COVID-19 Tweet Sentiment Analysis Using CNN, RNN and LSTM

Overview

This repository contains a Jupyter Notebook that demonstrates sentiment analysis on COVID-19-related tweets using deep learning techniques. The goal is to classify tweets into sentiment categories (e.g., Negative, Positive, Other) by exploring and comparing three deep learning architectures:

• Convolutional Neural Network (CNN)
• Artificial Neural Network (ANN)
• Long Short-Term Memory (LSTM) Network

These models help reveal patterns in public sentiment during the COVID-19 pandemic.

Dataset

The dataset is sourced from Kaggle: Coronavirus tweets NLP - Text Classification. It comprises tweets related to the COVID-19 pandemic that have been manually tagged for sentiment. Key columns in the dataset include:

• Location: The origin of the tweet.
• Tweet At: Timestamp of when the tweet was posted.
• Original Tweet: The text content of the tweet.
• Label: The manually assigned sentiment (e.g., Negative, Positive, Other).

Features

• Data Preprocessing: Clean and tokenize tweet texts to prepare for model input.
• Modeling: Implementation and comparative analysis of three models:
  • CNN: To capture local features from text sequences.
  • ANN: A baseline deep learning model for text classification.
  • LSTM: To capture sequential dependencies and context in tweets.
• Evaluation & Visualization: Analyze model performance using accuracy metrics and visualize sentiment distribution and trends.

Requirements

• Python 3.x
• Jupyter Notebook or JupyterLab

Required Python packages:

• pandas
• numpy
• nltk
• tensorflow (or keras)
• matplotlib
• seaborn

Installation & Setup

1. Clone the Repository:

   git clone https://github.com/YourUsername/COVID19-Sentiment-Analysis.git

2. Navigate to the Project Directory:

   cd COVID19-Sentiment-Analysis

Usage

• Launch Jupyter Notebook:

  jupyter notebook

• Open the sentiment_analy.ipynb notebook and run the cells sequentially to preprocess the data, train the models, and visualize the results.

File Structure

• sentiment_analy.ipynb: The main notebook with the complete sentiment analysis workflow.
• data/: Directory to store the COVID-19 tweets dataset (or instructions to download it from Kaggle).