Overview
Explore advanced natural language processing concepts in this lecture on information extraction and knowledge-based question answering. Delve into the fundamentals of knowledge graphs and ontologies, learn techniques for relation extraction from embeddings, and discover methods for learning embeddings from relations. Investigate how to probe language models for knowledge. Gain insights into structured knowledge bases, distance supervision noise, and knowledge injection. Examine open information extraction, neural models for relations, and practical examples of relation extraction. Part of CMU's Advanced NLP course, this comprehensive lecture provides a deep dive into cutting-edge NLP techniques for working with knowledge and information.
Syllabus
Intro
Types of knowledge bases
Psych
Learning representations from knowledge bases
TransE method
Relation extraction
Relation classification
Distance supervision noise
Structured knowledge bases
Retraining embeddings
Knowledge injection
Knowledgebased QA
Open Information Extraction
Neural Models
Relations
Relation example
Taught by
Graham Neubig