[NLUlite]

By the way, the commercial version's parser scales almost linearly with the number of (independent) threads. The Wisdom.ask() method is also faster with the multithreaded version.