The State of Natural Language Understanding: Past, Present and Future
The advent of deep learning has changed the landscape of natural language processing research. I will describe what has changed, what we gained and at what cost particularly in the context of language understanding tasks such as semantic parsing, question answering, and dialogue systems. The success of deep learning for problems such as Machine Translation is not directly transferable to language understanding problems, and it is critical to understand the limitations of these models before deploying them.
Siva Reddy is a postdoc at the Stanford NLP group working with Prof. Christopher D. Manning. His research focuses on finding fundamental representations of language which are useful for NLP applications, especially machine understanding. He published many papers on Semantic Parsing and Question Answering. His research has been partly funded by grants from the Google NLU Team and the Facebook AI Research.