PROJECT | Natural Language Processing Challenge

Introduction

This project focuses on Natural Language Processing (NLP) techniques to classify news headlines as real or fake news. Processing text is a core skill for Data Scientists and AI Engineers, and this task allows you to apply these skills in practice.

Project Overview

• Dataset: The main dataset is located at in this Github Repo and contains the following columns:
  • label: 0 for fake news, 1 for real news
  • title: Headline of the news article
  • text: Full content of the article
  • subject: Category or topic of the news
  • date: Publication date of the article
• Goal: Building a classifier to predict whether a news article is real or fake.

Project Files

• model_W2V_XGBoosting.py: Contains the complete Python code for building and training the Word2Vec + XGBoost model.
• model_explanation.md: Explains all the steps in the model development pipeline.
• accuracy_estimation.md: Lists the model performance scores and evaluation metrics.
• fp_3.csv: Contains the predicted labels (0 or 1) for the validation dataset.