GenderPred-IN

GenderPred-IN is a Python package designed to predict the gender of a person based on their name. It is specifically tailored for Indian names, leveraging advanced machine learning techniques to provide prediction with accuracy ~96%.

PyPI

Features

• LSTM Model: Utilizes a Long Short-Term Memory (LSTM) neural network model to understand the sequential patterns in names.

• Tokenizer and Label Encoder: Processes names through a trained tokenizer and label encoder to convert names into a format suitable for the LSTM model.

• Pre-trained Model: Comes with a pre-trained model, eliminating the need for extensive training and setup.

• User-friendly: Simple and easy-to-use functions to get predictions with minimal setup.

Getting Started

Installation

You can install the package using pip (easy-peasy way):

pip  install  genderpred_in

or you can use github to install (harder way):

1. Clone the repository:

git clone https://github.com/DhrvM/GenderPred-India.git

cd GenderPred-India

3. Install the package:

pip install .

5. Verify the installation:

pip list

Usage

Import Package

from genderpred_in import classify_name, get_name, get_first_name, get_male_probability, get_female_probability, get_gender

Here is an example of how to use the package:

# Classify the name "Rohit"

result = classify_name("Rohit")



# Retrieve and print the results

full_name = get_name(result)

first_name = get_first_name(result)

male_prob = get_male_probability(result)

female_prob = get_female_probability(result)

gender = get_gender(result)



print(f"Full Name: {full_name}")

print(f"First Name: {first_name}")

print(f"Male Probability: {male_prob}")

print(f"Female Probability: {female_prob}")

print(f"Gender: {gender}")

Example Output:

Full Name: Rohit

First Name: ROHIT

Male Probability: 0.9916077852249146

Female Probability: 0.008392222225666046

Gender: male

Functions

• classify_name(full_name): Classifies the given full name and returns a dictionary with the name, first name, gender, and probabilities.

• get_name(result): Retrieves the full name from the classification result.

• get_first_name(result): Retrieves the first name from the classification result.

• get_male_probability(result): Retrieves the male probability from the classification result.

• get_female_probability(result): Retrieves the female probability from the classification result.

• get_gender(result): Retrieves the predicted gender from the classification result. (Output: male, female, unknown)

Versions

Version 1.0.2 Fixed Model loading error for Windows.\

Version 1.0.1 Uses LSTM model with a tokenized First-Name to Generate Predictions of Gender.

Built With

• TensorFlow - The machine learning framework used

• Keras - High-level neural networks API

• NumPy - Used for numerical computing

• Pandas - Data manipulation and analysis

Authors

Dhruv Malpani - Initial Work

LinkedIn

GitHub

License

This project is licensed under the MIT License - see the LICENSE.md file for details.

Acknowledgments

Praneeth Vasarla

Your article helped me create the initial model using Logistic Regression and n-grams. (article)