🎬 Blockbuster Movies Prediction

📌 Project Overview

This project aims to build a Machine Learning model that predicts whether a movie can be considered a blockbuster — both in terms of audience reception and financial success.

🔥 What Is a Blockbuster?

A blockbuster movie is generally defined as a film that achieves both high profitability and high public acclaim. To capture this, we define two measurable targets:

• IMDB Score: Proxy for audience satisfaction and critical reception.
• ROI (Return on Investment): Proxy for commercial success.

🎯 Project Goals

• Perform Exploratory Data Analysis (EDA) to understand patterns behind blockbuster movies.
• Build predictive models for:
  1. IMDB Score
  2. ROI
• Optionally, create a combined model to classify blockbuster likelihood based on both.

---

📂 Dataset

Datasets for Blockbuster Movies Analysis

Our goal is to build a comprehensive dataset of blockbuster movies and find a model that optimizes all the information we got. We'll combine information from multiple sources. Below are some datasets that align with our project requirements:

---

1. Movie Data Analysis Dataset

• Details about 7,668 movies, including:
  • Titles, ratings, genres, release years
  • IMDb scores, votes
  • Directors, writers, main stars
  • Production countries, budgets, gross earnings
  • Production companies, runtimes
• Source: GitHub Repository

---

2. Global Movie Franchise Revenue and Budget Data

• Comprehensive data on movie franchises worldwide between 2000–2020:
  • Lifetime gross, budget, rating
  • Runtime, release date, vote count/average
• Source: Kaggle Dataset

---

3. TMDB 5000 Movies Dataset

• Information on over 5,000 movies:
  • Budget, cast, director
  • Keywords, runtime, genres
  • Production companies, release dates
• Source: Hugging Face Dataset

---

4. Complete Movie Metadata Dataset

• Data on over 722,000 movies, including:
  • ID, title, genres, budget, revenue
• Suitable for analyzing trends in movie popularity, production companies, budgets, and revenues.
• Source: Gigasheet Dataset

---

5. Movie Revenue Analysis Dataset

• Approx. 1,800 movies released between 1915 and 2020:
  • Domestic and worldwide gross revenues
  • Production budgets, release dates
• Source: GitHub Repository

---

These are merged and cleaned into:

• imdb_score_features
• roi_features
• master_table

---

🛠️ Methods & Tools

• Python (Pandas, Scikit-learn)
• Feature engineering based on domain knowledge and EDA
• ML models (regression, classification)

---

🤖 Why IMDB Score and ROI?

• IMDB Score reflects whether people liked the movie — crucial for sustained popularity and brand value.
• ROI captures whether the film was a financial hit — crucial for studios and investors.

Combining these helps us define and detect "blockbusters" more holistically.

---

🚀 Future Work

• Incorporate marketing or release strategy data (e.g., release date, streaming vs theater)
• Refine model into binary classification: "Blockbuster vs Not"