Add basic text data augmentation functionality

[bdewilde]

Description

• Add a sub-package (textacy.augmentation) for basic text data augmentation; implements several transformations suitable for use in text classification tasks, with a higher-level function textacy.augmentation.apply() to call them
  • random synonym replacement
  • random synonym insertion
  • random item swapping
  • random item deletion
  • random sentence shuffling
  • Note: This code is provisional, and the API will almost definitely be changing.

Motivation and Context

I've been training spaCy TextCategorizer s on datasets that are too small, and data augmentation is a great way to improve model performance.

How Has This Been Tested?

Lots of manual validation and trial-by-error. Wrote some tests, and they pass (mostly...).

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• My code follows the code style of this project.
• [TODO] My change requires a change to the documentation, and I have updated it accordingly.