🛒 Predicting Consumer Purchase Intention using Twitter Data

Final Year Project II (FYP-II) — All material related to this project is uploaded here.

---

📖 Overview

This project leverages Twitter data and machine learning to predict whether a user has purchase intention (PI) towards a product. By analysing tweet content, we classify users as either showing purchase intention or not, and surface actionable insights through an interactive web application.

---

🎯 Project Details

1. 🐦 Data Collection — Scrape Twitter tweets using the Twitter Search API to build a labelled dataset.
2. 🏷️ Annotation — Label each tweet as PI (Purchase Intention) or No PI.
3. 📊 Exploratory Data Analysis — Inspect tweet types, check for class imbalance, generate word clouds and visualisations.
4. 🔧 Pre-processing — Apply text cleaning and feature-extraction techniques to build a corpus.
5. 🤖 Machine Learning Models — Train and evaluate multiple classifiers:
   • Support Vector Machine (SVM)
   • Naive Bayes
   • Logistic Regression
   • Decision Tree
   • Neural Network
6. 📋 Output — Produce a ranked list of customers who have shown purchase intention.
7. 🌐 Web Application — A Django-powered website that summarises results and lets users upload their own dataset to train/test the pre-built models.

---

🗂️ Repository Structure

Folder / File	Description
data/	Annotated CSV datasets used for training and testing
models/	Python scripts for each ML model and pre-processing pipeline
posters/	Project poster and standee artwork
SYMPOSIUM/	Presentation slides, final report, and project completion certificate
PurchaseIntention2.zip	Complete Django website source code
Model Evaluation - Sheet1.pdf	Comparative model evaluation results

---

⚙️ Setting Up Dependencies

Make sure you have Python (latest version) installed, then install the required packages.

📦 Install packages

pip3 install django numpy pandas nltk textblob scikit-learn

📥 Download TextBlob corpora

python -m textblob.download_corpora

📥 Download NLTK stopwords

Open a Python terminal and run:

import nltk
nltk.download('stopwords')
exit()

✅ All required dependencies are now installed.

---

🚀 How to Run the Website

1. Clone the repository and unzip PurchaseIntention2.zip.
2. Open Windows Command Prompt (cmd) and navigate to the PurchaseIntention2 folder.
3. Activate the virtual environment:

   Scripts\activate

4. Navigate to the Django project:

   cd djangoPIWebsite

5. Start the development server:

   python manage.py runserver

6. Open your browser and go to localhost:8000 🎉
7. To stop the server press Ctrl + C. To restart, re-run python manage.py runserver.

---

🧪 How to Run the Model Scripts Directly

1. Navigate to CIP\PurchaseIntention2 and activate the virtual environment:

   Scripts\activate

2. Navigate to the pages directory:

   cd CIP\PurchaseIntention2\djangoWebsite\pages

3. Open a Python terminal:

   python

4. Import and run the model test module:

   import ModelTest as mt
   mt.output_to_results("Annotated4.csv", "AnnotatedData2.csv", "TF-IDF", "Naive Bayes", "90", "80", "70")

5. 📈 The output will display prediction results and the accuracy score for the tested model.

---

🤝 Contributing

Contributions are welcomed and encouraged! Whether it's fixing a bug, improving the documentation, adding a new feature, or suggesting enhancements — all input is appreciated.

To contribute:

1. Fork this repository.
2. Create a new branch for your feature or fix:

   git checkout -b feature/your-feature-name

3. Make your changes and commit them with a clear message:

   git commit -m "Add: description of your change"

4. Push your branch to your fork:

   git push origin feature/your-feature-name

5. Open a Pull Request against the main branch of this repository and describe what you've changed.

Please make sure your contributions align with the project goals and maintain code quality. For major changes, consider opening an issue first to discuss your proposal.

---

📜 License

This project is licensed under the terms of the LICENSE file included in this repository.