🎯 Tailored Career Prediction Based on Emotional & Personality Traits

This project predicts suitable career roles for an individual based on their personality profile and emotional traits using the Big Five Personality Model (OCEAN).
The model takes in personality scores or processes a short video interview to extract traits, then predicts multiple career options.

---

📌 Overview

The system uses:

• Machine Learning (Random Forest, SVM, KMeans, KNN)
• Video & Audio Processing (extract personality cues from speech and facial expressions)
• Multi-label classification to recommend more than one career if applicable

---

📂 Dataset

Note: The dataset is not included in this repository for privacy reasons.

How it was created:

• Collected 409+ interview videos from public sources (e.g., YouTube).
• Trimmed videos to show only the person speaking.
• Three annotators scored each person on the Big Five personality traits:
  • Openness
  • Conscientiousness
  • Extraversion
  • Agreeableness
  • Neuroticism
• Annotators discussed disagreements until consensus was reached.
• Mapped personality profiles to one or more suitable career roles.
• Stored data in a CSV file (merged_big5_scores.csv) with the following columns:

  Video_ID, Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism, Career Role
  

• This CSV was used to train, validate, and test the ML models.

You can replicate this dataset creation process or adapt it with your own video data.

---

✨ Features

• 📹 Video-based personality analysis
• 🧠 Machine Learning prediction models (Random Forest, KMeans, SVM, KNN)
• 📊 Multi-label classification for multiple career suggestions
• 🖥 Jupyter Notebook & Python script implementations
• 🔍 Optional integration with DeepFace for facial emotion recognition
• ⚡ Predictions in under 5 minutes per video

---

🛠 Tech Stack

Languages: Python 3.8+
Libraries & Frameworks:

• pandas, numpy, scikit-learn, seaborn, matplotlib
• joblib (model saving/loading)
• moviepy, SpeechRecognition (audio/video processing)
• deepface (emotion detection)
• ipywidgets (video upload in Jupyter)

---

📦 Installation

# Clone the repository
git clone https://github.com/ashish10-aga/tailored-career-prediction.git
cd tailored-career-prediction

# Create a virtual environment (optional)
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

---

🚀 Usage

1️⃣ Running via Python script

python model_train.py

Ensure you have a CSV in the same format as described in the Dataset section.

2️⃣ Running via Jupyter Notebook

jupyter notebook model_train_video.ipynb

Upload a video through the provided widget to get predictions.

---

📂 Project Structure

personality-model/
│-- model_train.py              # Main ML pipeline script
│-- model_train_video.ipynb     # Notebook with video upload & prediction
│-- requirements.txt            # Dependencies
│-- README.md                   # Project documentation
│-- .gitignore                  # Files to ignore in Git

---

📊 Example Prediction

Input Traits:
Openness=3.5, Conscientiousness=4.0, Extraversion=3.2, Agreeableness=4.1, Neuroticism=2.5

Predicted Careers:
['Data Analyst', 'Business Analyst', 'Project Manager']

---

📈 Results

• Random Forest Accuracy: ~84%
• KMeans Clustering: Groups candidates into 4 clusters mapped to career roles
• Multi-label Classification: Supports multiple job predictions per individual

---

📌 Future Work

• Deploy as a web or mobile app
• Add multilingual support
• Expand dataset for more diverse career coverage
• Improve accuracy with deep learning models

Trigger language refresh