Provides a UI for training a naive Bayes classifier on a set of text items.
Takes a list of items to classify, options by which to classify them, and presents the user with a UI for classifying.
When the user is done, they can export their classifier settings to JSON.
- Install dependencies with 'npm install'.
- Put your JSON array of data to classify in conf/data.json. There is a set of Trump Tweets there as an example.
- In src/Constants, replace the OPTIONS object with the categories for classification.
- Run the app with 'npm start'.
- In your browser, visit http://localhost:3000
- Classify some text.
- Profit!
I included a script for converting Tweets to text items. Invoke it with:
npm run convert-tweets -- /path/to/tweets.json