Neural network based language models ease the sparsity dilemma Incidentally they encode inputs. Phrase embedding levels make an arbitrary sized vector of each and every term that comes with semantic interactions as well. These continuous vectors produce the Significantly necessary granularity in