S. Takase, J. Suzuki, and M. Nagata, “Character n-Gram Embeddings to Improve RNN Language Models”, AAAI, vol. 33, no. 01, pp. 5074-5082, Jul. 2019.