Skip to content

bassmanitram/recastai-recognizer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

recastai-recognizer

An MS Bot Framework IIntentRecognizer implementation for Recast.ai

Prereqs

Installation

npm install recastai-recognizer

usage

The recognizer is used in basically the same way as the LUIS recognizer plugin packaged with the MS Bot Framework SDK. It uses the Recast.ai node.js client module to communicate with your Recast.ai bot:

/*
 * The Recast.ai-specific bit
 */
var rairec = require('recastai-recognizer')
var recognizer = new rairec.RecastRecognizer(YOUR_TOKEN, YOUR_LANGUAGE)

/*
 * Standard MS Bot Framework stuff
 */
var builder = require('botbuilder')
var dialog = new builder.IntentDialog({ recognizers: [recognizer] });
var bot = new builder.UniversalBot(MS_BOT_FMWK_CONNECTOR)
bot.dialog('/', intents)

/*
 * Matching against one Recast.ai intent
 */
dialog.matches('YOUR_INTENT', function(session, args) {
    ...
}

/*
 * Matching against one of a number of Recast.ai intents
 */
dialog.matchesAny(['YOUR_INTENT_1', 'YOUR_INTENT_2' ...], function(session, args) {
    ...
}
...
  • YOUR_TOKEN should be the client token string from your Recast.aia bot
  • YOUR_LANGUAGE should be the language in which the utterenaces are made (e.g. "en" or "fr")

This allows the Recast.ai API to be used with the botbuilder intent matching facilities, making dialog and bot development with Recast.ai much more intuitive and closer to the recommended style of the botbuilder framework.

Notes

  • Entity information from Recast.ai far exceeds that defined for the IEntity interface. Consequently, for each returned entity, the entity (string) property of the IEntity structure is set to the raw property of the Recast entity object, while an additional property, details, is set to the full Recast entity object.
  • The full Recast response structure can be found in the recastResult property of the returned IIntentRecognizerResult object.

Contributing

In lieu of a formal styleguide, take care to maintain the existing coding style. That said, my own style is rather old-school, so any improvements that retain compatibility with the minimum platform requirements of the Recast.ai framework are more than welcome.

Add unit tests for any new or changed functionality. Lint and test your code - some of which I've been particularly careful to observe in these early releases!

Release History

  • 0.0.4 Breaking change! "description" in IEntity object changed to "details"
  • 0.0.3 Better Recast Entity support, plus tests!
  • 0.0.2 Documentation fixes!
  • 0.0.1 Initial release

About

An MS Bot Framfork IIntentRecognizer for Recast.ai

Resources

Stars

Watchers

Forks

Packages

No packages published