-, Ayush (2024) Hindi FinBERT: A Pre-trained Language Model for Financial Text Classification. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (916kB) | Preview |
Preview |
PDF (Configuration manual)
Download (967kB) | Preview |
Abstract
This research proposes Hindi language FinBERT, a pre-trained language model designed specifically for financial based text sentiment analysis in Hindi. It addresses the need for a specialized tool which is elegant and capable of understanding the intricacies of financial language for Hindi context, a niche that happens to be underrepresented in the realm of language modelling. The development of the very Hindi FinBERT was driven by the gap in existing infrastructure for language models which were not adequately tuned to handle the unique aspects of financial texts for Hindi. Given The significance of the financial sector in India and the dominance of Hindi as language in the space, there was a clear need for a model that could accurately interpret financial terminology with context in this language Hindi FinBERT was meticulously trained and designed using a large corpus of financial documents, including reports, news, and statements. There is assurance that the model is fine tuned to the linguistic subtleties and technicalities of the financial domain that exists within the Hindi language framework. Upon evaluation, Hindi FinBERT demonstrated. superior performance in tasks like sentiment analysis and when compared to general-purpose Hindi language models This underlines its enhanced outcomes and capability in accurately understanding and analysing the very financial texts. This model would prove to be a significant contribution to the field of language processing, particularly. in the finance generic domain. It not only bridges a crucial linguistic gap but also enhances our understanding of how niche-specific language models can be developed and optimized For practitioners. In the financial sector, Hindi FinBERT offers a powerful tool for analysing financial texts with greater accuracy with cultural relevance. It holds substantial potential for long horizon application in various financial analyses and decision-making. processes within Hindi-speaking countries' markets. While Hindi FinBERT thus marks a substantial advancement, though there are areas for further research and exploration, which could be as extending its application to even more diverse financial datasets or adapting. It's very architecture for other sectors. tasks within the language universe.
Item Type: | Thesis (Masters) |
---|---|
Supervisors: | Name Email Haque, Rejwanul UNSPECIFIED |
Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing |
Divisions: | School of Computing > Master of Science in Artificial Intelligence |
Depositing User: | Tamara Malone |
Date Deposited: | 03 Apr 2025 16:24 |
Last Modified: | 03 Apr 2025 16:24 |
URI: | https://norma.ncirl.ie/id/eprint/7357 |
Actions (login required)
![]() |
View Item |