Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add measures #18

Open
yhamoudi opened this issue Oct 29, 2014 · 3 comments
Open

Add measures #18

yhamoudi opened this issue Oct 29, 2014 · 3 comments

Comments

@yhamoudi
Copy link
Member

Add measures of: https://github.com/ProjetPP/Documentation/blob/master/module-communication.md#measures

@Ezibenroc
Copy link
Member

Not important yet. The core might use the measure in the future to choose between the triple produced by NLP-classical and NLP-ML, but NLP-classical will be the only NLP module for the midterm.

Do not focus on this issue for the moment.

@yhamoudi
Copy link
Member Author

Just some ideas:

  • attribute a score to each dependency relation (for ex: dep relation is not good because it means the stanford parser failed to parse the whole sentence)
  • attribute a score to each triple production rule
  • attribute a score to each question word (for ex: yes/no questions are very bad for the moment)
  • attribute a score to the final result (depth of the tree, number of unknowns)
  • attribute a score to the merging operations
  • take into account the score given by users (or it's the role of the core?)

Mix all this scores together to obtain the measures

@Ezibenroc
Copy link
Member

Another idea:

  • Reduce the score if predicates were obtained with nounificationAuto.pickle

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants