keyword overlap (percentage of aligned keywords), target token overlap (percentage of aligned target tokens), learner token overlap (percentage of aligned learner tokens), token match (percentage of token alignments that are token identical), lemma match, synonym match, type match, target triple overlap, learner triple overlap, target chunk overlap, learner chunk overlap, target bigram overlap, learner bigram overlap, target trigram overlap, learner trigram overlap, and variety of alignment (number of different token alignment types); the n-gram features are the only new features