If you're looking for the version 3 docs, you can find them here Version 3
"NLP.js" is a general natural language utility for nodejs. Currently supporting:
!
Version 4 is very different from previous versions. Before this version, NLP.js was a monolithic library. The big changes:
normalize -> tokenize -> removeStopwords -> stem -> arrToObj
hear
and say
. Examples of connectors included: Console Connector, Microsoft Bot Framework Connector and a Direct Line Offline Connector (this one allows you to build a web chatbot using the Microsoft Webchat, but without having to deploy anything in Azure).If you're looking to use NLP.js in your Node application, you can install via NPM like so:
npm install node-nlp
There is a version of NLP.js that works in React Native, so you can build chatbots that can be trained and executed on the mobile even without the internet. You can install it via NPM:
npm install node-nlp-rn
Some limitations:
You can see a great example of use in the folder /examples/02-qna-classic
. This example is able to train the bot and save the model to a file, so when the bot is started again, the model is loaded instead of being trained again.
You can start to build your NLP from scratch with a few lines:
const { NlpManager } = require('node-nlp');
const manager = new NlpManager({ languages: ['en'], forceNER: true });
// Adds the utterances and intents for the NLP
manager.addDocument('en', 'goodbye for now', 'greetings.bye');
manager.addDocument('en', 'bye bye take care', 'greetings.bye');
manager.addDocument('en', 'okay see you later', 'greetings.bye');
manager.addDocument('en', 'bye for now', 'greetings.bye');
manager.addDocument('en', 'i must go', 'greetings.bye');
manager.addDocument('en', 'hello', 'greetings.hello');
manager.addDocument('en', 'hi', 'greetings.hello');
manager.addDocument('en', 'howdy', 'greetings.hello');
// Train also the NLG
manager.addAnswer('en', 'greetings.bye', 'Till next time');
manager.addAnswer('en', 'greetings.bye', 'see you soon!');
manager.addAnswer('en', 'greetings.hello', 'Hey there!');
manager.addAnswer('en', 'greetings.hello', 'Greetings!');
// Train and save the model.
(async() => {
await manager.train();
manager.save();
const response = await manager.process('en', 'I should go now');
console.log(response);
})();
This produces the following result in a console:
{ utterance: 'I should go now',
locale: 'en',
languageGuessed: false,
localeIso2: 'en',
language: 'English',
domain: 'default',
classifications:
[ { label: 'greetings.bye', value: 0.698219120207268 },
{ label: 'None', value: 0.30178087979273216 },
{ label: 'greetings.hello', value: 0 } ],
intent: 'greetings.bye',
score: 0.698219120207268,
entities:
[ { start: 12,
end: 14,
len: 3,
accuracy: 0.95,
sourceText: 'now',
utteranceText: 'now',
entity: 'datetime',
resolution: [Object] } ],
sentiment:
{ score: 1,
comparative: 0.25,
vote: 'positive',
numWords: 4,
numHits: 2,
type: 'senticon',
language: 'en' },
actions: [],
srcAnswer: 'Till next time',
answer: 'Till next time' }
By default, the neural network tries to avoid false positives. To achieve that, one of the internal processes is that words never seen by the network are represented as a feature that gives some weight to the None
intent. So, if you try the previous example with "I have to go" it will return the None
intent because 2 of the 4 words have never been seen while training.
If you don't want to avoid those false positives, and you feel more comfortable with classifications into the intents that you declare, then you can disable this behavior by setting the useNoneFeature
to false:
const manager = new NlpManager({ languages: ['en'], nlu: { useNoneFeature: false } });
You can also add a log progress, so you can trace what is happening during the training. You can log the progress to the console:
const nlpManager = new NlpManager({ languages: ['en'], nlu: { log: true } });
Or you can provide your own log function:
const logfn = (status, time) => console.log(status, time);
const nlpManager = new NlpManager({ languages: ['en'], nlu: { log: logfn } });
You can read the guide for how to contribute at Contributing.
Made with contributors-img.
You can read the Code of Conduct at Code of Conduct.
?
This project is developed by AXA Group Operations Spain S.A.
If you need to contact us, you can do it at the email jesus.seijas@axa.com
Copyright (c) AXA Group Operations 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.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。