Lecture
This lecture covers various algorithms and techniques for information extraction, including the Viterbi algorithm, named entities recognition, part-of-speech tags, and word n-grams. It explores the use of hand-written patterns, supervised machine learning, bootstrapping, and distant supervision. The instructor discusses the challenges of semantic drift and the use of knowledge bases for distant supervision. Matrix factorization and Bayesian personalized ranking are introduced as methods for relation extraction. The lecture also delves into linking text to knowledge bases, creating matrix representations, and utilizing relation embeddings.