My interest is twice:
i. I want to catch possible errors when a sentence is parsed and the parsing fails in the middle of the process. In other words, if a partial tree, or trees, is generated I would like to get knowledge about these "partial/s tree/s".
ii. I want to include some "dynamic" vocabulary. Sometimes the parsing process fails because of misspelling or because the sentence includes a unknown lexical term. I could fix such a situation programmatically, sometimes.