Fuzzy Data Matching with SQL

Chapter 3. Names, Names, Names

What can I say? Names are hard. Is it James or Jim? Spell-checking is impossible: people name their kids anything. Add in cross-cultural differences, and it becomes very hard to do much with names; but we must try! The rest of this chapter is going to assume you’re dealing with some system where your customer records are stored with fields similar to these if dealing with a human.

What’s in a Name?

You’ve had to fill out forms with your name since kindergarten. You know the drill on how they are supposed to work. The first three are the most common:

Last name

Or family name or surname. Maybe you have fields for matronymics and patronymics, too.

First name

Or given name. May be optional.

Middle name

Or middle initial or middle names. Optional.

Nickname(s)

Optional.

Suffix

Optional.

Titles and honorifics

Optional, and you’ll learn why we’ll ignore them.

Full name

Often synthesized from the others, but woe to you if your incoming data only has this; we will talk about it at the end!

Or if dealing with businesses, simply this:

Company name

While we commonly tear apart human names into their constituent parts, rarely are entity names held in more than one field. No matter the structure of the entity—corporation, partnership, trust, whatever—we jam it into one field. Except sometimes there is another field that looks like company name.

DBA

“Doing Business As,” often used by an individual who hasn’t established a more formal entity ...