." Mathematics 12, no. 21 (2024): 3332. http://dx.doi.org/10.3390/math12213332.
Full textAbstract:
In the current era, the environmental component of ESG is recognized as a major driver due to the pressing challenges posed by climate change, population growth, global warming, and shifting weather patterns. The environment must be considered a critical factor, and as evidenced by existing research, it is regarded as the dominant component within ESG. In this study, the ESG score is derived primarily from the environmental score. The increasing importance of the environmental, social, and governance (ESG) factors in financial markets, along with the growing need for sentiment analysis in sustainability, has necessitated the development of advanced sentiment analysis techniques. A predictive model has been introduced utilizing a nested sentiment analysis framework, which classifies sentiments towards eco-friendly and non-eco-friendly products, as well as positive and negative sentiments, using FinBERT. The model has been optimized with the AdamW optimizer, L2 regularization, and dropout to assess how sentiments related to these product types influence ESG metrics. The “black-box” nature of the model has been addressed through the application of explainable AI (XAI) to enhance its interpretability. The model demonstrated an accuracy of 91.76% in predicting ESG scores and 99% in sentiment classification. The integration of XAI improves the transparency of the model’s predictions, making it a valuable tool for decision-making in making sustainable investments. This research is aligned with the United Nations’ Sustainable Development Goals (SDG 12 and SDG 13), contributing to the promotion of sustainable practices and fostering improved market dynamics.
44
Stander, Yolanda S. "A News Sentiment Index to Inform International Financial Reporting Standard 9 Impairments." Journal of Risk and Financial Management 17, no. 7 (2024): 282. http://dx.doi.org/10.3390/jrfm17070282.
Full textAbstract:
Economic and financial narratives inform market sentiment through the emotions that are triggered and the subjectivity that gets evoked. There is an important connection between narrative, sentiment, and human decision making. In this study, natural language processing is used to extract market sentiment from the narratives using FinBERT, a Python library that has been pretrained on a large financial corpus. A news sentiment index is constructed and shown to be a leading indicator of systemic risk. A rolling regression shows how the impact of news sentiment on systemic risk changes over time, with the importance of news sentiment increasing in more recent years. Monitoring systemic risk is an important tool used by central banks to proactively identify and manage emerging risks to the financial system; it is also a key input into the credit loss provision quantification at banks. Credit loss provision is a key focus area for auditors because of the risk of material misstatement, but finding appropriate sources of audit evidence is challenging. The causal relationship between news sentiment and systemic risk suggests that news sentiment could serve as an early warning signal of increasing credit risk and an effective indicator of the state of the economic cycle. The news sentiment index is shown to be useful as audit evidence when benchmarking trends in accounting provisions, thus informing financial disclosures and serving as an exogenous variable in econometric forecast models.
45
Díaz Berenguer, Abel, Yifei Da, Matías Nicolás Bossa, Meshia Cédric Oveneke, and Hichem Sahli. "Causality-driven multivariate stock movement forecasting." PLOS ONE 19, no. 4 (2024): e0302197. http://dx.doi.org/10.1371/journal.pone.0302197.
Full textAbstract:
Our study aims to investigate the interdependence between international stock markets and sentiments from financial news in stock forecasting. We adopt the Temporal Fusion Transformers (TFT) to incorporate intra and inter-market correlations and the interaction between the information flow, i.e. causality, of financial news sentiment and the dynamics of the stock market. The current study distinguishes itself from existing research by adopting Dynamic Transfer Entropy (DTE) to establish an accurate information flow propagation between stock and sentiments. DTE has the advantage of providing time series that mine information flow propagation paths between certain parts of the time series, highlighting marginal events such as spikes or sudden jumps, which are crucial in financial time series. The proposed methodological approach involves the following elements: a FinBERT-based textual analysis of financial news articles to extract sentiment time series, the use of the Transfer Entropy and corresponding heat maps to analyze the net information flows, the calculation of the DTE time series, which are considered as co-occurring covariates of stock Price, and TFT-based stock forecasting. The Dow Jones Industrial Average index of 13 countries, along with daily financial news data obtained through the New York Times API, are used to demonstrate the validity and superiority of the proposed DTE-based causality method along with TFT for accurate stock Price and Return forecasting compared to state-of-the-art time series forecasting methods.
46
Dorosh, Volodymyr, Roman Vavryk, and Olena Stankevych. "Developing a Sentiment Analyzer Using ChatGPT for a Stock Market." Computer Design Systems. Theory and Practice 6, no. 1 (2024): 107–16. http://dx.doi.org/10.23939/cds2024.01.107.
Full textAbstract:
Today, an important problem of financial successis to find effective trading approaches that can adapt to rapidly changing market conditions and ensure high investment returns. Based on the literature analysis, ChatGPT is identified as a promising technology that is more effective than FinBert in being used as a component for conducting sentiment analysis of stocks. The research also shows satisfactory efficiency and productivity of ChatGPT. Existing sources do not provide a detailed description of the automation of the sentiment analysis process and testing of the ChatGPT model on big data. The purpose of the performed research is to develop an automated system for sentiment analysis based on ChatGPT with an integrated news aggregator for collecting and analyzing financial data.The study details the creation of a comprehensive solution sketch. A plan is presented that covers the entire range of the proposed system. A preliminary application architecture has been developed that provides a visual and structural representation of how the various components of the solution interact and function in a coordinated manner. This architectural plan serves as a roadmap for the implementation and deployment of the automated sentiment analyzer, ensuring clarity and accuracy in its design. Initial diagrams of the relationships between the entities in the system have been developed and an algorithm for the system has been proposed. Further research will focus on creating a minimum working system for the sentiment analyzer and evaluating its efficiency and quality of work.
47
Yan, Jin, and Yuling Huang. "MambaLLM: Integrating Macro-Index and Micro-Stock Data for Enhanced Stock Price Prediction." Mathematics 13, no. 10 (2025): 1599. https://doi.org/10.3390/math13101599.
Full textAbstract:
Accurate stock price prediction requires the integration of heterogeneous data streams, yet conventional techniques struggle to simultaneously leverage fine-grained micro-stock features and broader macroeconomic indicators. To address this gap, we propose MambaLLM, a novel framework that fuses macro-index and micro-stock inputs through the synergistic use of state-space models (SSMs) and large language models (LLMs). Our two-branch architecture comprises (i) Micro-Stock Encoder, a Mamba-based temporal encoder for processing granular stock-level data (prices, volumes, and technical indicators), and (ii) Macro-Index Analyzer, an LLM module—employing DeepSeek R1 7B distillation—capable of interpreting market-level index trends (e.g., S&P 500) to produce textual summaries. These summaries are then distilled into compact embeddings via FinBERT. By merging these multi-scale representations through a concatenation mechanism and subsequently refining them with multi-layer perceptrons (MLPs), MambaLLM dynamically captures both asset-specific price behavior and systemic market fluctuations. Extensive experiments on six major U.S. stocks (AAPL, AMZN, MSFT, TSLA, GOOGL, and META) reveal that MambaLLM delivers up to a 28.50% reduction in RMSE compared with suboptimal models, surpassing traditional recurrent neural networks and MAMBA-based baselines under volatile market conditions. This marked performance gain highlights the framework’s unique ability to merge structured financial time series with semantically rich macroeconomic narratives. Altogether, our findings underscore the scalability and adaptability of MambaLLM, offering a powerful, next-generation tool for financial forecasting and risk management.
48
Kim, Young Min, Changwan Kang, and Yong-Seok Choi. "Stock Price Prediction utilizing Sentiment Scores Based on Fin-BERT Models." Korean Data Analysis Society 27, no. 1 (2025): 31–43. https://doi.org/10.37727/jkdas.2025.27.1.31.
Full textAbstract:
Economic and technological advancements have led to a growing interest in stocks for many people. Investors want to make profits, which has led to the creation of many stock price prediction models using machine learning and deep learning models. Most studies have used news data to create stock price prediction models, but news data tends to be too neutral to accurately classify sentiment. In this study, we aim to create a stock price prediction model using NAVER Stock Discussion Room data, a community data where people of all ages can freely write opinions. In order to more accurately classify sentiment in the financial domain, we used KR-FinBERT, which is trained on Korean financial data, to perform sentiment classification. Since stock prices are most affected by recent data, we used a weighted moving average as a weight to create the final sentiment score. To check whether the sentiment scores generated from the NAVER stock discussion board data have an impact on stock price prediction, we compared the models generated from the data with sentiment scores to the models generated from the data without sentiment scores using three different analysis methodologies. Random Forest, XGBoost, and LSTM. The evaluation metrics used to compare the models were RMSE and MAE. The results of the analysis showed that the evaluation metrics of the data with sentiment scores were better in the model comparison using the three analysis methods, confirming that sentiment scores have an impact on stock price prediction.
49
Khan, Usama Waheed, Muhammad Bilal Saeed, and Aleena Nadeem. "Stock Price Prediction Model: Assessing the Performance of a Hybrid Deep Learning Model Employing Multi-Stream Data." NICE Research Journal 17, no. 1 (2024): 40–63. http://dx.doi.org/10.51239/nrjss.v17i1.459.
Full textAbstract:
Purpose - The study investigates the effectiveness of the ConvLSTM model in the next-day closing price prediction for stocks using a novel combination of input features. These features include past prices, prices of related stocks, technical indicators of the target stock, mutation point impact on closing price, stock market sentiment, stock market index, interest rate, and dollar exchange rate Study Design/Methodology/Approach - Sentiment analysis of financial news related to the Pakistan Stock Exchange (PSX) was performed using the pre-trained FinBERT model. Relevant stocks were identified through historical prices using Random Forests and XGBoost algorithms. A comparative analysis was conducted involving ConvLSTM1D and ConvLSTM2D models, as well as LSTM, CNN, and CNN+LSTM models, all having the same network structure. Findings - The study found that there is not a universal best model for every stock. However, the ConvLSTM1D model exhibited the best average performance for Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) across the selected stocks. Practical Implications- The findings suggest that while no single model is optimal for all stocks, the ConvLSTM1D model may be particularly effective for improving prediction accuracy in stock closing prices. This can aid investors and analysts in making more informed decisions based on the model's performance metrics. Originality/Novelty- This study provides a novel approach by integrating a diverse set of input features and comparing multiple models in a comprehensive manner. The use of sentiment analysis and advanced machine learning techniques such as ConvLSTM adds significant value to the field of stock price prediction.
50
Yang, Junwen, Yunmin Wang, and Xiang Li. "Prediction of stock price direction using the LASSO-LSTM model combines technical indicators and financial sentiment analysis." PeerJ Computer Science 8 (November 16, 2022): e1148. http://dx.doi.org/10.7717/peerj-cs.1148.
Full textAbstract:
Correctly predicting the stock price movement direction is of immense importance in the financial market. In recent years, with the expansion of dimension and volume in data, the nonstationary and nonlinear characters in finance data make it difficult to predict stock movement accurately. In this article, we propose a methodology that combines technical analysis and sentiment analysis to construct predictor variables and then apply the improved LASSO-LASSO to forecast stock direction. First, the financial textual content and stock historical transaction data are crawled from websites. Then transfer learning Finbert is used to recognize the emotion of textual data and the TTR package is taken to calculate the technical indicators based on historical price data. To eliminate the multi-collinearity of predictor variables after combination, we improve the long short-term memory neural network (LSTM) model with the Absolute Shrinkage and Selection Operator (LASSO). In predict phase, we apply the variables screened as the input vector to train the LASSO-LSTM model. To evaluate the model performance, we compare the LASSO-LSTM and baseline models on accuracy and robustness metrics. In addition, we introduce the Wilcoxon signed rank test to evaluate the difference in results. The experiment result proves that the LASSO-LSTM with technical and sentiment indicators has an average 8.53% accuracy improvement than standard LSTM. Consequently, this study proves that utilizing historical transactions and financial sentiment data can capture critical information affecting stock movement. Also, effective variable selection can retain the key variables and improve the model prediction performance.