It does some fairly computationally cheap manipulation to return a "stem" word for most (English) words, e.g.: "biologically" matches "biologist."
He has a Web page for his better-established algorithm:
It has numerous links to implementations in various programming languages.
By his own comparison with LovinsStemming, he traded a shorter reach (i.e.: fewer characters) to his matching and transformation rules, for dramatically more (conceptual) steps in the algorithm.
He has a separate Web page for his newer algorithm:
It has links to implementations in far fewer languages.