Bach4Popper

Bach4Popper is a coordination-based framework for Distributed Inductive Logic Programming (ILP). It extends the Popper ILP system by decomposing its generate–test–constrain loop into coordinated roles executed across multiple clients. In this setting, hypotheses are generated centrally, evaluated locally on partitioned datasets, and aggregated symbolically without sharing raw data. The system combines the Popper ILP engine (v1.1.0) with BLPy, a Python implementation of the Bach coordination model providing a shared tuple-space store and coordination primitives. This separation enables federated symbolic learning while preserving the semantics of centralized ILP. The artefact is distributed as a self-contained Docker environment to ensure platform-independent reproducibility and includes all scripts and datasets required to reproduce the experiments reported in the accompanying paper. A recorded demonstration video (bach4popper-demo-final) is included in the Zenodo archive to illustratethe execution workflow and distributed coordination behavior.