One hypothesis we could make is that these companies, or at least a good part of them, are from the same industry. Is that true?
We actually have the data to figure this out. If we look back at our information dataset, we can easily see that some of the records start with the industry token. These are the records reproducing the line related to the customer's industry, which is contained within every customer card.
Let's filter out all the other records to retain only those records that specify the industry of the company:
information %>% filter(grepl("industry", text))
This is fine; nonetheless, ...
