- Gender Classification
python scripts/gender_classifier.py etc/params.txt
- Country Classfication
tar zxf data/model/country.pkl_02.npy.tgz
python scripts/country_classifier.py etc/params.txt
The model was a decison tree with n-gram features, trained on Wikipedia. We subsampled on male names to balance to ratio of male to female. We inferred gender from the number of occurences of masculine words like "he", "him" or "his" and feminine words like "she", "her". We inferred country based on nationality and place of birth.