GreenAI – An NLP approach to ESG financing

  • Nicolaas Ruberg BNDES
  • Rafael B. Pereira BNDES
  • Mauro L. Stein BNDES

Resumo


Environmental, Social, and Governance (ESG) factors are critical for investors and financing institutions like the Brazilian Development Bank (BNDES). Such institutions are currently working on setting up a framework to assess companies' ESG factors in their financing evaluation. In this study, we identify an opportunity to use Natural Language Processing (NLP) to improve the framework. This opportunity stems from the fact that the key documents for ESG analysis, such as the company's activity report (RAA), Environmental Impact Study (EIA), and Environmental Impact Report (RIMA), undergo manual screening and decomposition whilst being analyzed by specialists. By incorporating NLP, we aim to automate the classification of text passages from these reports and enhance the efficiency of the analysis process.

Referências

Bird, S., Klein, E., & Loper, E. (2009). Natural Language Processing with Python. O'ReillyO'Reilly (01 de 04 de 2021). BNDES Sustainability Bond Framework. Acesso em 16 de 11 de 2021, disponível em [link].

BNDES. (16 de 11 de 2021). Sustainable Development. Fonte: BNDES Homepage: [link]

Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., Krueger, G. (2020). Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems, 1877-1901.

Devlin, J., Chang, M., Lee, K., & Toutanova, K. (2018). BERT: Pretraining of Deep Bidirectional Transformers for Language Understanding. CoRR.

Dhar, A., Mukherjee, H., Dash, N., & Kaushik, R. (2021). Text categorization: past and present. Artificial Intelligence Review volume 54 (pp. 3007–3054). Springer.

Diksha Khurana, A. K. (2022 de 07 de 2022). Natural language processing: state of the art, current trends and challenges. Multimedia Tools and Applications, pp. 3713–3744.

Fang, X., Xu, M., & Xu, S. (2019). A deep learning framework for predicting cyber attacks rates. EURASIP Journal on Information Security.

Hildebrand, P., Polk, C., Deese, B., & Boivin, J. (02 de 2020). Sustainability: The tectonic shift transforming investing. Fonte: BlackRock: [link].

Inderst, G., & Stewart, F. (2018). Incorporating Environmental, Social and Governance Factors into Fixed Income Investment. Fonte: World Bank: [link].

Martin, J. H., & Jurafsky, D. (2008). Speech and Language Processing. Pearson International Edition.

Meager, E. (31 de 08 de 2021). Capital Monitor. Fonte: Capital Monitor AI: [link].

Napoletano, E., & Curry, B. (01 de 03 de 2021). FORBES Advisor. Fonte: FORBES: [link]

Ruberg, N. (3 de 12 de 2021). Bert goes sustainable: an NLP approach to ESG financing. Acesso em 30 de 01 de 2023, disponível em AMSLaurea Institutional Theses Repository: [link]

Seth, Y. (19 de 07 de 2019). BERT Explained – A list of Frequently Asked Questions. Fonte: A blog on data science, machine learning and artificial intelligence: [link]

Souza, F. a. (2020). BERTimbau: Pretrained BERT Models for Brazilian Portuguese. Intelligent Systems, pp. 403-417.

Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Polosukhin, I. (2017). Attention Is All You Need. Advances in Neural Information Processing Systems 30, pp. 5998–6008.

wikipedia. (03 de 02 de 2023). Bag-of-words model. Fonte: Wikipedia The Free Encyclopedia : [link]

Wikipedia. (03 de 02 de 2023). Global Reporting Initiative. Fonte: Wikipedia- The Free Encyclopedia: [link]

Wissler, L. &. (2014). The Gold Standard in Corpus Annotation. Passau, Germany: 5th IEEE Germany Student Conference.

Zhu, Y., Kiros, R., Zemel, R., Salakhutdinov, R., Urtasun, R., Torralba, A., & Fidler, S. (22 de 06 de 2015). Aligning Books and Movies: Towards Story-like Visual Explanations by Watching Movies and Reading Books. Fonte: arXiv.org: [link]
Publicado
06/08/2023
RUBERG, Nicolaas; PEREIRA, Rafael B.; STEIN, Mauro L.. GreenAI – An NLP approach to ESG financing. In: BRAZILIAN WORKSHOP ON ARTIFICIAL INTELLIGENCE IN FINANCE (BWAIF), 2. , 2023, João Pessoa/PB. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2023 . p. 37-48. DOI: https://doi.org/10.5753/bwaif.2023.229922.