Rakuten Technology Conference 2017

Understanding human language is considered to be a true hallmark of machine intelligence. In the last decade, for the first time in AI history, tools for natural language processing (NLP) have transitioned from research labs to industrial applications, and are used by millions of people on a daily basis. How did deep learning impact this transition? In this talk, I will aim to answer this question. I will start by providing a short tutorial on deep learning and its uses in NLP. Next, I will discuss our recent work on addressing the limitations of this technology, including "black-box" predictions and reliance on large amounts of training data.  This will cover our work on interpretability and transfer learning.