🌀 DisasterTweet : Detection 🔍 of Fake Disaster Tweets

This project was developed as a part of a Kaggle Playground Competition. DisasterTweet is a project that is developed to classify a tweet as a Real or Fake Disaster Tweet.
The project makes use of various NLP techniques to train ML models and then has chosen the best performing models.The following two methods were implemented during this project:
1️⃣ Bag Of Words Model
2️⃣ Word Embeddings

Competition Ranking

Ranked 441 with a Score of 0.81060 (F1-Score).
This implementation crossed the AutoML benchmark at the score of 0.80753.

Libraries Used For Development

1️⃣ Spacy
2️⃣ Keras
3️⃣ Tensorflow
4️⃣ Sklearn
5️⃣ Pandas