Trapets

Trapets helps in meeting compliance requirements, reduce risk and support for customer onboarding, ongoing due diligence (ODD) and enhanced due diligence (EDD) processes.

This data provider features screening for person and organisations against the PEP, sanction and adverse media lists, thus providing valuable information in the onboarding journey.

Trapets is suitable for both one-time and continuous screening of natural persons and organisations.

Properties

Data attributes

The Data Verification API provides the following attributes from Trapets:

• Screening

API requests

To make an API call to the Organisation Screening endpoint and retrieve data from Trapets, build your API request URL like:

https://api.signicat.com/info/lookup/countries/{countries}/organizations/screening?name={name}&listType={listType}&source={source}

Note the following parameters in the request:

• country: One or more ISO 3166-1 alpha-2 country codes, separated by commas. For example: se,no.
• source: The data provider or screening source (trapets).
• name: The name of the person or organisation to screen.
• listType: One or more list categories to check against. Supported values: pep, sanction, adverseMedia. Use commas to separate multiple values.

For example, you can screen Katrine Thorstensen for PEP, sanctions and adverse media in Sweden by sending this request:

https://api.signicat.com/info/lookup/organizations/screening?countries=se%2Cno&name=Katrine%20Thorstensen&source=trapets&listType=pep%2Csanction%2CadverseMedia&rawJson=true

Matching logic

Trapets uses two types of name-matching logic when screening:

• Exact match (also called Regular)
• Phonetic match (also called Fuzzy)

Exact match

In an exact match, all name parts in the searched name must be equal to name parts in a single list entry. The order of name parts does not matter.

Definition

"Equal" means character-by-character identical.

If a name part appears more than once in the searched name, then it must also appear the same number of times in the list entry.

Examples

• John John Andersson matches John John Andersson
• John Andersson matches John John Andersson
• John John Andersson (search) does not match John Andersson (list entry)

Phonetic match

Phonetic matching uses fuzzy logic to find names that sound (phonetically) similar, based on the Metaphone3 algorithm.

The system looks for at least one name part in the searched name that phonetically matches a name part in a list entry. Then, it calculates similarity using the Levenshtein distance algorithm, which measures how many edits are needed to change one word into another.

Each name part's distance is weighted by its length. This means that differences in longer name parts affect the total score more than shorter ones.

Recommended threshold

A similarity score (hit rate) between 0.75 and 0.9 is typically recommended for fuzzy matching.

Score

The score represents the confidence level of a match, based on Trapets' internal Hit Rating system.

How it works

• The Hit Rating is an integer from 1 to 5.
• The score is calculated by multiplying the Hit Rating by 20 to produce a percentage. For example, 5 x 20 = 100%.

Examples

• A Hit Rating of 5 (score 100) means a perfect match, typically when a unique identifier, such as social security number (SSN) or company registration number, is used.
• If no SSN is provided and the match is based only on name and country, the Hit Rating may be lower (1 or 2) and the total score affected (20% or 40%).

Hit rating

Indicates how relevant this list item is to the query. The higher number, the more accurate the match, with 5 being the most accurate.

Hit rating breakdown

Natural persons

* Date of birth can be provided as full date, YYYY-MM or just YYYY.

Organisations

Screening Logic

The KYC screening engine processes input parameters in order of reliability. The flow differs slightly between PEP and sanctions screening.

PEP screening

When screening for politically exposed persons (PEPs), the engine evaluates the input fields in the following order:

1. Social Security Number (SSN)
   • If an SSN is provided and matched, it is considered a definitive match. All other input (name, country, date of birth) is ignored.
   • If no match is found, the engine excludes list entries with non-matching SSNs.
2. Date of birth
   • If provided, the engine keeps only entries with matching dates of birth or no dates of birth at all.
3. Country
   • If provided, the engine filters out entries from non-matching countries.
4. Name
   • Finally, the engine performs name matching based on either exact or phonetic logic.

Sanction screening

Sanctions screening follows the same process, with one key difference:

• SSNs are not used, as sanctions lists do not include SSNs.

Test data