Simple way to parse a person's name into its component parts? [closed]

Answer 1

This is a fools errand. Too many exceptions to be able to do this deterministically. If you were doing this to pre-process a list for further review I would contend that less would certainly be more.

1. Strip out salutations, titles and generational suffixes (big regex, or several small ones)
2. if only one name, it is 'last'.
3. If only two names split them first,last.
4. If three tokens and middle is initial split them first, middle, last
5. Sort the rest by hand.

Any further processing is almost guaranteed to create more work as you have to go through recombining what your processing split-up.

Answer 2

The real solution here does not answer the question. The portent of the information must be observed. A name is not just a name; it is how we are known.

The problem here is not knowing exactly what parts are labled what, and what they are used for. Honorable prefixes should be granted only in personal corrospondences; Doctor is an honorific that is derived from a title. All information about a person is relavent to their identity, it is just determining what is relavent information. You need a first and last name for reasons of administration; phone number, email addresses, land descriptions and mailing addresses; all to the portent of identity, knowing who you are dealing with.

The real problem here is that the person gets lost in the administration. All of a sudden, after only entering their personal information into a form and submitting it to an arbitrary program for processing, they become afforded all sorts of honorifics and pleasentries spewed out by a prefabricated template. This is wrong; honorable Sir or Madam, if personal interest is shown toward the reason of corrospondence, then a letter should never be written from a template. Personal corrospondance requires a little knowledge about the recipient. Male or female, went to school to be a doctor or judge, what culture in which they were raised.

In other cultures, a name is made up from a variable number of characters. The person's name to us can only be interpretted as a string of numbers where the spaces are actually determined by character width instead of the space character. Honorifics in these cases are instead one or many characters prefixing and suffixing the actual name. The polite thing to do is use the string you are given, if you know the honorific then by all means use it, but this again implies some sort of personal knowledge of the recipient. Calling Sensei anything other than Sensei is wrong. Not in the sense of a logic error, but in that you have just insulted your caller, and now you should find a template that helps you apologize.

For the purposes of automated, impersonal corrospondence, a template may be devised for such things as daily articles, weekly issues or whatever, but the problem becomes important when corrospondence is instigated by the recipient to an automated service.

What happens is an error. Missing information. Unknown or missing information will always generate an Exception. The real problem is not how do you seperate a person's name into its seperate components with an expression, but what do you call them.

The solution is to create an extra field, make it optional if there is already a first and last name, and call it "What may we call you" or "How should we refer to you as". A doctor and a judge will ensure you address them properly. These are not programming issues, they are issues of communication.

Ok, bad way to put it, but in my opinion, Username, Tagname, and ID are worse. So my solution; is the missing question, "What should we call you?"

This is only a solution where you can afford to make a new question. Tact prevails. Create a new field upon your user forms, call it Alias, label for the user "What should we call you?", then you have a means to communicate with. Use the first and last name unless the recipient has given an Alias, or is personally familiar with the sender then first and middle is acceptable.

To Me, _______________________ (standard subscribed corrospondence)
To Me ( Myself | I ), ________ (standard recipient instigated corrospondence)
To Me Myself I, ______________ (look out, its your mother, and you're in big trouble;
                                nobody addresses a person by their actual full name)

Dear *(Mr./Mrs./Ms./Dr./Hon./Sen.) Me M. I *(I),
To Whom it may Concern;

Otherwise you are looking for something standard: hello, greetings, you may be a winner.

Where you have data that is a person's name all in one string, you don't have a problem because you already have their alias. If what you need is the first and last name, then just Left(name,instr(name," ")) & " " & Right(name,instrrev(name," ")), my math is probably wrong, i'm a bit out of practice. compare left and right with known prefixes and suffixes and eliminate them from your matches. Generally the middle name is rarely used except for instances of confirming an identity; which an address or phone number tells you a lot more. Watching for hyphanation, one can determine that if the last name is not used, then one of the middle ones would be instead.

For searching lists of first and last names, one must consider the possibility that one of the middle ones was instead used; this would require four searches: one to filter for first & last, then another to filter first & middle, then another to filter middle & last, and then another to filter middle & middle. Ultimately, the first name is always first, and the last is always last, and there can be any number of middle names; less is more, and where zero is likely, but improbable.

Sometimes people prefer to be called Bill, Harry, Jim, Bob, Doug, Beth, Sue, or Madonna; than their actual names; similar, but unrealistically expected of anyone to fathom all the different possibilities.

The most polite thing you could do, is ask; What can we call you?

Answer 3

You can do the obvious things: look for Jr., II, III, etc. as suffixes, and Mr., Mrs., Dr., etc. as prefixes and remove them, then first word is first name, last word is last name, everything in between are middle names. Other than that, there's no foolproof solution for this.

A perfect example is David Lee Roth (last name: Roth) and Eddie Van Halen (last name: Van Halen). If Ann Marie Smith's first name is "Ann Marie", there's no way to distinguish that from Ann having a middle name of Marie.

Answer 4

I would say Strip out salutations from a list then split by space, placing list.first() as first name, list.last() as last name then join the remainder by a space and have that as a middle name. And ABOVE ALL display your results and let the user modify them!

Answer 5

There are a few add-ins we have used in our company to accomplish this. I ended up creating a way to actually specify the formats for the name on our different imports for different clients. There is a company that has a tool that in my experience is well worth the price and is really incredible when tackling this subject. It's at: http://www.softwarecompany.com/ and works great. The most efficient way to do this w/out using any statistical approach is to split the string by commas or spaces then: 1. strip titles and prefixes out 2. strip suffixes out 3, parse name in the order of ( 2 names = F & L, 3 names = F M L or L M F) depending on order of string().

Answer 6

There is a 3rd party tool for this kind of thing called NetGender that works surprisingly well. I used it to parse a massive amount of really mal-formed names in unpredictable formats. Take a look at the examples on their page, and you can download and try it as well.

http://www.softwarecompany.com/dotnet/netgender.htm

I came up with these statistics based on a sampling of 4.2 million names. Name Parts means the number of distinct parts separated by spaces. A very high percentage were correct for most names in the database. The correctness went down as the parts went up, but there were very few names with >3 parts and fewer with >4. This was good enough for our case. Where the software fell down was recognizing not-well-known multi-part last names, even when separated by a comma. If it was able to decipher this, then the number of mistakes in total would have been less than 1% for all data.

Name Parts | Correct | Percent of Names in DB
 2             100%       48%
 3              98%       42%
 4              70%        9%
 5              45%        0.25%