top of page

Geospatial Aspect-Based Sentiment Analysis for Big NLP Data

To understand and analyse potential biases in machine learning, I created a Jupyter Notebook (hosted on GitHub) that details the following:

 

  • An Exploratory Data Analysis

  • Creating a Machine Learning Classifier (Rapidminer Go)

  • Investigating Bias

  • Interpretation and Recommendations

To perform the bias and fairness audit, the Aequitas Python library was used:

http://www.datasciencepublicpolicy.org/projects/aequitas/ 

aequitas_header.png
Mediamodifier-Design-Template (35).png
bottom of page