Tools
Tools
Khan, Saif Ali (2023) Investigating the Application of Natural Language Processing in Analysing Sentiment within Financial News. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (1MB) | Preview |
Abstract
Making informed financial decisions requires an understanding of market mood and public opinion. Natural Language Processing (NLP) is a strong method for extracting sentiment from massive amounts of text data, revealing market sentiments. More research, however, is required to properly investigate the effectiveness and applicability of NLP in the financial sector.
This study tries to fill these gaps by comparing NLP-based sentiment analysis to traditional market indicators for predicting stock price fluctuations. It also looks into the viability of incorporating domain-specific financial data and ontologies into NLP models in order to improve accuracy and interpretability.
The study addresses the challenges of sentiment analysis using NLP approaches and suggests practical solutions to improve its effectiveness. It highlights the importance of understanding market sentiment and investor behavior during economic events and crises, as well as its impact on short-term market volatility and price swings.
The findings demonstrate the use of NLP-based sentiment analysis in financial news, demonstrating its superiority to traditional indicators in predicting stock price fluctuations. Furthermore, Our comprehensive exploration of sentiment analysis and topic modeling in the financial domain yielded notable insights. Random Forest emerged as a strong performer, achieving 86.52% accuracy in sentiment analysis. SVM also demonstrated commendable accuracy at 85.8%. LSTM faced challenges, while LDA showcased subject modeling prowess and LSA explained 24% of the dataset's variation. Random Forest and SVM's robust performances validate their significance in sentiment prediction. Looking forward, addressing the challenges faced by LSTM, incorporating temporal dynamics, and exploring advanced models can further enhance applicability and effectiveness in the financial realm. Integrating domain-specific financial data with natural language processing (NLP) models enhances accuracy and interpretability.
This study highlights the effectiveness of NLP-driven sentiment analysis in financial decision-making, particularly during economic downturns. While it provides valuable insights, there are opportunities for further research to improve NLP models and incorporate new data sources.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Muntean, Cristina Hava UNSPECIFIED |
| Subjects: | H Social Sciences > HG Finance Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing H Social Sciences > Economics |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 14 May 2025 13:24 |
| Last Modified: | 14 May 2025 13:24 |
| URI: | https://norma.ncirl.ie/id/eprint/7546 |
Actions (login required)
 |
View Item |