Gender Classifier for India

In-group bias in the Indian judiciary: Evidence from 5 million criminal cases

Elliott Ash, Sam Asher, Aditi Bhowmick, Sandeep Bhupatiraju, Daniel Chen, Tanaya Devi, Christoph Goessmann, Paul Novosad, and Bilal Siddiqi

For every name that you provide, this tool will calculate the probability that the name belongs to a female/male individual. The classifier is a bidirectional LSTM trained on a dataset of 13.7 million names from the delhi voter rolls, and returns a value between 0 and 1 which is the probability that the name is male.

For more information and for replication code, please see the paper codebase.

How to provide your data

Please only upload files that abide by the following standards (sample data_structure.csv):

• Format: *.csv (with one header row)
• Maximum file size: < 100 KB (corresponds to a few thousand names)
• Structure: One column, with first row containing the string "name" as column title
• Note: be sure to remove special characters! These cause the model to break