Text Moderation / Guides

Text Moderation - Quick start guide

In this guide

The Text Moderation API can be used to moderate any kind of text content. From usernames to long form texts. Choose the guide corresponding to the type of content you need to moderate:

GUIDE

ML Text moderation

Use Deep Learning to moderate chats, messages, articles and other long-form texts.

GUIDE

Rule-based Text moderation

Use advanced rule-based pattern matching to moderate chats, messages, articles and other long-form texts.

GUIDE

Username Moderation

Moderate usernames. Make sure your users choose usernames that are not derogatory, sexual or offensive.

Rule-based Text Moderation with Sightengine

Simply send a POST request containing the UTF-8 formatted text along with the ISO 639-1 language code (such as en for english) and the comma-separated list of countries for phone number detection (such as us,gb,fr for the United States, United Kingdom and France). Here is an example:


curl -X POST 'https://api.sightengine.com/1.0/text/check.json' \
  -F 'text=Contact rick(at)gmail(dot)com to have s_*_x' \
  -F 'lang=en' \
  -F 'opt_countries=us,gb,fr' \
  -F 'mode=rules' \
  -F 'api_user={api_user}' \
  -F 'api_secret={api_secret}'


# this example uses requests
import requests
import json

data = {
  'text': 'Contact rick(at)gmail(dot)com to have s_*_x',
  'mode': 'rules',
  'lang': 'en',
  'opt_countries': 'us,gb,fr',
  'api_user': '{api_user}',
  'api_secret': '{api_secret}'
}
r = requests.post('https://api.sightengine.com/1.0/text/check.json', data=data)

output = json.loads(r.text)


$params = array(
  'text' => 'Contact rick(at)gmail(dot)com to have s_*_x',
  'lang' => 'en',
  'opt_countries' => 'us,gb,fr',
  'mode' => 'rules',
  'api_user' => '{api_user}',
  'api_secret' => '{api_secret}',
);

// this example uses cURL
$ch = curl_init('https://api.sightengine.com/1.0/text/check.json');
curl_setopt($ch, CURLOPT_POST, true);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $params);
$response = curl_exec($ch);
curl_close($ch);

$output = json_decode($response, true);


// this example uses axios and form-data
const axios = require('axios');
const FormData = require('form-data');

data = new FormData();
data.append('text', 'Contact rick(at)gmail(dot)com to have s_*_x');
data.append('lang', 'en');
data.append('opt_countries', 'us,gb,fr');
data.append('mode', 'rules');
data.append('api_user', '{api_user}');
data.append('api_secret', '{api_secret}');

axios({
  url: 'https://api.sightengine.com/1.0/text/check.json',
  method:'post',
  data: data,
  headers: data.getHeaders()
})
.then(function (response) {
  // on success: handle response
  console.log(response.data);
})
.catch(function (error) {
  // handle error
  if (error.response) console.log(error.response.data);
  else console.log(error.message);
});

The request response will contain a JSON body that you can then use to determine if the text was acceptable or not.

Objectionable content is categorized into a few main categories, with each category having different types (subcategories) that you can use to determine what was found:

Category	Types
profanity	sexual for sexual content discriminatory for discriminatory and derogatory content insult for insults grawlix for grawlix inappropriate for inappropriate language other_profanity for other types of profanity Learn more
personal	email for email addresses phone_number_** for phone numbers, where the last two characters are the country code for the matching country username for usernames ipv4 for IPs (version 4) ipv6 for IPs (version 6) ssn for US social security numbers Learn more
link	url for URLs Learn more

As an example, here is the JSON request that you would receive for the above request. The message contains a sexual term along with a slightly obfuscated email address, both of which are detected and returned:


{
  "status": "success",
  "request": {
    "id": "req_6cujQglQPgGApjI5odv0P",
    "timestamp": 1471947033.92,
    "operations": 1
  },
  "profanity": {
    "matches": [
      {
        "type": "sexual",
        "match": "sx",
        "intensity": "medium",
        "start": 38,
        "end": 42,
      }
    ]
  },
  "personal": {
    "matches": [
      {
        "type": "email",
        "match": "rick(at)gmail(dot)com",
        "start": 8,
        "end": 28
      }
    ]
  },
  "link": {
    "matches": []
  },
}

You can also create your own custom disallow list to filter custom text items, in addition to the words that we detect out-of-the-box. More on custom disallow lists

Moderating Usernames with Sightengine

Simply send a POST request containing the UTF-8 formatted username along with the ISO 639-1 language code (such as en for english). Here is an example:


curl -X POST 'https://api.sightengine.com/1.0/text/check.json' \
  -F 'text=urCr0tch' \
  -F 'lang=en' \
  -F 'mode=username' \
  -F 'api_user={api_user}' \
  -F 'api_secret={api_secret}'


# this example uses requests
import requests
import json

data = {
  'text': 'urCr0tch',
  'mode': 'username',
  'lang': 'en',
  'api_user': '{api_user}',
  'api_secret': '{api_secret}'
}
r = requests.post('https://api.sightengine.com/1.0/text/check.json', data=data)

output = json.loads(r.text)


$params = array(
  'text' => 'urCr0tch',
  'lang' => 'en',
  'mode' => 'username',
  'api_user' => '{api_user}',
  'api_secret' => '{api_secret}',
);

// this example uses cURL
$ch = curl_init('https://api.sightengine.com/1.0/text/check.json');
curl_setopt($ch, CURLOPT_POST, true);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $params);
$response = curl_exec($ch);
curl_close($ch);

$output = json_decode($response, true);


// this example uses axios and form-data
const axios = require('axios');
const FormData = require('form-data');

data = new FormData();
data.append('text', 'urCr0tch');
data.append('lang', 'en');
data.append('mode', 'username');
data.append('api_user', '{api_user}');
data.append('api_secret', '{api_secret}');

axios({
  url: 'https://api.sightengine.com/1.0/text/check.json',
  method:'post',
  data: data,
  headers: data.getHeaders()
})
.then(function (response) {
  // on success: handle response
  console.log(response.data);
})
.catch(function (error) {
  // handle error
  if (error.response) console.log(error.response.data);
  else console.log(error.message);
});

The request response will contain a JSON body that you can then use to determine if the username was acceptable or not.

Objectionable content is categorized into a few main categories, with each category having different types (subcategories) that you can use to determine what was found:

Category	Types
profanity	sexual for sexual content discriminatory for discriminatory and derogatory content insult for insults grawlix for grawlix inappropriate for inappropriate language other_profanity for other types of profanity Learn more
misleading	misleading for usernames that might be misleading when viewed by other users. Learn more
personal	email for email addresses phone_number_** for phone numbers, where the last two characters are the country code for the matching country ipv4 for IPs (version 4) ipv6 for IPs (version 6) ssn for US social security numbers Learn more

As an example, here is the JSON request that you would receive for the above request. The username contains an obfuscated sexual term, which is detected and returned:


{
  "status": "success",
  "request": {
    "id": "req_6cujQglQPgGApjI5odv0P",
    "timestamp": 1471947033.92,
    "operations": 1
  },
  "profanity": {
    "matches": [
      {
        "type": "sexual",
        "match": "crotch",
        "intensity": "medium",
        "start": 0,
        "end": 7
      }
    ]
  },
  "personal": {
    "matches": []
  },
  "link": {
    "matches": []
  },
  "misleading": {
    "matches": []
  }
}

Was this page helpful?

Products

MODERATION

REDACTION

REFERENCE

Text Moderation - Quick start guide

In this guide

ML Text moderation

Rule-based Text moderation

Username Moderation

Rule-based Text Moderation with Sightengine

Moderating Usernames with Sightengine