computational antitrust

I believe that there is an immensely rich, and largely unexplored source of information in unstructured textual data that would be hugely useful for applied microeconomic works, including my own area of IO and competition policy.