Learning Interpretable Models in the Property Specification Language