Name Parser

Name parsing is the process of splitting a person's name string into its components, such as first name, middle name, last name, suffix, salutation, and title.

Developer: see the technical specification of the REST service.

The Solution

A process that begins by identifying the individual terms, matching them against defined name structures, selecting the best options based on culture and statistics.

Large dictionaries are used to identify terms, covering the names and name parts from this world, in all languages and writing systems. Cultural and statistical attributes and data are used to weight the options and score and sort the results.

Challenges

The world exhibits a variety of names and naming systems, which have become interconnected due to globalization. While colonization and globalization have contributed to a more standardized approach to naming, there are still numerous cultural peculiarities that make this process challenging.

1. Cultural variations

Names can have different structures, order, and conventions across various cultures. First names can come after the last name. Parsing names from different cultural backgrounds requires understanding and accommodating these variations.

2. Name components

Different cultures use different naming systems, not all countries have first names, middles and last names.

3. Ambiguous names and name parts

Names that exist in multiple cultures, having different roots, and therefore different attributes (culture, gender, age). Names can be ambiguous, with multiple possible interpretations.

4. Spelling variations

Names can have various spelling variations or alternate forms.

5. Lost detail

Detail is lost through transcription (or ASCIIfication), rendering distinctive names the same.

6. Unconventional names

Some individuals have unconventional or unique names.

7. Misspellings

Sometimes a misspelling can be linked to the actual name with high confidence, other times contradicting options are viable.

Use Cases

Parsing a person’s name is the first step for additional processing, such as name matching, genderizing, culture classification, and is also a core component of our SearchCluster product.

The Name Parser has numerous practical applications, some of which include:

• Name splitting: identifying and extracting first names, last names, prefixes (such as Mr. or Dr.), and suffixes (such as Jr. or III) from a full name string.
• Identifying the gender of a person: this can be a complex task due to the fact that the same name may have one gender in a certain culture and the opposite gender in another.
• Classifying natural and legal persons (businesses): differentiating between natural and legal persons by analyzing the components of a name string.
• Identifying multiple individual people within a name string: the name may contain 2 natural people in the same field and the syntax used can significantly differ depending on the country. For instance, in the context of bank account names, it is possible to encounter situations where two people are linked together as joint account holders.