This repository contains benchmark instances for evaluating planning action model learning algorithms.
In particular, it contains the benchmarks used in our paper "Planning Domain Model Acquisition from State Traces without Action Parameters," published at the Special Track on Reasoning, Learning, & Decision Making at the 21st International Conference on Knowledge Representation and Reasoning (KR-2024).
The most up-to-date version of the paper is avaliable here: https://arxiv.org/abs/2402.10726. Compared to KR-2024 version the experimental evaluation results in Table 2 have been updated.