Closed metazool closed 3 years ago
I know you've got a lot of other work on your stack @mobiuscreek and I wasn't planning to haul you into this, however @rachelheaven is unavailable for the next week or 2, so if you had half an hour minutes to cast your eyes over this PR before it builds up into too much of a pile I'd appreciate that. I can return the favour
Most of the code you've seen before while looking at the changes to entity-resolver
- I know there's a lot that could be improved (from the string closeness methods to the way the results are passed around) but this is the smallest step to move forwards with. I got CoreNLP Server working as a docker service for the tests. It was the right call sticking with it for now, thanks.
Thank you for sending it my way. Just by skimming through I see what you mean with the string matching but this improvement isn't for this PR. I'll have a look either later today or tomorrow.
Thank you, appreciate it, not really expecting more than a skim through
setup.py
wasn't installing the dependencies so I added them and I also added versioneer for automatic code version. I hope it doesn't change a lot. Everything else is fine.
This is more or less as-is the existing code from the
entity-resolver
project that provides an interface to CoreNLP Server's NER and matches names by Levenshtein distance