Dynamic Player Transfer Value Prediction using AI and Multi-source Data.
👉 Try Transfer-IQ on Streamlit
Transfer-IQ is a data-driven analytics project designed to predict and analyze football player market values. By integrating multi-source data—including on-pitch performance metrics, injury histories, and social media sentiment—the project aims to build a robust, machine-learning-ready dataset that uncovers the hidden factors influencing a player's valuation in the modern transfer market.
To build a comprehensive profile for each player, data was aggregated from multiple domains:
- StatsBomb: Player performance and match data.
- Transfermarkt: Historical and current market value valuations.
- Social Media: Public sentiment signals processed via NLP (VADER / TextBlob).
- Injury Records: Player availability, historical absences, and future risk assessment.
Our dataset (~270 players) relies on advanced feature engineering to capture true player value:
- Goal Conversion Rate & Minutes Per Goal: Efficiency metrics for attackers.
- Performance Index: A normalized score combining various on-ball and off-ball actions.
- Injury Risk Score: A predictive metric based on past physical reliability.
- Sentiment Polarity: Public perception scores to gauge marketability and reputation.
- Multi-source data collection and alignment.
- Player identity standardization across different APIs and datasets.
- Missing value handling, imputation, and deduplication.
- Initial base feature engineering.
- NLP Sentiment analysis using VADER and TextBlob.
- Creation of advanced performance efficiency metrics.
- Injury risk feature generation.
- Finalized modeling-ready dataset.
- Predictive ML modeling (Regression/XGBoost).
- Model evaluation and hyperparameter tuning.
- Streamlit web application deployment.
| Category | Technologies |
|---|---|
| Language | Python |
| Data Processing | Pandas, NumPy |
| Machine Learning | Scikit-learn |
| NLP | VADER, TextBlob |
| Deployment | Streamlit |
| Version Control | Git & GitHub |
Transfer-IQ/
│
├── data/
│ ├── raw/ # Original, immutable data dumps
│ ├── interim/ # Intermediate data that has been transformed
│ └── processed/ # The final, canonical datasets for modeling
│
├── notebooks/ # Jupyter notebooks for EDA and prototyping
├── src/ # Source code for data pipeline and feature engineering
├── reports/ # Generated analysis as HTML, PDF, or markdown
├── app.py # Streamlit application entry point
└── README.md # Project documentation
To run this project locally, follow these steps:
-
Clone the repository:
git clone https://github.com/springboardmentor67/Transfer-IQ.git cd Transfer-IQ -
Install dependencies:
pip install -r requirements.txt
-
Run the Streamlit App:
streamlit run app.py
Abhishek Kumar
Infosys Springboard Internship Project