Monday, March 26, 2012

Gender of Names

Does anyone know of a way to identify the gender of a person's name in a table? I have found an application that does it but I have to export it and use another program to analyze the data and then importing it back into SQL. We have a table that has a list of customers names that we use to market and we need this tool to analzye the customers by gender.

Thank you in advance for your help.

xroadtripx wrote:

Does anyone know of a way to identify the gender of a person's name in a table? I have found an application that does it but I have to export it and use another program to analyze the data and then importing it back into SQL. We have a table that has a list of customers names that we use to market and we need this tool to analyze the customers by gender.

There's nothing inherent in SQL Server to estimate the gender of a name for a given language.

Probably the most common method is simply a look-up, into a table to names and gender association or probability. There's a few gender tables on the Web if that's the approach you want to take.

Besides a straight look-up, there name analysis based on language and statistics--which really just breaks a name into several look-ups to account for "unseen" names. For example, http://cl4.org/comp/gender/, describes an algorithm for estimating the gender of a name by combining the gender probability of each trigraph in a name. There's also been some research on gender estimation based on hyphenation. See http://junobeach.cs.uwaterloo.ca/~warren/publications/warren:isi:2004/check_name_gender.pdf

I think a combination of look-up and statistical analysis would be the most accurate.

|||This sounds to me like I would manually have to enter names for it to work. Is this correct? I think I should have been more clear in my question. We have a table of about 1 milllion records we need to check to determine the gender of the names and, I was hoping I guess, to be able to run some sort command or something to check and genderize the names in the table. Thank you for your help and maybe I am a little unclear on how your answer would work..

No comments:

Post a Comment