Name Matcher
The person matcher compares the data sets of two people and computes scores of similarity and differences on multiple levels, including names.
| Examples by culture: Examples: |
Developer: see the technical specification of the REST service.
The matcher compares the data items that are handed over, including the name, gender, age.
There are various reasons why two name pairs are not precise matches, yet stand for the same person:
- one name is incomplete (only one of two surnames written)
- a given name is abbreviated or missing
- a nickname is used instead of the official given name
- a title or qualifier is added
- another alphabet is used such as Cyrillic, or non-English extra characters are missing
- one contains a spelling mistake (various reasons, see Name Validator)
The name matcher computes a myriad of numbers to come to a result of telling whether two name pairs likely or possibly belong to the same person:
- name variants
- nick names
- variant transcriptions
- chance of mistyping
- phonetic distance
- simple string difference (Levenshtein, Damerau algorithm)
Applications
- Tracing duplicates
- Prevent duplicates
- Filter and sort a search list by relevance