jaotheboss / Drug-Sentiment-Analysis

This project was done under the NUS module DSA4262, of which it focuses on projects that center around medicine and healthcare.

Drug-Sentiment-Analysis

This project was done under the NUS module DSA4262, of which it focuses on projects that center around medicine and healthcare. I wrote an article regarding this topic, here is a link.

Overview

1. The nature of this project is natural language processing
2. Each review has a rating between 1 - 10
3. Objective is to predict/classify other reviews as to whether they're positive or negative

Methodology

1. Tokenizing (Cleaning/Preprocessing)
2. Exploratory Data Analysis (Exploration)
3. Stemming (Cleaning/Preprocessing)
4. Lemmatizing (Cleaning/Preprocessing)
5. Vectorizing/OneHotEncoding (Cleaning/Preprocessing)
6. Modelling (Cleaning/Preprocessing)
7. Predicting (Machine Learning)
8. Exploration of Final Model (Exploration)
9. Feature Analysis (Exploration)

Models Used:

1. Logistic Regression
2. Decision Trees
3. XGBoost