Date extracted by CRFEntityExtractor instead of DucklingHTTPExtractor

Hello, I was able to setup duckling and made sure its working.

However, when the Rasa extracted dates, CRFEntityExtractor is used instead and the result is wrong:

  1. confidence: 0.8715912209

  2. end: 23

  3. entity: “date_jobdate”

  4. extractor: “CRFEntityExtractor”

  5. processors: Array(1)

  6. 0: “EntitySynonymMapper”

  7. length: 1

  8. start: 15

  9. value: “2019-06-14T00:00:00.000-04:00”

The text is “I want to move tomorrow”

It pickups tomorrow but convert it to the wrong date since tomorrow is 2019-06-27

I have Duckling in the pipeline:


  • name: “SpacyNLP”
  • name: “SpacyTokenizer”
  • name: “SpacyFeaturizer”
  • name: “SklearnIntentClassifier”
  • name: “CRFEntityExtractor”
  • name: “EntitySynonymMapper”
  • name: “RegexFeaturizer”
  • name: “DucklingHTTPExtractor” url: “” locale: “en-US” timezone:“US/Eastern”

But it doesn’t use duckling to extract data If I use duckling directly I’m getting correct parsing: [ { “body”: “tomorrow”, “start”: 15, “value”: { “values”: [ { “value”: “2019-06-27T00:00:00.000-07:00”, “grain”: “day”, “type”: “value” } ], “value”: “2019-06-27T00:00:00.000-07:00”, “grain”: “day”, “type”: “value” }, “end”: 23, “dim”: “time”, “latent”: false } ]

Please help

I have the exact same problem if someone could help :slight_smile: ?

maybe put IntentClassifier in the last row of your pipeline