I am looking for rule-based methods or any other methods to identify all mentions in a text. I have found several libraries that give coreferences but no exact options for only mentions. What I want is something like below:

Input text:

[This painter]'s indulgence of visual fantasy, and appreciation of different historic architectural styles can be seen in his 1840 Architect's Dream. After a series of paintings on The Last of the Mohicans, [he] made a three year trip to Europe in 1829, but [he] is better known for a trip four years earlier in which [he] journeyed up the Hudson River to the Catskill Mountains. FTP, name [this painter of The Oxbow] and The Voyage of Life series.

*The square brackets highlight mentions.

Please suggest a way to deal with this. Also, how is it different from coreferences? It would be really helpful if someone could post links to the concerned papers.

  • 3
    This is coreference. It is exactly what coreference resolution is intended to do. "This painter" coreferences "he", etc. – Adam_G May 15 '15 at 21:16
  • Thank you Adam_G for your reply. But my problem is little different as I want all the mentions. Does coreference gives all the mentions? Actually when I tried Stanford core nlp coreference module, it was giving me only co-referent mentions and not all. – kahini wadhawan May 16 '15 at 21:06
  • 3
    It is extremely difficult to understand what you're looking for. Coreference resolution is a hard task, and is not 100% robust. I'd encourage you to do some reading about the topic, and then submit a new question. – Adam_G May 16 '15 at 21:11

Does "any other method" include manual work? Then I recommend to train some students to annotate the coreference chains. Take care that every text is annotated twice by different annotators, and compute the inter-annotator agreement (e.g., by determining Cohen's kappa).

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.