Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Feature: This PR introduces a new transformer that scores utterances with questions (0 otherwise).
The BERT similarity method computes the degree of openness of a given question by analyzing how "diverse" the reply can be, given the question. It uses BM25 method to look for 10 different questions that are the most similar to the question at hand. Then, it will compute the cosine similarity of the responses that these 10 questions received. It hinges upon the idea that the more open a question is, there is a multiple ways to answer the question; the more closed a question is, there is a fixed way that a respondent can reply -- A high similarity score will indicate closedness of the question.
The PR introduces the new transformer that accomplishes this task.
Motivation and Context
This new transformer aims to capture and score the openness of questions.
How has this been tested?
It has been tested locally through the demo.
Other information
This was made in part of the requirement of A8 of INFO 4350.