pyJedAI

What pyJedAI is

pyJedAI is a free, open-source Python framework for entity resolution and link discovery. Its documented workflow reads structured or semi-structured data, including pandas DataFrames, then builds and cleans blocks, matches entities, clusters results, and presents scores and visualizations. The project lists supervised and unsupervised machine-learning techniques, and says matching and blocking can also use deep learning and natural language processing. A tutorial demonstrates embedding-based blocking with PyTorch and Gensim vectorizers, followed by nearest-neighbor search through FAISS. Documentation also covers schema matching through Valentine methods such as Similarity Flooding, Cupid, COMA, EmbDI, and SemProp. A geospatial interlinking component includes spatial filtering, initialization, and verification. Installation options include PyPI, a local Git checkout, and Docker. The project says it has been tested on Windows and Linux and requires Python 3.8 or later; Windows also requires Microsoft Visual C++ 14.0. pyJedAI is a research project by the AI-Team of the University of Athens Department of Informatics and Telecommunications, and recommends GitHub Discussions for general questions and its issue tracker for bugs.