Highly interpretable linguistic knowledge bases optimization: Genetic tuning versus solis-wetts. Looking for a good interpretability-accuracy trade-off