This website demos the NLP model created from my artificial intelligence project class.

The model is trained on Stanford's Wikipedia Neutrality Corpus.

Find out more about the project here: https://aidenywl.github.io/neutbias/

1. Obtain the extensions:

• Prettier: Code formatter, configured via .prettierrc
• TSLint: TS linter, configured via .tslint.json
• Stylelint: CSS linter, configured via .stylelintrc
• PostCSS syntax: Enable postCSS syntax to use Javascript variables inside CSS files.

2. Update workspace's settings.json with this:

   "editor.formatOnSave": true,

3. Reload your editor

1. Pull the repository
2. Run yarn to install all packages.
3. Run yarn start and go to your browser to see the starting website.

1. There are deployment scripts in package.json.
2. Simply run yarn deploy to deploy to github pages with gh-pages.

This website extends the typescript react boilerplate I've written here.

• Typescript modules and all relevant react modules for development with typescript are installed.
• Webpack is configured to build and serve Typescript and CSS files.
• Babel is set up for ES6 syntax.
• CSS Loader and POSTCSS Loader is used to shift away from global css and localized css for each component.
• Further CSS set-up: autoprefixer and autoprefixer normalizes styles for each browser and takes away the trouble of writing specific css classes for each browser
• Redux Saga and Redux is used to abstract away the data layer and api calls.

Although Jest is set up, minimal tests have been written for this static page.

• Jest is set up and ready to go via yarn test.
• React Testing Library is used to write tests instead of enzyme for flexibility and being able to better simulate the DOM as seen by a user.
• Using RTL allows us to move away from snapshot testing and test each component mocking components as required.

The backend is running on Flask on a GCP compute server serving the OpenNMT-py model trained on the corpus data.