1.2.0

• Change structure of corpus module #102

1.1.4

• Use glob instead of walk to get files #101

1.1.3

Update version to 1.1.3

1.1.2

• Switch from ruamel.yaml to PyYAML #57
• Various corpus contributions.

1.1.0

• Additional dialog contributions from contributors
• Removed math and swear words data files.

1.0.1

Bug fixes

• Fix where category is stored #39

Testing

• Add line length test to ensure all dialog is under 400 characters #36

1.0.0

Breaking changes

• Change format of corpus files from json to yml #25

Features

• Allow corpus name to be accessed as a category #26
  • Add support for multiple category tags for a corpus #28

Testing

• Add test cases to verify that file paths to custom corpora can be specified #27

0.0.2

• Various corpus contributions

0.0.1

First Release of the ChatterBot Corpus

This is the first release of this dialog corpus module which was originally a part of the ChatterBot package.

Splitting this data out into its own module is intended to improve the rate at which new training data can be released to ChatterBot users (because chatterbot-corpus can now be updated independently of the chatterbot module).

In addition, this change helps to make the corpus dialog data more widely available to other projects. As this project moves forward I am hoping to make this collection of data as useful as possible to the community of Python developers, and possibly even other programming languages as well.

Additional changes

• Addition of a Bangla corpus from @orinify #10