troll
Language sentiment analysis and neural networks... for trolls.
Troll is a tool for performing sentiment analysis (ie: "is this naughty or nice") on arbitrary blocks of text and associating it with a unique user. Using this data, combined with a rather naïve neural network and some training data, users can be indentified as "trolls".
Installation
Troll uses Redis for data storage. Once Redis is up and running, you can install Troll using NPM:
npm install troll
String Analysis
var troll = require('troll');
troll.analyze('This is totally awesome!', 'user123', function (err, result) {
console.log(result); // 2.5
});
troll.analyze('This is lame.', 'user456', function (err, result) {
console.log(result); // -1.333
});
Training
Before attempting to classify a user, you'll need to train Troll. You can specify your own training data or use a basic set that is included. To load the included training set:
troll.train(function (err, result) {
console.dir(result); // { error: 0.0049931996067587685, iterations: 802 }
});
User Classification
Once trained, now you can classify:
troll.classify('user123', function (err, result) {
console.dir(result); // { total: 9, sum: 36, troll: 0.010294962292857838 }
});
The value returned for the troll
key represents the probability of that user being a troll. In other words, a value of 0
would likely represent a particularly friendly user, while a value of 1
would be... uh, Ted Dziuba?
Modifing The Sentiment Engine
The underlying sentiment analysis module supports "injecting" additional key/value pairs. This is useful in certain situations where you may want to exclude or even blacklist certain words based on a particular use case. For example:
troll.inject({
'stupid': -5,
'lame': -5
});
troll.analyze('This is totally stupid and lame!', 'user123', function (err, result) {
console.log(result); // -5.833
});
Redis Connection Options
Troll uses your environment by looking at process.env
for connection settings. If none are found, default Redis connection settings are used:
TROLL_HOST: null
TROLL_PORT: null
TROLL_PASS: null
Testing
npm test
Credits
- Neural network by Heather Arthur
- Training data inferred and subsequently condensed by scraping Boing Boing's reader comments.