Document Classification using predictive classifier

INTRODUCTION

This project develops a supervised machine learning model to classify sports-related text documents into categories such as Soccer, Basketball, and Tennis. We will use a labeled dataset to train the model, and its performance is evaluated by predicting categories for newly collected sports headlines. Our goal is to demonstrate how text classification methods can generalize to unseen data.

PLANNED APPROACH

To successfully solve this sports-related classification problem, our approach will consist of using a standard text classification pipeline in R as followed:

• First, a labeled dataset of sports documents will be preprocessed by cleaning and normalizing the text.

• Next, the preprocessed text will be transformed into numerical features using a Document-Term Matrix or Term Frequency–Inverse Document Frequency(TF-IDF) representation.

• We will then train A supervised learning algorithm on the labeled data.

• Finally, our trained model will be applied to new, unseen documents such as scraped sports headlines to predict their corresponding categories including Soccer, Basketball,Tennis, etc…