search

huggingface / neuralcoref

✨Fast Coreference Resolution in spaCy with Neural Networks
https://huggingface.co/coref/
MIT License
2.84k stars 474 forks source link

The results are missing a lot of basic co-references ... #290

Open ohmeow opened 3 years ago

ohmeow commented 3 years ago

Hi there,

Running this against many trivial examples and it seems to miss obvious co-references ... here's an example of what I mean:

Input: "My dogs love the beach. They really love going there."

Results: The only co-reference it finds is "They" => "My dogs". But it completely misses that "there" => "the beach".

I'm noticing that the model seems to really have issues with correctly resolving what words like "it", "that", "this", and "there" refer too. Is this something with how the model was trained or is there something I can do to tune it to capture those words as well?

Thanks.

Hevia commented 3 years ago

Having similar issues, anything we can do to improve performance/fine-tune on our end?

stale[bot] commented 2 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.