Relations

Overview

A relationship extraction task requires the detection and classification of semantic relationship mentions within a set of artifacts, typically from text or XML documents.

Get Relations Using Catalyst UI

• Goto AmplifyReach Dashboard -> Bot Studio -> Left Navigation Sidebar
• Select Core Services > Text Analysis.
• Select URL or Text tab to enter the link or text for Text Analysis. 

• Press Analyze button.
• Text Analysis Results will be shown.

• Check Relation section to find out relations identified by Text Analysis.
• For Developers, you can view JSON by clicking </> View JSON on right corner.
• Following relations are detected
  1. Is Positive
  2. Is Negative
  3. Is Question
  4. Subject
  5. Object
  6. Entity
  7. Answer Type

Extracting Relations Using AmplifyReach Catalyst API

Tip: Register your account on AmplifyReach and contact us at kick-start@amplifyreach.com to get Developer ID and Access Token.

URL & Objects

The following URLs are available.

PUT /textanalysis/v1/relations

PUT https://launch.amplifyreach.com/textanalysis/v1/relations

Headers:

PUT body

cURL-PUT Request for URL Request

curl \
-H "Content-Type: application/json" \
-H "cid:<YOUR_DEVELOPER_ID>" \
-H "authtoken:<YOUR_DEVELOPER_ACCESS_TOKEN>" \
-X PUT \
-d '{"url":"<SAMPLE_URL_FOR_ANALYSIS>"}'\
https://launch.amplifyreach.com/textanalysis/v1/relations

cURL-PUT Request for Text Request

curl \
-H “Content-Type: application/json” \
-H “cid:” \
-H “authtoken:” \
-X PUT \
-d ‘{“text”:””}’ \
https://launch.amplifyreach.com/textanalysis/v1/relations

Put Response

The taxonomy put API response JSON structure.

Example: PUT Response

{
“status”:”success”,
“uid”:”4f057e01d88ff7aa82a8610dc651e8fe”,
“license”:”By accessing AmplifyReach or using information generated by AmplifyReachAPI,
you are agreeing to be bound by the AmplifyReach Terms of Use.”
}

GET /textanalysis/v1/relations/{UID}

GET https://launch.amplifyreach.com/textanalysis/v1/relations/{UID}

Headers:

cURL-GET Request for URL Request

Response Time:
Free Trial Account
– For URL as a payload, you can retrieve response after 10 seconds
– For Text as a payload, you can retrieve response after 1 seconds
Paid account:
– Higher QPS as per applicable plan

curl \
-H “Content-Type: application/json” \
-H “cid:” \
-H “authtoken:” \
-X GET \
https://apps.amplifyreach.com/textanalysis/v1/relations/{UID}

Relations Object

The relations response JSON structure:

Example: Relations Object

Example-1
{
“status”: “success”,
“relations”: [{
“sentence”: “I want to book ticket from Mumbai to Pune.”,
“knowledge”: [{
“isNegative”: false,
“isSentence”: true,
“isQuestion”: false,
“tense”: “SPR”,
“subject”: “I”,
“action”: “book”,
“object”: “ticket”,
“supportedClause”: “from Mumbai”,
“entity”: [“Mumbai”, “Pune”]
}]
}]
}
Example-2
{
“status”: “success”,
“relations”: [{
“sentence”: “His fingers poised over the keys several times as if he were going to type more.”,
“knowledge”: [{
“isNegative”: false,
“isSentence”: true,
“isQuestion”: false,
“subject”: “His fingers”,
“action”: “poised”,
“object”: “several times”,
“supportedClause”: “over the keys”,
“entity”: []
},
{
“isNegative”: false,
“isSentence”: true,
“isQuestion”: false,
“subject”: “He”,
“action”: “type”,
“object”: “more”,
“entity”: []
}]
}]
}
Example-3
{
“status”: “success”,
“relations”: [{
“sentence”: “How many days are required to complete this task”,
“knowledge”: [{
“isNegative”: false,
“isSentence”: false,
“isQuestion”: true,
“action”: “complete”,
“object”: “this task”,
“answerType”: {
“answerType”: “number_count”,
“answerReferance”: “days”
},
“entity”: []
}]
}]
}
Example-4
{
“status”: “success”,
“relations”: [{
“sentence”: “I am buying Ice Cream, Do you want?”,
“knowledge”: [{
“isNegative”: false,
“isSentence”: true,
“isQuestion”: false,
“subject”: “I”,
“action”: “buying”,
“object”: “Ice Cream”,
“entity”: []
},
{
“isNegative”: false,
“isSentence”: false,
“isQuestion”: true,
“subject”: “you”,
“action”: “want”,
“answerType”: {
“answerType”: “yes_no”,
“answerReferance”: “days”
},
“entity”: []
}]
}]
}