Towards Generating Financial Reports from Tabular Data Using Transformers

Author: C. L. Chapman, L. Hillebrand, M. R. Stenzel, T. Deusser, D. Biesner, C. Bauckhage, R. Sifa
Journal: CD-MAKE 2022: Machine Learning and Knowledge Extraction
Year: 2022

Citation information

C. L. Chapman, L. Hillebrand, M. R. Stenzel, T. Deusser, D. Biesner, C. Bauckhage, R. Sifa,
CD-MAKE 2022: Machine Learning and Knowledge Extraction,
2022,
221-232,
https://doi.org/10.1007/978-3-031-14463-9_14

Financial reports are commonplace in the business world, but are long and tedious to produce. These reports mostly consist of tables with written sections describing these tables. Automating the process of creating these reports, even partially has the potential to save a company time and resources that could be spent on more creative tasks. Some software exists which uses conditional statements and sentence templates to generate the written sections. This solution lacks creativity and innovation when compared to recent advancements in NLP and deep learning. We instead implement a transformer network to solve the task of generating this text. By generating matching pairs between tables and sentences found in financial documents, we created a dataset for our transformer. We were able to achieve promising results, with the final model reaching a BLEU score of 63.3. Generated sentences are natural, grammatically correct and mostly faithful to the information found in the tables.