Factorial Hidden Markov Models for Learning Representations of Natural Language | Read Paper on Bytez