Json file containing labels not accepted custom NER

Marten 1 Reputation point
2022-04-15T13:09:51.787+00:00

I am trying to make a custom NER project using Azure. When I want to use my own file for the tags of the data, it responds "InvalidArgument: Invalid project json object. Could not find member 'extractors' on object of type 'JsonProjectAssets'" when I click create. I've tried different formats I found online and even tried some example files they give in the Quickstart documentation. But I keep receiving the same error. Are they busy with changing the format of the tagged file or am I doing something wrong? I hope someone knows how to fix it!

Azure AI Language
Azure AI Language
An Azure service that provides natural language capabilities including sentiment analysis, entity extraction, and automated question answering.
359 questions
Azure AI services
Azure AI services
A group of Azure services, SDKs, and APIs designed to make apps more intelligent, engaging, and discoverable.
2,415 questions
{count} votes

1 answer

Sort by: Most helpful
  1. Pat 6 Reputation points
    2022-04-21T14:10:53.857+00:00

    I ran into this same issue while following the custom text classification quick start guide. I got around the issue by first creating my project in Language Studio with no mapping file (I selected "No, I need to tag my files as part of this project."). Then I used the REST API to add the mapping file. Within the body of the request, you can import the JSON file. I found that the keys need to match the example in the assets object or else I received an error.

    "assets": {  
        "classes": [  
          {  
            "category": "string"  
          }  
        ],  
        "entities": [  
          {  
            "category": "string"  
          }  
        ],  
        "documents": [  
          {  
            "location": "string",  
            "language": "string",  
            "class": {  
              "category": "string"  
            },  
            "classes": [  
              {  
                "category": "string"  
              }  
            ],  
            "entities": [  
              {  
                "regionOffset": 0,  
                "regionLength": 0,  
                "labels": [  
                  {  
                    "category": "string",  
                    "offset": 0,  
                    "length": 0  
                  }  
                ]  
              }  
            ],  
            "dataset": "string"  
          }  
        ]  
      }  
    
    1 person found this answer helpful.
    0 comments No comments