This document discusses named entity recognition (NER) as a case study for natural language processing (NLP) tasks. It outlines the general methodology for solving NLP problems, including formalizing insights, studying algorithms mathematically, developing and implementing algorithms, and testing on real data. For NER, this involves collecting annotated training data, designing informative features from the text, and using these features to train a model like conditional random fields to label named entities. Evaluation, improving features and models, exploiting unlabeled data, and addressing domain transfer are also discussed.