/gender_newspaper

This is files for me undergraduate thesis on detecting gender bias in Bangladeshi newspapers.

Primary LanguageJupyter Notebook

gender_newspaper

word2vec_similarity_statistics.ipynb > youngladesh@gmail.com

Description:

• Load 2 trained models
• Calculate similarity between 2 different Gensim word2vec model using numpy
• Make statistical calculations with the models, such as Mean, Standar Deviation
• Calculate number of words outside Standard Deviation
• Print the words outside norm
• Plot the histogram of words using Matplotlib
• Runtime calculation

newspaper word2vec gensim.ipynb > youngladesh@gmail.com

Description

• Install Gensim and Download Newspaper data
• Train Gensim word2vec with the Newspaper data
• Train FastText with Newspaper data
• Save and Load trained model
• Failed attempt to Data Visualization

word2vec gensim.ipynb > youngladesh@gmail.com

Description

• Gensim simple tutorial
• Gensim training with 16MB TED talk dataset
• Basics of Gensim
• FastText

word2vec with name replaced dataset.ipynb > masum.mc2@gmail.com

Description

• Download the Newspaper dataset
• Make 4 different dataset with it
• Train 4 different word2vec models with the dataset
• Most important file for me thesis

word2vec_diff.ipynb > masum.mc2@gmail.com

Description

• Many kind of operations with Gensim word2vec model
• Playing with word2vec
• Comparison between 2 models