node package manager



AFINN-based sentiment analysis for Node.js

Build Status Coverage Status Greenkeeper badge

Sentiment is a Node.js module that uses the AFINN-165 wordlist and Emoji Sentiment Ranking to perform sentiment analysis on arbitrary blocks of input text. Sentiment provides several things:

  • Performance (see benchmarks below)
  • The ability to append and overwrite word / value pairs from the AFINN wordlist
  • A build process that makes updating sentiment to future wordlists trivial


npm install sentiment


var sentiment = require('sentiment');
var r1 = sentiment('Cats are stupid.');
console.dir(r1);        // Score: -2, Comparative: -0.666
var r2 = sentiment('Cats are totally amazing!');
console.dir(r2);        // Score: 4, Comparative: 1

Adding / overwriting words

You can append and/or overwrite values from AFINN by simply injecting key/value pairs into a sentiment method call:

var sentiment = require('sentiment');
var result = sentiment('Cats are totally amazing!', {
    'cats': 5,
    'amazing': 2  
console.dir(result);    // Score: 7, Comparative: 1.75


A primary motivation for designing sentiment was performance. As such, it includes a benchmark script within the test directory that compares it against the Sentimental module which provides a nearly equivalent interface and approach. Based on these benchmarks, running on a MacBook Pro with Node v6.9.1, sentiment is twice as fast as alternative implementations:

sentiment (Latest) x 448,788 ops/sec ±1.02% (88 runs sampled)
Sentimental (1.0.1) x 240,103 ops/sec ±5.13% (81 runs sampled)

To run the benchmarks yourself:

make benchmark


While the accuracy provided by AFINN is quite good considering it's computational performance (see above) there is always room for improvement. Therefore the sentiment module is open to accepting PRs which modify or amend the AFINN / Emoji datasets or implementation given that they improve accuracy and maintain similar performance characteristics. In order to establish this, we test the sentiment module against three labelled datasets provided by UCI.

To run the validation tests yourself:

make validate

Rand Accuracy (AFINN Only)

Amazon:  0.70
IMDB:    0.76
Yelp:    0.67

Rand Accuracy (AFINN + Additions)

Amazon:  0.72 (+2%)
IMDB:    0.76 (+0%)
Yelp:    0.69 (+2%)


npm test