The annual society of toxicology conference has 288 company booths. We built a phylogeny of those companies. Companies are clustered by similarity of their products and descriptions. Example products and companies are given for each cluster.
How was this built?
A Sysrev.com Doc2Vec model was used to transform company descriptions into vectors (blog.sysrev.com/machine-learning) which were then compared with cosine similarity. The APE R package was used to build the phylogenies.
Scroll below for a better view of the individual companies.
Contact me at firstname.lastname@example.org if you want access to this data.