datamining

This project shows how we can classify bad and good comments from extracted and put in a CSV file. The data is in Tunisian Dialects, which makes it really hard to clean, as there are different ways of writing a word. We used a basic machine learning model to divide comments, bad ones are labeled as -1 whereas good ones were labeled as +1. At the very least, the model gave an accuracy of 75%.