Overview
The course covers the following learning outcomes and goals:
- Understanding the concept of a "word" in natural language processing.
- Learning about tokenization and its importance in text processing.
- Exploring morphology and morphological analysis in linguistic studies.
- Studying unsupervised subword segmentation techniques for language processing.
The course teaches the following individual skills or tools:
- Traditional word segmentation methods.
- Language typology and its relevance in linguistic analysis.
- Finite state automata for recognizing linguistic patterns.
- Two-level morphology and finite state transducers for morphological analysis.
- Spelling rules and their application in linguistic tasks.
The teaching method of the course involves a lecture format where the instructor covers theoretical concepts and practical applications in the field of advanced NLP.
The intended audience for this course includes students and professionals interested in delving deeper into natural language processing, linguistics, and computational language analysis.
Syllabus
Introduction
Whats a word
Word Segmentation
Traditional Word Segmentation
Morphology
Language typology
Eskimo
Fusional languages
Proto IndoEuropean
Type Tokens
Linguistic Side
Recognition vs Parsing
Spelling Changes
Finite State Automata
Adjectives
Linguistic Analysis
Meaning Analysis
TwoLevel Morphology
Finite State Transducer
Spelling Rules
E Compose
Taught by
Graham Neubig