anonymous userKumarDave-6852 With the above information you can extract entities using the named entity recognition API. In most cases the phone numbers, names, email ids are identified without any additional processing but with numbers like customer id or invoice id you might want to perform some validation of the data as it is returned as a number in the response. In the above case I used your example sentence as is with the API and the following is the response of the entities.
With a bit of standard formatting in the input to the API you can also readily categorize the numbers as either customer ids or invoice ids based on the offset returned with the numbers. I hope this helps!!
"documents": [{
"id": "2",
"entities": [{
"text": "Rajev dave",
"category": "Person",
"offset": 11,
"length": 10,
"confidenceScore": 0.61
}, {
"text": "customer",
"category": "PersonType",
"offset": 50,
"length": 8,
"confidenceScore": 0.51
}, {
"text": "10400068",
"category": "Quantity",
"subcategory": "Number",
"offset": 69,
"length": 8,
"confidenceScore": 0.8
}, {
"text": "10400068",
"category": "Phone Number",
"offset": 69,
"length": 8,
"confidenceScore": 0.8
}, {
"text": "email",
"category": "Skill",
"offset": 82,
"length": 5,
"confidenceScore": 0.8
}, {
"text": "raajeevxxx@gmail.com",
"category": "Email",
"offset": 99,
"length": 20,
"confidenceScore": 0.8
}, {
"text": "billing information",
"category": "Skill",
"offset": 143,
"length": 19,
"confidenceScore": 0.8
}, {
"text": "next month",
"category": "DateTime",
"subcategory": "DateRange",
"offset": 227,
"length": 10,
"confidenceScore": 0.8
}, {
"text": "30087",
"category": "Quantity",
"subcategory": "Number",
"offset": 260,
"length": 5,
"modelVersion": "2 "confidenceScore": 0.8
}],
"warnings": []
}],
"errors": [],