End-to-end Neural Coreference Resolution

This is the demo for the coreference resolution system described in:

End-to-end Neural Coreference Resolution
Kenton Lee, Luheng He, Mike Lewis and Luke Zettlemoyer
In Proceedings of the Conference on Empirical Methods in Natural Language Process (EMNLP), 2017.

Submit an input document below. A simple example is provided.

On the right side, the highest-scoring spans are shown. On the left side, the predicted coreference cluster (mentions of the same entity) are shown.

Hover over the mentions on either side to highlight their location in the document and render the attention weights of the words.

The code is available here.

Input Document

     Render attention weights on hover