bert-similarity

This is a repository with materials used in a seminar I lectured on July 4th in University of Alcala

It gives an introduction on how to use AI applied to M&A in order to get peers, or similar companies. It runs with sentence transformers library. Structure of notebook has been taken from this article of James Briggs https://towardsdatascience.com/bert-for-measuring-text-similarity-eec91c6bf9e1

Also a small notebook on how to assess similarity among images is uploaded, thanks to Ben Chamblee https://towardsdatascience.com/what-is-cosine-similarity-how-to-compare-text-and-images-in-python-d2bb6e411ef0

Both cases are a first approach. Data is dummy in case of companies, although you can substitute with your dataset, preferably using pickle instead of csv in case is a big number of registers.

Hope anyone can find it useful!