---

🔍 Twitter Sentiment Analysis & ML Model Evaluation 🚀

📌 A powerful Streamlit-based web application for analyzing Twitter data, performing sentiment analysis, extracting key insights using TF-IDF, and building machine learning models to predict engagement metrics (likes, retweets, replies, etc.).

---

📜 Table of Contents

• 📌 Features
• 🚀 Installation
• 📂 Dataset Requirements
• ⚙️ How It Works
• 📊 Data Processing & Visualization
• 🤖 Machine Learning Models
• 📥 Downloading Processed Data
• 📸 Screenshots
• 🛠 Technologies Used
• 🙌 Contributing
• 📜 License

---

📌 Features

✅ Load Twitter Data from CSV 📂
✅ Text Preprocessing & Normalization (Tokenization, Stopword Removal, etc.)
✅ TF-IDF Vectorization for Keyword Extraction 🔠
✅ WordCloud & Bar Charts for Word Frequency Analysis 🌥📊
✅ VADER Sentiment Analysis (Positive, Negative, Neutral) 🎭
✅ Sentiment Trend Over Time 📉
✅ Correlation Heatmap of Twitter Features 🔥
✅ Machine Learning Models for Engagement Prediction 🧠
✅ Confusion Matrix, ROC Curve, and Feature Importance Analysis 📈
✅ Download Processed Data & TF-IDF Features as CSV 📥

---

🚀 Installation

📌 1️⃣ Clone the Repository

git clone https://github.com/yourusername/twitter-sentiment-ml.git
cd twitter-sentiment-ml

📌 2️⃣ Create a Virtual Environment

python -m venv venv
source venv/bin/activate   # MacOS/Linux
venv\Scripts\activate      # Windows

📌 3️⃣ Install Dependencies

pip install -r requirements.txt

📌 4️⃣ Run the Streamlit App

streamlit run app.py

📌 Open the URL in your browser: https://app3-xv4r4cqlzqddwfyrux9qgv.streamlit.app/

---

📂 Dataset Requirements

The application requires a CSV file named filtered_data.csv with the following columns:

Column Name	Description
text	Tweet content
timestamp	Date & time of the tweet
likes	Number of likes on the tweet
retweets	Number of retweets
replies	Number of replies
quotes	Number of quote tweets
isReply	1 if it's a reply, 0 otherwise
isRetweet	1 if it's a retweet, 0 otherwise
isQuote	1 if it's a quote tweet, 0 otherwise
verified	1 if the user is verified, 0 otherwise

---

⚙️ How It Works

1️⃣ Load Data 📂

• Reads the Twitter dataset (filtered_data.csv) into a Pandas DataFrame.
• Prepares text data for analysis.

2️⃣ Preprocessing & Tokenization 📝

• Tokenizes text using spaCy.
• Removes stopwords and punctuation.

3️⃣ TF-IDF Vectorization 🔠

• Converts tweets into numerical features using TF-IDF (Term Frequency - Inverse Document Frequency).
• Extracts the top 500 important words from the dataset.

4️⃣ Sentiment Analysis 🎭

• Uses VADER Sentiment Analysis to categorize tweets into Positive, Negative, or Neutral.
• Plots sentiment distribution using Seaborn.

5️⃣ Machine Learning Model Training 🧠

• Trains Linear Regression, Logistic Regression, Random Forest, and XGBoost models to predict engagement metrics.
• Evaluates models with MSE, Accuracy, Confusion Matrix, and ROC Curve.

6️⃣ Download Processed Data 📥

• Allows users to download cleaned & processed data for further analysis.

---

📊 Data Processing & Visualization

🔍 TF-IDF Feature Importance

• Displays the Top 20 most important words in the dataset using a bar chart.

🌥 WordCloud for Top Words

• Generates a WordCloud to visually represent the most frequent words in tweets.

📉 Sentiment Trend Over Time

• Plots Average Sentiment Score over time to identify trends in public opinion.

🔥 Feature Correlation Heatmap

• Shows how different features (likes, retweets, replies) correlate with each other.

---

🤖 Machine Learning Models

Model	Type	Usage
Linear Regression	Regression	Predicts number of likes on a tweet.
Logistic Regression	Classification	Predicts if a tweet will get above-average likes.
Random Forest	Classification	Predicts high-engagement tweets using ensemble learning.
XGBoost	Classification	Predicts engagement with boosting techniques.

---

📥 Downloading Processed Data

The app allows you to download:
✅ TF-IDF Features in CSV format.
✅ Processed Twitter Data with sentiment scores.

---

🛠 Technologies Used

✅ Python 🐍
✅ Streamlit 🎨
✅ Pandas, NumPy 🏗
✅ spaCy 🔠 (Text Processing)
✅ VADER Sentiment Analysis 🎭
✅ Matplotlib, Seaborn 📊 (Data Visualization)
✅ Scikit-learn, XGBoost 🤖 (Machine Learning)
✅ WordCloud 🌥 (Text Visualization)

---

🙌 Contributing

📌 Want to improve this project? Contributions are welcome!

• Fork this repository.
• Create a feature branch (git checkout -b new-feature).
• Commit your changes (git commit -m "Added new feature").
• Push to the branch (git push origin new-feature).
• Submit a pull request! 🚀

---

📜 License

This project is licensed under the MIT License.

---

🚀 Enjoy Analyzing Twitter Data with AI & Machine Learning! 🎉

---