pyJedAI
github.com
What it is
pyJedAI is a free, open-source Python framework for creating entity-resolution workflows. It addresses record linkage, entity resolution and link discovery across structured and semi-structured data. Its collection of algorithms includes supervised and unsupervised machine-learning methods. The project says its blocking and matching methods can apply deep learning and NLP without requiring users to provide labeled instances. Installation is available through pip, Git checkout or Docker. The README says pyJedAI has been tested on Windows and Linux and requires Python 3.8 or later; Windows also requires Microsoft Visual C++ 14.0. It uses packages including NLTK, pandas, transformers, sentence-transformers and FAISS CPU. pyJedAI is licensed under Apache-2.0. The project describes it as suitable for both expert and novice users, and its package metadata also identifies developer and science/research audiences. GitHub Discussions are recommended for general questions, while bugs can be reported through GitHub Issues. The README points Java users to JedAI, which includes a web app for building workflows interactively.
Spec sheet
- Purpose
- pyJedAI is an open-source Python framework for building end-to-end entity resolution workflows.Read 4 Oct 2026 · github.com
- Audience
- The project describes pyJedAI as intended for expert and novice users, and its package metadata lists developers and science/research audiences.Read 4 Oct 2026 · github.com
- Integrations
- The project uses Python data-science and machine-learning packages, including NLTK, pandas, transformers, sentence-transformers, and FAISS CPU.Read 4 Oct 2026 · github.com
- Support
- The project recommends GitHub Discussions for general questions and discussions and GitHub Issues for bugs.Read 4 Oct 2026 · github.com
- License
- pyJedAI is released under the Apache-2.0 license.Read 4 Oct 2026 · github.com
- Algorithms
- It offers multiple algorithms and supervised and unsupervised machine learning techniques.Read 4 Oct 2026 · github.com
- Data types
- It can process structured and semi-structured data.Read 4 Oct 2026 · github.com
- Install
- The project provides installation through pip, Git checkout, or Docker.Read 4 Oct 2026 · github.com
- Learning methods
- The project says its blocking and matching methods can use deep learning and NLP techniques without requiring labeled instances from the user.Read 4 Oct 2026 · github.com
- Operating systems
- The README says pyJedAI has been tested on Windows and Linux.Read 4 Oct 2026 · github.com
- Problem types
- The framework targets record linkage, entity resolution, and link discovery problems.Read 4 Oct 2026 · github.com
- Related web app
- The README points Java users to JedAI, which includes a web application for interactive creation of entity-resolution workflows.Read 4 Oct 2026 · github.com
- Requirements
- The README lists Python 3.8 or later and says Windows requires Microsoft Visual C++ 14.0.Read 4 Oct 2026 · github.com
- Security policy
- The security policy lists version 0.0.x as supported with security updates and directs users to open an issue to report a bug.Read 4 Oct 2026 · github.com
Plans and prices
Strong points and limits
Strong points
- Works with structured and semi-structured data
- Offers supervised and unsupervised methods
- Can use deep learning and NLP without labeled instances
- Install via pip, Git checkout or Docker
- Free under the Apache-2.0 license
Limits
- Requires Python 3.8 or later
- Windows requires Microsoft Visual C++ 14.0
- Security policy lists supported version 0.0.x
Who it suits
It suits developers and researchers, as well as expert or novice users building record-linkage and entity-resolution workflows in Python. Windows users need Microsoft Visual C++ 14.0 in addition to Python 3.8 or later.
Verdict
pyJedAI provides several ways to build entity-resolution workflows, with data formats, methods and installation options stated. Check its Python and Windows requirements before setting it up.
Where Specifiction places it
- Identity Resolution Software#10 of 257.6