BLOG

A NEW VERSION OF OUR API IS AVAILABLE - PART III

In our third announcement, we would like to introduce our next improvement: Enhanced support for variant forms of spelling.

Sometimes customers enjoy writing their names in different ways. A very common way to do this is using Leetspeak to replace chars with similar glyphs.

For example, the name John is written as J0hn.

To enhance support for names spelt like this, we added a new field called name_sanitized to the API response.

This field gives you the name we found and used for genderization after the name was sanitized in different ways.

IMPORTANT UPDATES ABOUT THE GENERAL DATA PROTECTION REGULATION (GDPR)

The new European General Data Protection Regulation (GDPR) goes into effect on May 25, 2018. We can assure you that all Gender-API.com services will be in full compliance with GDPR by this date.

Gender-API.com provides a Data Processing Agreement (DPA), which enables you to comply with your GDPR obligations.

In addition, we have updated our Privacy Policy and General Terms and Conditions to comply with the new requirements of GDPR.

A NEW VERSION OF OUR API IS AVAILABLE - PART I

Good news, during the last couple of months, our team has worked very hard on a full rewrite of our name lookup engine.

We improved so many great details that we couldn't just put them into one single announcement anymore.

So here is our first improvement, we would like to share with you: Reduced split call failure rate.

Our split API is one of our customers' most used features. However, we understand that it can be very tricky to split a full name into its parts.

With the new API version, we are able to reduce the failure rate of all split calls by about 40%. With the old API version, about 90% of the calls returned the correct result. The new version improves this performance to 94% as internal tests show.

To achieve this improvement, we completely refactored our codebase and developed specific micro-services, which can also be more easily improved in the future.

The advancement also includes improved support for countries where Gender-API.com is already heavily in use, for example Germany or the Netherlands. Names with an affix like "Thomas van der Weijden" are now properly resolved into their consisting parts.

WE IMPROVED THE ACCURACY OF OUR EMAIL API BY 30%

During the last few months, we revised and completely redesigned the backend of our email interface in 3 steps. With the new interface it is now possible to break up email addresses that do not have a clearly distinguishable separator between the first and last name. Before the update, names had to be separated by a dot or hyphen for the email address to be properly broken up. With the new interface, email addresses such as kathleenevans44@gmail.com can now be broken up without any problems. How do we do this: We implemented a multi-stage process using machine learning to check the frequency of any possible first and last name combination to provide the best possible result for each case. Internal tests and analysed have shown that result quality improved by more than 30%. 
GET https://gender-api.com/get?email=kathleenevans44@gmail.com&key=<your api key>
The request will be returned as: 
{"email":"kathleenevans44@gmail.com","lastname":"evans","mailprovider":"gmail","name":"kathleen","gender":"female","samples":533,"accuracy":99,"duration":"211ms"}

OUR SPLIT API LEFT ITS BETA STATE

After a year of testing, our split API left its beta state. The API was constantly improved since we first launched it with some of our biggest customers a year ago. The API can split full names (first and last name) into its parts. If you have a combined field for first and last name on your page, use this API to extract the parts: 
GET https://gender-api.com/get?split=theresa%20miller&key=<your api key>
The request will be returned as: 
{"last_name":"Miller","first_name":"Theresa","strict":false,"name":"theresa","gender":"female","samples":8065,"accuracy":98,"duration":"56ms"}
fast_rewind
Newer entries Older entries
fast_forward
Chat