The large-scale language model tags each token with its part of speech (POS), maps dependencies in a sentence, and structures and represents the sentence. The meaning of the dependency strings is ...