node-nlp
Version:
Library for NLU (Natural Language Understanding) done in Node.js
212 lines (188 loc) • 8.27 kB
HTML
<html lang="en">
<head>
<meta charset="utf-8">
<title>JSDoc: Source: sentiment/sentiment-analyzer.js</title>
<script src="scripts/prettify/prettify.js"> </script>
<script src="scripts/prettify/lang-css.js"> </script>
<!--[if lt IE 9]>
<script src="//html5shiv.googlecode.com/svn/trunk/html5.js"></script>
<![endif]-->
<link type="text/css" rel="stylesheet" href="styles/prettify-tomorrow.css">
<link type="text/css" rel="stylesheet" href="styles/jsdoc-default.css">
</head>
<body>
<div id="main">
<h1 class="page-title">Source: sentiment/sentiment-analyzer.js</h1>
<section>
<article>
<pre class="prettyprint source linenums"><code>/*
* Copyright (c) AXA Shared Services Spain S.A.
*
* Permission is hereby granted, free of charge, to any person obtaining
* a copy of this software and associated documentation files (the
* "Software"), to deal in the Software without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Software, and to
* permit persons to whom the Software is furnished to do so, subject to
* the following conditions:
*
* The above copyright notice and this permission notice shall be
* included in all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
* LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
* OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
* WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*/
const NlpUtil = require('../nlp/nlp-util');
/**
* Class for a Sentiment Analyzer.
* Sentiment analysis can use 3 different type of files:
* - AFINN
* - Senticon
* - Pattern
*/
class SentimentAnalyzer {
/**
* Constructor of the class.
* @param {Object} settings Settings to initialize the instance.
*/
constructor(settings) {
this.settings = settings || {};
if (!this.settings.language) {
this.settings.language = 'en';
}
if (!this.settings.tokenizer) {
this.settings.tokenizer = NlpUtil.getTokenizer(this.settings.language);
}
if (!this.settings.type) {
switch (this.settings.language) {
case 'en':
this.settings.type = 'senticon';
break;
case 'es':
this.settings.type = 'senticon';
break;
case 'it':
this.settings.type = 'pattern';
break;
case 'fr':
this.settings.type = 'pattern';
break;
case 'nl':
this.settings.type = 'pattern';
break;
case 'de':
this.settings.type = 'senticon';
break;
default:
this.settings.type = 'senticon';
}
}
if (!SentimentAnalyzer.loadedFiles) {
SentimentAnalyzer.loadedFiles = {};
}
this.loadVocabulary();
}
/**
* Load the vocabulary and negation files based on the type of files and language.
*/
loadVocabulary() {
this.vocabularyFileName = `./languages/${this.settings.language}/${
this.settings.type
}_${this.settings.language}.json`;
this.negationFileName = `./languages/${this.settings.language}/negations_${
this.settings.language
}.json`;
try {
// eslint-disable-next-line
this.vocabulary = require(this.vocabularyFileName);
} catch (ex) {
this.vocabulary = undefined;
}
if (
this.vocabulary &&
!SentimentAnalyzer.loadedFiles[this.vocabularyFileName]
) {
SentimentAnalyzer.loadedFiles[this.vocabularyFileName] = true;
if (
this.settings.type === 'senticon' ||
this.settings.type === 'pattern'
) {
const propertyName =
this.settings.type === 'senticon' ? 'pol' : 'polarity';
Object.keys(this.vocabulary).forEach(word => {
this.vocabulary[word] = Number.parseFloat(
this.vocabulary[word][propertyName]
);
});
}
}
try {
// eslint-disable-next-line
this.negations = require(this.negationFileName).words || [];
} catch (ex) {
this.negations = [];
}
}
/**
* Given an utterance, return the sentiment analysis of the utterance.
* @param {String} utterance Utterance to be analyzed.
*/
getSentiment(utterance) {
const words = Array.isArray(utterance)
? utterance
: this.settings.tokenizer.tokenize(utterance);
if (!this.vocabulary) {
return {
score: 0,
numWords: words.length,
numHits: 0,
comparative: 0,
type: this.settings.type,
language: this.settings.language,
};
}
let score = 0;
let negator = 1;
let nrHits = 0;
words.forEach(token => {
const lowerCased = token.toLowerCase();
if (this.negations.indexOf(lowerCased) !== -1) {
negator = -1;
nrHits += 1;
} else if (this.vocabulary[lowerCased] !== undefined) {
score += negator * this.vocabulary[lowerCased];
nrHits += 1;
}
});
return {
score,
numWords: words.length,
numHits: nrHits,
comparative: score / words.length,
type: this.settings.type,
language: this.settings.language,
};
}
}
module.exports = SentimentAnalyzer;
</code></pre>
</article>
</section>
</div>
<nav>
<h2><a href="index.html">Home</a></h2><h3>Classes</h3><ul><li><a href="BinaryNeuralNetworkClassifier.html">BinaryNeuralNetworkClassifier</a></li><li><a href="Classifier.html">Classifier</a></li><li><a href="ConversationContext.html">ConversationContext</a></li><li><a href="DutchStemmer.html">DutchStemmer</a></li><li><a href="EnglishStemmer.html">EnglishStemmer</a></li><li><a href="EnumNamedEntity.html">EnumNamedEntity</a></li><li><a href="Evaluator.html">Evaluator</a></li><li><a href="HungarianStemmer.html">HungarianStemmer</a></li><li><a href="ItalianStemmer.html">ItalianStemmer</a></li><li><a href="Language.html">Language</a></li><li><a href="LogisticRegressionClassifier.html">LogisticRegressionClassifier</a></li><li><a href="Matrix.html">Matrix</a></li><li><a href="MemoryConversationContext.html">MemoryConversationContext</a></li><li><a href="NamedEntity.html">NamedEntity</a></li><li><a href="NerManager.html">NerManager</a></li><li><a href="NlgManager.html">NlgManager</a></li><li><a href="NlpClassifier.html">NlpClassifier</a></li><li><a href="NlpManager.html">NlpManager</a></li><li><a href="NorwegianStemmer.html">NorwegianStemmer</a></li><li><a href="PortugueseStemmer.html">PortugueseStemmer</a></li><li><a href="Recognizer.html">Recognizer</a></li><li><a href="RegexNamedEntity.html">RegexNamedEntity</a></li><li><a href="RomanianStemmer.html">RomanianStemmer</a></li><li><a href="RussianStemmer.html">RussianStemmer</a></li><li><a href="SentimentAnalyzer.html">SentimentAnalyzer</a></li><li><a href="SentimentManager.html">SentimentManager</a></li><li><a href="SimilarSearch.html">SimilarSearch</a></li><li><a href="SlotManager.html">SlotManager</a></li><li><a href="StemmerJa.html">StemmerJa</a></li><li><a href="SwedishStemmer.html">SwedishStemmer</a></li><li><a href="Tokenizer.html">Tokenizer</a></li><li><a href="TrimNamedEntity.html">TrimNamedEntity</a></li><li><a href="TurkishStemmer.html">TurkishStemmer</a></li><li><a href="Vector.html">Vector</a></li><li><a href="XTable.html">XTable</a></li></ul><h3>Global</h3><ul><li><a href="global.html#endsinArr">endsinArr</a></li><li><a href="global.html#prelude">prelude</a></li><li><a href="global.html#regions">regions</a></li><li><a href="global.html#stem">stem</a></li><li><a href="global.html#stopwords">stopwords</a></li></ul>
</nav>
<br class="clear">
<footer>
Documentation generated by <a href="https://github.com/jsdoc3/jsdoc">JSDoc 3.5.5</a> on Sat Oct 13 2018 19:14:51 GMT+0200 (CEST)
</footer>
<script> prettyPrint(); </script>
<script src="scripts/linenumber.js"> </script>
</body>
</html>