Chapter 13 – Natural language understanding

Contents

13.1  Overview
13.2  What Is Natural Language Understanding?
13.3  Why Do We Need Natural Language Understanding?
13.4  Why Is Natural Language Understanding Difficult?
13.5  An Early Attempt at Natural Language Understanding: SHRDLU
13.6  How Does Natural Language Understanding Work?
13.7  Syntactic Analysis
13.7.1  Grammars
13.7.2  An Example: Generating a Grammar Fragment
13.7.3  Transition Networks
13.7.4  Context-sensitive Grammars
13.7.5  Feature Sets
13.7.6  Augmented Transition Networks
13.7.7  Taggers
13.8  Semantic Analysis
13.8.1  Semantic Grammars
13.8.1.1  An Example: A Database Query Interpreter Revisited
13.8.2  Case Grammars
13.9  Pragmatic Analysis
13.9.1  Speech Acts
13.10  Grammar-free Approaches
13.10.1  Template Matching
13.10.2  Keyword Matching
13.10.3  Predictive Methods
13.10.4  Statistical Methods
13.11  Summary
13.12  Solution to SHRDLU Problem

Glossary items referenced in this chapter

ambiguity!lexical, ambiguity!pragmatic, ambiguity!referential, ambiguity!semantic, ambiguity!syntactic, augmented transition network, big data, bottom-up reasoning, case grammar, chatbot, constraints, context-free grammar, context-sensitive grammar, database, database query, deep neural network, definite clause grammar, disambiguation, document retrieval, ELIZA, feature sets, finite state machine, formal grammar, frame problem, GPT-3, grammar, grammar fragment, grammar rules, grammar!rules, grammar!semantic, grammar!syntactic, grammar-free approaches, heuristic evaluation function, hidden Markov model, home automation, Human Computer Interaction, keyword matching, large language model, latent semantic analysis, latent space, lexical processing, lexicon, machine learning, n-gram, natural language database query, natural language processing, natural language processing!information management, natural language processing!intent, natural language processing!tagger, natural language processing!tagset, natural language understanding, neural network, non-terminal symbols, parse tree, parser, part-of-speech tagger, pattern matching, Pereira, Fernando, POS tagger, pragmatic, pragmatic ambiguity, pragmatic analysis, principal components analysis, production system, Prolog, query interpreter, recommender systems, restricted language, robotics, script, semantic analysis, semantic grammar, sentence parsing, sentence parsing methods, sentence parsing!bottom up, sentence parsing!systems, sentence parsing!top down, sentence-level processing, SHRDLU, speech acts, statistical methods, statistical techniques, syntactic analysis, syntactic grammar, template matching, terminal symbols, transition network, trigger, user interface, web search, Winograd, Terry, word vector, word2vec