By Kyle McCardle
Issues with Natural Language Basic Components Syntax The Earley Parser Transition Network Parsers Augmented Transition Networks Stochastic Tools Applications
A large amount of human knowledge is assumed Language is pattern based ◦ Syllables -> words -> phrases -> sentences Language occurs in a complex environment
Language lexicon ◦ Complete catalog of recognized words Parser and set of grammar rules ◦ Pulls apart sentences for internal representation Semantic theory ◦ Tool for deriving meaning from internal representation
Terminals Nonterminals Top-down derivation Parse tree
Dynamic Programming ◦ Memoization Dotted grammar rules ◦ Predict, Scanned, Completed Sentence → ∙ Noun Verb Noun → ∙ john Noun → john ∙ Sentence → Noun ∙ Verb
Addresses issue of semantic relationships Terminal and nonterminal arcs Context-sensitive grammars
Extend transition networks by utilizing stored procedures associated with arcs Procedures assign grammatical qualities to a given word (part of speech, root, number) Case frames provide semantic context
View language as a random process ◦ p(t 1, …, t n | w 1, …, w n ) Markov model approach Decision tree approach Probabilistic approaches
Reading and comprehension Translators Relational database front end Information extraction from the web
Earley parser. (n.d.). Retrieved from ki100k/docs/Earley_parser.html Luger, G. (2009). Artificial intelligence: Structures and strategies for complex problem solving. (6th ed.). Boston: Pearson Education, Inc. Stochastic semantic analysis. (n.d.). Retrieved from ntic_analysis Zhaoyin, Z. (2009). Rule-based natural language understanding based on fuzzy evaluation of teaching quality International Forum on Computer Science-Technology and Applications