Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/thisandagain/troll
Language sentiment analysis and neural networks... for trolls.
https://github.com/thisandagain/troll
javascript moderation neural-network nlp sentiment sentiment-analysis
Last synced: about 2 months ago
JSON representation
Language sentiment analysis and neural networks... for trolls.
- Host: GitHub
- URL: https://github.com/thisandagain/troll
- Owner: thisandagain
- License: other
- Created: 2012-10-02T18:40:40.000Z (almost 12 years ago)
- Default Branch: master
- Last Pushed: 2013-01-18T19:46:49.000Z (over 11 years ago)
- Last Synced: 2024-04-14T13:11:40.436Z (5 months ago)
- Topics: javascript, moderation, neural-network, nlp, sentiment, sentiment-analysis
- Language: JavaScript
- Size: 130 KB
- Stars: 330
- Watchers: 30
- Forks: 19
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
## troll
#### Language sentiment analysis and neural networks... for trolls.
[](http://travis-ci.org/thisandagain/troll)
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](http://redis.io/) for data storage. Once Redis is up and running, you can install Troll using NPM:
```bash
npm install troll
```
### String Analysis
```javascript
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:
```javascript
troll.train(function (err, result) {
console.dir(result); // { error: 0.0049931996067587685, iterations: 802 }
});
```
### User Classification
Once trained, now you can classify:
```javascript
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](https://github.com/thisandagain/sentiment) 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:
```javascript
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](http://redis.io/) connection settings are used:
```
TROLL_HOST: null
TROLL_PORT: null
TROLL_PASS: null
```
---
### Testing
```bash
npm test
```
### Credits
- Neural network by [Heather Arthur](https://github.com/harthur)
- Training data inferred and subsequently condensed by scraping [Boing Boing's](http://boingboing.net) reader comments.