Sentiment Analysis is the process of determining whether a piece of writing is positive, negative or neutral. [TBC]

We use positive and negative indicators from an VADER sentiment lexicon, which is is sensitive both the **polarity** and the **intensity** of sentiments expressed in social media contexts. But it is also generally applicable to sentiment analysis in other domains.

more information about the algorithm can be found in this paper or in this Python package site

See demo here: Developer or Instructor

To use this web service, you can either POST a JSON with the following format:

{
    "review":"review text...."
}

or

{
    "sentences": ["sentence 1", "sentence 2",...]
}

It returns an overall sentiment of the review, which is called “overall_compound”. Additionally, it returns the sentiment of each sentence as well as the individual polarity (negative, neutral, positive). The service response is formatted in JSON format, as the following example shows:

  {
      "overall_compound": -1.0,
      "sentiments": {
        "After a 20 mile ride, both hands had sores because of this.": {
          "compound": 0.0,
          "neg": 0.0,
          "neu": 1.0,
          "pos": 0.0
        }, {
        ....
        }
    }