Provides positive, negative, or neutral classification for a sample of English language text.

Current version is based off of Natural, the node-based NLP library.

There are two current versions of this application. One deployed through Vercel which can be found here

The other version can be found in the directory titled node-nlp-api which can be run locally as an API on your local machine. Most tools associated with this project are written in Nodejs or bash.

git clone https://gitlab.com/zoverlvx/sentiment-analysis-engine.git

cd sentiment-analysis-engine/node-nlp-api && npm i

npm start 

To request sentiment analysis from the API
Send POST to localhost:PORT/api/nlp
PORT defaults to 3000 if PORT is not in environment

Request

{
    data: String|String[] 
}

Response

{
    sentiment: Sentiment|Sentiment[]
}

Example Request:

curl -X POST \
-H "Content-Type: application/json" \
-d '{"data": ["good", "bad", "beautiful"]}'\
localhost:3000/api/nlp

This API is currently equipped with stress tests. The script test will use text examples from lockdown.txt in order to test the response rate of the API.

• 12 Factor
• 12 factor slides
• twelve-factor methodology used at cloud.gov
• Ben Awad on CI/CD clean-up video
• Spacy
• Natural Language Processing