East Carolina University
Department of Computer Science

CSCI 4140
Natural Language Processing
Standard Syllabus

Catalog entry

P: CSCI 2540; MATH 2228 or MATH 2283. Fundamental algorithms and computational models for core tasks in natural language processing. Word and sentence tokenization, parsing, information and meaning extraction, spelling correction, text summarization, question answering, and sentiment analysis.

Course summary

This course focuses on algorithms and tools for text-based processing of natural language. Topics on both syntax and semantics are covered. At present the course makes use of the Python programming language and the Natural Language Toolkit. In addition, students have the opportunity to engage in a research-based investigation in natural language processing. The current research topic is natural language dialog.

Course topics

• (Interspersed) Python programming for natural language processing applications.
• (Interspersed) Text processing fundamentals using the Natural Language Toolkit.
• Corpus processing.
• Part of speech tagging.
• Text classification.
• Information extraction fundamentals: chunking, named entity recognition, and relation extraction.
• Optional topics: parsing, natural language understanding, corpus management, dialog processing, and/or instructor research.

Student learning outcomes

• Design and program solutions to text-based processing problems using the Python programming language and the Natural Language toolkit.
• Describe language structure using regular expressions.
• Develop simple feature extractors for machine learning approaches for solving text-based natural language processing problems.
• Complete a research-oriented investigation into a suitably scoped problem that requires application of natural language processing techniques. Investigation should require both programming as well as a summary writeup that demonstrates an ability to self-critique the success of the investigation relative to the original project goal.

Textbook

Natural Language Processing with Python, Steven Bird, Ewan Klein, and Edward Loper (available online at http://www.nltk.org/book/); hard copy book for earlier version of NLTK was published in 2009 and available from O'Reilly.

The text interleaves text-based natural language processing fundamentals with an introduction to the Python programming language and the Natural Language Toolkit, an API for use with Python. It is very applied in nature and somewhat limited in its discussion of lingusitics. A supplementary text in linguistics might be desired depending upon instructor focus.

Grading

Grading will be based on midterm exam, final exam, homework assignments and quizzes, and a comprehensive project. The midterm exam will count 15%, the final exam 25%, the homework assignments and quizzes 40% and the project 20%.

Grade meanings