Name Analyzer takes any first name (e.g. Christian, Mike, Anna) as input, applies some machine learning magic and returns three things: A prediction if the name is real or fake, a likelihood if the name is male or female and a mapping to one of nine regions: German, English, Iberian (including Spain, Portugal and Latin America), Italian, French, African, Arabic (from Tunesia to Iran), East-Asian (from India to Japan and South-East-Asia) and Slavic (Eastern Europe and Russia). Name Analyzer is using Azure Machine Learning as a backend, the machine learning models were created in Azure Machine Learning Studio. As classification method, a two-class boosted decision tree was used to label the data. For training, this data file was used.
Name Analyzer is not using any pre-trained models, the training, scoring and evaluation was done based on the file linked above in Machine Learning Studio. A basic introduction into machine learning in general can be found here: A Friendly Introduction to Machine Learning . If you want to learn more about machine learning on azure, try this training: Using Azure Machine Learning to Predict Who Will Survive the Titanic.