-------------
Features
V 1.0.4 - Origin
- Classify text as hate speech.
- Send items for prediction
- Access raw output from model server consisting of the probability array and basic aggregate stats
- Filter results and return only 'hate' 'offensive' or 'neither ' results.
- Group results by Class
- Map extra data to prediction results
- Submit reclassed text items for use in further model training.
- Submit reclassed items.
- Create a HTML form template for item reclassification in a web environment.
- Use library function to parse the template form data.
- Twitter/Tweepy API handling.
- Make request for tweets related to a specific keyword.
- Make request for tweets from a specific user's timeline.
- Modify count of tweets to include in query results.