This book is great. Really mind warping at first read. Fernando Pereira has had an incredible influence across NLP for his whole career. Here is an offhand lis

This book is great. Really mind warping at first read. Fernando Pereira has had an incredible influence across NLP for his whole career. Here is an offhand list of papers to check out. * Conditional random fields: Probabilistic models for segmenting and labeling sequence data (2001) - Central paper of structured supervised learning in the 2000s era * Weighted finite-state transducers in speech recognition (2002) - This work and OpenFST are so clean * Non-projective dependency parsing using spanning tree algorithms (2005) - Influential work connecting graph algorithms to syntax. Less relevant now, but still such a nice paper. * Distributional clustering of English words (1994) - Proto word embeddings. * The Unreasonable Effectiveness of Data (2009) - More high-level, but certainly explains the last 15 years