Connectors

BabelStreet Connector

The BabelStreet connector lets you use the BabelStreet Analytics API in your PingOne DaVinci flow.

The connector offers a suite of tools for name and address similarity, name transliteration, and language identification.

Setup

Resources

You can find more information and setup help in the following:

Requirements

To use the connector, you’ll need a BabelStreet API Key.

Configuring the BabelStreet connector

Add the connector in DaVinci as shown in Adding a connector, then configure it as follows.

Connector configuration

API Key

The API Key for the BabelStreet Analytics API.

Using the connector in a flow

Comparing name similarity

A screen capture of the complete name similarity flow.

This flow prompts a user to enter two names into a form, optionally specifying the language for each. For greater accuracy, the node can also be configured with an expected entity type and gender. The node then compares the names and outputs a similarity score from 0 to 1.

Test the flow by clicking Save, Deploy, and Try Flow.

Comparing address similarity

A screen capture of the complete address similarity flow.

This flow prompts a user to enter the details of two addresses into a form. The accuracy of the comparison improves as more fields are completed. The node then compares the two addresses and outputs a similarity score from 0 to 1.

Test the flow by clicking Save, Deploy, and Try Flow.

Converting a name from one writing system to another

A screen capture of the complete name transliteration flow.

This flow prompts a user to enter a name and a target language. For a more accurate transliteration, the user can optionally provide the name’s source language and configure the node to expect a specific entity type. The node then outputs the transliterated name in the target language, along with a confidence score.

Test the flow by clicking Save, Deploy, and Try Flow.

Identifying the language of a text sample

A screen capture of the complete language identification flow.

This flow prompts a user to enter a text sample. The node then outputs the language detected, along with a confidence score.

Test the flow by clicking Save, Deploy, and Try Flow.

Capabilities

Name Similarity

Name Similarity compares two entity names (Person, Location, or Organization) and returns a match score from 0 to 1.

Show details

  • Properties

  • Input Schema

  • Output Schema

Name 1 textField required

The first name to be compared

Name 1 - Language textField

The language of the first name

Name 1 - Entity Type dropDown

The type of the first name

  • Person (Default)

  • Location

  • Organization

Name 1 - Gender dropDown

The gender associated with the first name

  • Male

  • Female

  • Nonbinary

Name 2 textField required

The second name to be compared

Name 2 - Language textField

The language of the second name

Name 2 - Entity Type dropDown

The type of the second name

  • Person (Default)

  • Location

  • Organization

Name 2 - Gender dropDown

The gender associated with the second name

  • Male

  • Female

  • Nonbinary

  • default object

    • properties object

      • textName1 string required

        The first name to be compared. This can be the name of a person, location, or organization (e.g., "Barbara Jensen", "New York City", or "Babel Street").

      • languageName1 string

        The first name’s language specified by the three-letter ISO 639-3 code.

      • entityTypeName1 string

        The type of the first name (use "PERSON", "LOCATION", or "ORGANIZATION").

      • genderName1 string

        The gender associated with the first name (use "MALE", "FEMALE", or "NONBINARY").

      • textName2 string required

        The second name to be compared. This can be the name of a person, location, or organization (e.g., "Barbara Jensen", "New York City", or "Babel Street").

      • languageName2 string

        The second name’s language specified by the three-letter ISO 639-3 code.

      • entityTypeName2 string

        The type of the second name (use "PERSON", "LOCATION", or "ORGANIZATION").

      • genderName2 string

        The gender associated with the second name (use "MALE", "FEMALE", or "NONBINARY").

Input Example

{
  "properties": {
    "textName1": "Barbara Jensen",
    "languageName1": "eng",
    "entityTypeName1": "PERSON",
    "genderName1": "FEMALE",
    "textName2": "Hanson Dolores",
    "languageName2": "eng",
    "entityTypeName2": "PERSON",
    "genderName2": "MALE"
  }
}

  • output object

    • rawResponse object

      The full, unmodified JSON object returned directly from the external API call.

    • statusCode number

      The HTTP status code returned by the API request.

    • score number

      A numerical score from 0.0 (no similarity) to 1.0 (a perfect match) indicating how similar the two names are.

Output Example

{
  "rawResponse": {
    "score": 0.43314963
  },
  "statusCode": 200,
  "score": 0.43314963
}

Address Similarity

Address Similarity compares two addresses and returns a match score from 0 to 1.

Show details

  • Properties

  • Input Schema

  • Output Schema

Address 1 - Unit textField

The unit, apartment, or suite number for the first address

Address 1 - House Number textField

The house or building number on the street for the first address

Address 1 - Street textField

The street name for the first address

Address 1 - City textField

The city or town for the first address

Address 1 - State textField

The state, province, or primary administrative region for the first address

Address 1 - Postal Code textField

The postal code or ZIP code for the first address

Address 1 - Country textField

The country for the first address

Address 2 - Unit textField

The unit, apartment, or suite number for the second address

Address 2 - House Number textField

The house or building number on the street for the second address

Address 2 - Street textField

The street name for the second address

Address 2 - City textField

The city or town for the second address

Address 2 - State textField

The state, province, or primary administrative region for the second address

Address 2 - Postal Code textField

The postal code or ZIP code for the second address

Address 2 - Country textField

The country for the second address

  • default object

    • address1Unit string

      The unit, apartment, or suite number for the first address (e.g., "4B" or "Suite 1200").

    • address1HouseNo string

      The house or building number on the street for the first address (e.g., "123").

    • address1Road string

      The street name for the first address (e.g., "Main St").

    • address1City string

      The city or town for the first address (e.g., "Victoria").

    • address1State string

      The state, province, or primary administrative region for the first address (e.g., "BC" or "California").

    • address1PostCode string

      The postal code or ZIP code for the first address (e.g., "V8W 1A1" or "90210").

    • address1Country string

      The country for the first address. Can be a full name or a code (e.g., "Canada" or "CAN").

    • address2Unit string

      The unit, apartment, or suite number for the second address.

    • address2HouseNo string

      The house or building number on the street for the second address.

    • address2Road string

      The street name for the second address.

    • address2City string

      The city or town for the second address.

    • address2State string

      The state, province, or primary administrative region for the first address.

    • address2PostCode string

      The postal code or ZIP code for the second address.

    • address2Country string

      The country for the second address.

Input Example

{
  "properties": {
    "address1Unit": "Suite 300",
    "address1HouseNo": "721",
    "address1Road": "Government Street",
    "address1City": "Victoria",
    "address1State": "BC",
    "address1PostCode": "V8W 1W5",
    "address1Country": "Canada",
    "address2Unit": "300",
    "address2HouseNo": "721",
    "address2Road": "Government St",
    "address2City": "Victoria",
    "address2State": "British Columbia",
    "address2PostCode": "V8W 1W5",
    "address2Country": "CA"
  }
}

  • output object

    • rawResponse object

      The full, unmodified JSON object returned directly from the external API call.

    • statusCode number

      The HTTP status code returned by the API request.

    • score number

      A numerical score from 0.0 (no similarity) to 1.0 (a perfect match) indicating how similar the two addresses are.

Output Example

{
  "rawResponse": {
    "score": 0.86834957
  },
  "statusCode": 200,
  "score": 0.86834957
}

Name Transliteration

Name Transliteration converts a name from one writing system to another, preserving its approximate pronunciation.

Show details

  • Properties

  • Input Schema

  • Output Schema

Name textField required

The name to be transliterated

Target Language textField required

The language to transliterate the name into

Entity Type dropDown required

The type of entity: Person, Location, or Organization

  • Person (Default)

  • Location

  • Organization

Source Language of Origin textField

The name’s original language

  • default object

    • properties object

      • name string required

        The name of the entity to transliterate (e.g., "Barbara Jensen" or "北京").

      • targetLanguage string required

        The three-letter ISO 639-3 code for the target language (e.g., "eng" or "kor").

      • entityType string required

        Specifies the entity’s type to improve accuracy (use "PERSON", "LOCATION", or "ORGANIZATION").

      • sourceLanguageOfOrigin string

        The name’s native language, as a three-letter ISO 639-3 code (e.g., "ara" (Arabic) for the name "Mahmoud Abbas").

Input Example

{
  "properties": {
    "name": "Борис Петров",
    "targetLanguage": "eng",
    "entityType": "PERSON",
    "sourceLanguageOfOrigin": "rus"
  }
}

  • output object

    • rawResponse object

      The full, unmodified JSON object returned directly from the external API call.

    • statusCode number

      The HTTP status code returned by the API request.

    • translation string

      The transliterated name in the target language.

    • confidence number

      A numerical score from 0.0 (lowest) to 1.0 (highest) indicating the API’s certainty in the accuracy of the transliteration.

Output Example

{
  "rawResponse": {
    "extendedInformation": {
      "processedByBabel": true
    },
    "sourceScript": "Cyrl",
    "sourceLanguageOfOrigin": "rus",
    "sourceLanguageOfUse": "rus",
    "targetLanguage": "eng",
    "targetScript": "Latn",
    "targetScheme": "IC",
    "translation": "Boris Petrov",
    "confidence": 1
  },
  "statusCode": 200,
  "translation": "Boris Petrov",
  "confidence": 1
}

Language Identification

Language Identification identifies the language of a given text sample, returning the language code and a confidence score.

Show details

  • Properties

  • Input Schema

  • Output Schema

Content textArea required

The text sample to be analyzed for language identification

  • default object

    • properties object

      • content string required

        The text sample to be analyzed for language identification.

Input Example

{
  "properties": {
    "content": "明日は明日の風が吹く"
  }
}

  • output object

    • rawResponse object

      The full, unmodified JSON object returned directly from the external API call.

    • statusCode number

      The HTTP status code returned by the API request.

    • language string

      The three-letter ISO 639-3 code for the detected language (e.g., "eng" for English).

    • confidence number

      A numerical score from 0.0 (lowest) to 1.0 (highest) indicating the API’s certainty in the accuracy of the language identification.

Output Example

{
  "rawResponse": {
    "languageDetections": [
      {
        "language": "jpn",
        "confidence": 1
      }
    ]
  },
  "statusCode": 200,
  "language": "jpn",
  "confidence": 1
}