Benchmark and Hard Datasets

Description

it is used to evaluate PSM-DFA and RPNI passive DFA learning algorithms

Steps to reproduce

Positive and negative sample strings were generated by rejection sampling against a target membership rule for each dataset (e.g., "contains substring ab", "ends with ab") until the required count was reached. Each .txt file lists the positive and negative samples for one benchmark dataset used to construct a Prefix Tree Acceptor and evaluate PSM-DFA and RPNI state-merging algorithms.

Categories

Licence