Master's Thesis presentation, Colin Eriksson and Max Johansson

Abstract: Having extensive line-of-business classification of companies is of critical value for many private equity investors, yet mapping this market to a corporate taxonomy is tedious and complex if done manually. There are millions of companies in the world and many require specific domain knowledge in order to decipher their business.

This project aims to explore the automation of line-of-business classification by using natural-language processing (NLP) techniques to extract features from texts describing companies’ business practices.