From a single string to structured components
The Name Parser decomposes a full name into its salutation, title, first name, middle names, last name and suffix. It works across cultures, writing systems and naming conventions.
Internally it leverages extensive dictionaries covering names across languages and scripts. Identified terms are matched against known structures, and options are weighted using cultural and statistical data to deliver ranked results.
The parser is the foundation for downstream services: name matching, gender classification, culture identification, and entity disambiguation.
Why parsing names is hard
Cultural variations
Different societies organise names differently — some place surnames before given names, others use patronymics or compound family names.
Naming system diversity
Not all cultures employ identical component structures. Some have given + family, others given + patronymic + family.
Ambiguous entries
Many names appear across multiple cultures with different meanings — Andrea is male in Italian, female in German.
Spelling variants
Alternative forms of the same name — Catherine / Katherine / Katarina.
Transcription loss
Detail disappearing through ASCII conversion: Müller → Mueller or Muller.
Unconventional names
Unique or non-standard personal names — pseudonyms, single-token names, artistic names.
Misspellings
Errors requiring high-confidence matching — the parser still recognises the underlying name.
Joint accounts
Multiple people in a single field — "Mr. and Mrs. John Smith", "Anna & Peter Meier".
Person vs. organisation
Distinguish natural persons from companies that look like names — "John Deere Inc.".
What the components unlock
Form pre-fill
Auto-populate first/last name fields from a single full-name input.
Name matching
Feed structured names into the Name Matcher for high-quality record reconciliation.
Gender determination
Identify the given name to drive gender prediction across cultural contexts.
Personalisation
Address customers correctly: salutation, title, given name — in newsletters, transactional e-mail, support.
Person vs. company
Route legal-person records to a different processing pipeline than natural persons.
Joint detection
Recognise that one name field actually contains two people and split into two records.
A simple HTTP call
Request
POST https://api.nameapi.org/rest/v5.3/parser/personnameparser?apiKey=YOUR_API_KEY
{
"inputPerson": {
"type": "NaturalInputPerson",
"personName": {
"nameFields": [
{ "string": "Dr. John F. Kennedy",
"fieldType": "FULLNAME" }
]
}
}
}Response (excerpt)
{
"matches": [
{
"parsedPerson": {
"personType": "NATURAL",
"personName": {
"terms": [
{ "string": "Dr.", "termType": "TITLE" },
{ "string": "John", "termType": "GIVENNAME" },
{ "string": "F.", "termType": "GIVENNAMEINITIAL" },
{ "string": "Kennedy", "termType": "SURNAME" }
]
},
"gender": { "gender": "MALE", "confidence": 0.97 }
},
"parserDisputes": [],
"likeliness": "LIKELY",
"confidence": 0.92
}
]
}Parse names in your stack
We can walk you through the API, run a benchmark on your data, and help with onboarding.