Improvement of the recovery of Pragmatic Segmenter sentence segmentation text wrt to the original text offsets #701
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Outdated suggestions cannot be applied.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.
The pragmatic segmenter seems to modify the output string which makes the extraction of the sentence offsets more complicated.
This PR makes two changes:
Recovery of segmentation offset using Pragmatic Segmenter
This implementation uses an external library to compute the diff of the sentence and the original text.
For example
original = "This is the original text. Some spaces are going to be removed. sentence = "This is the original text."
The idea is to get the starting and ending within
original
taking as referencesentence
.After translating the diff in a char based string, it compute the starting and ending of the sentence within the text.
The diff is a list of string where each element is structured as [operation, space, character], having
(e.g. comparing the string
operation
=+
,-
ora
andab
would result in the following:diff = [' a', '- b']
After the diff is computed using a two-pass heuristic:
The heuristic also is limited to a subset of the string, should a sentence has been identified before.