This document discusses progress in using natural language processing (NLP) techniques to extract information from unstructured materials science text data. It notes that text data is not readily interpretable by machines and manual extraction is expensive. It describes using machine learning approaches to automatically extract synthesis parameters for materials from body text. It outlines a continuum of NLP approaches from rules to deep learning and challenges in applying NLP to materials domains due to issues like ambiguity and linked recipes. Examples of applying NLP to materials include generating plausible negative synthesis examples and exploring conditions to stabilize desired materials.