A Deep Compositional Framework for Human-like Language Acquisition in Virtual Environment | Read Paper on Bytez