Duckling extraction does not gives timestamp instead the value is extracted as text itself

Rasa Open Source
1

I’m using rasa for building my reminder chatbot where when a user gives a time , the time is extracted as the text what user gives instead of extracting timestamp…What is the issue and how it can be solved

image
image1920×1080 213 KB

refer the image for appointment_time slot and look the time format , it is text instead of timestamp

2

Yes, this is the case in rasa==1.10. It’s actually more reliable though, because time can be a single value or an interval, and if it is an interval then in rasa<=1.10 it was represented in a pretty strange way in the value. The timestamp (or timestamps, if you get an interval) are still available in the additional_info portion of the entity. For an example of backward-compatible parsing of the duckling time entity, see parsing.py in the financial-demo bot custom actions

3

Thanks much for the solution

4

@mloubser Why is it that when I run DucklingHTTPExtractor on my local machine with rasa==1.10.1, I get different results than when I run it on my rasax server running on a GCP VM (I’m running duckling on the vm with docker).

Below is an example: When I run the message “$2000-3000” using rasa shell nlu, I get the following output:

"entities": [
    {
      "start": 0,
      "end": 5,
      "text": "$2000",
      "value": 2000,
      "confidence": 1.0,
      "additional_info": {
        "value": 2000,
        "type": "value",
        "unit": "$"
      },
      "entity": "amount-of-money",
      "extractor": "DucklingHTTPExtractor"
    },
    {
      "start": 1,
      "end": 10,
      "text": "2000-3000",
      "value": {
        "to": 3000,
        "from": 2000
      },
      "confidence": 1.0,
      "additional_info": {
        "to": {
          "value": 3000,
          "unit": "unknown"
        },
        "from": {
          "value": 2000,
          "unit": "unknown"
        },
        "type": "interval"
      },
      "entity": "amount-of-money",
      "extractor": "DucklingHTTPExtractor"
    }
  ]

But when I run the same in the Interactive Learning section of the RasaX UI on my GCP VM, I don’t get any of these entities. In fact, it doesn’t even catch the same text as an entity.

Do you know why this is the case or how I can fix this?

5

Rasa version 1.10 has these updates where it gets extracted as text and we can also call the value using additional info

6

But in this case it’s not even extracting the amount-of-money entity from the same text. Do you know why that would be?

7

If you want the value of the amount of money u can directly extract from “from” or “to” inside the value dict.

So the code works like tis entity[‘from’][‘value’] to get the “from” value entity[‘to’][‘value’] to get the “to” value

8

I understand, but what I’m trying to say is that for some reason when I run rasa shell nlu on my local machine with $2000-3000 as text, DucklingHTTPExtractor extracts an amount-of-money entity, but on my GCP instance, it doesn’t even recognize $2000-3000 as an entity in the first place, so I can’t get any value from it at all.

9

I have a feeling this is happening because the Rasa/duckling fork doesn’t have the same rules as the main facebook version of duckling, because the facebook version is what I downloaded on my local machine and I believe the duckling docker image uses the Rasa fork.

10

Please check the version of rasa on your GCP