Predicting Gender by First Name Using Character-level Machine Learning

TL;DR

This paper explores various machine learning models, including deep neural networks, to accurately predict gender from first names, demonstrating that recurrent models outperform feedforward ones with over 95% accuracy.

Contribution

It systematically compares multiple machine learning and deep learning models for gender prediction from first names, highlighting the effectiveness of recurrent neural networks.

Findings

Recurrent neural networks outperform feedforward models.

Models achieve over 95% accuracy in gender prediction.

Feature extraction from names as string sets is effective.

Abstract

Predicting gender by the first name is not a simple task. In many applications, especially in the natural language processing (NLP) field, this task may be necessary, mainly when considering foreign names. In this paper, we examined and implemented several machine learning algorithms, such as extra trees, KNN, Naive Bayes, SVM, random forest, gradient boosting, light GBM, logistic regression, ridge classifier, and deep neural network models, such as MLP, RNN, GRU, CNN, and BiLSTM, to classify gender through the first name. A dataset of Brazilian names is used to train and evaluate the models. We analyzed the accuracy, recall, precision, f1 score, and confusion matrix to measure the models' performances. The results indicate that the gender prediction can be performed from the feature extraction strategy looking at the names as a set of strings. Some models accurately predict gender in more than 95% of the cases. The recurrent models overcome the feedforward models in this binary classification problem.