Entity matching with both spacy and HF NER model
#1
by
nickmuchi
- opened
Really good demo, have adopted the entity matching into my article text summarizer as well. Wondering why you used both spacy and HF NER model to determine the named entities?
Hello, sounds cool! We decided to combine 2 NER models because at the time of implementation, we found that the resulting list of entities was (slightly) better than when you would get using a single model.
Makes sense, thank you! Also noticed a few issues with highlighting entities that are numeric, the raw markdown is generated instead:
Summary
Summary with highlighted entities
