Information Extraction: An application to the domain of hyper-local financial data on developing countries
Keywords: named entity recognition, computational finance, information extraction, large language models
Abstract: Despite the need for financial and economic data in developing countries for development research and economic analysis, such data remains limited and silo-ed. In this paper, we develop and evaluate two Natural Language Processing (NLP) based approaches to address this issue. First, we curate a custom dataset specific to the domain of financial text data on developing countries and explore multiple approaches for information extraction. We then explore a text-to-text approach leveraging the transformer-based T5 model with the goal of undertaking simultaneous Named Entity Recognition and relation extraction. We find that this model is able to learn the custom text structure output data corresponding to the entities and their relations, resulting in an accuracy of 92.44\%, a precision of 68.25\% and a recall of 54.20\% from our best T5 model on the combined task. Secondly, we explore an approach with sequential NER and relation extration. For the NER, we run pre-trained and fine-tuned models using SpaCy, and wee develop a custom relation extraction model using SpaCy's Dependency Parser output and some heuristics to determine entity relationships. We obtain an accuracy of 84.72\%, a precision of 6.06\% and a recall of 5.57\% on this sequential task. Overall, LLMs such as T5 show tremendous promise in bridging the identified data gap.
Submission Number: 23
Loading