Speech and Natural Language Processing for Educational Applications
(CS 3710/ISSP 3565: Advanced Topics in Artificial Intelligence)

Instructor:  Diane Litman
Time: MW 1-2:15

	"NLP-based applications in educational environments continue to develop
	at a fast pace. Initial work began as early as the 1960Â^Òs on
	evaluating written essay-length and short-answer assessments,
	text-based intelligent tutoring systems (ITS), and proofreading
	tools. These fields continue to progress using innovative NLP
	techniques - statistical, rule-based, or some combination of the
	two. More recently, new technologies have made it possible to include
	speech in both assessment and ITS. Using a somewhat different
	approach, NLP techniques are being used to generate assessments and
	tools for curriculum development. As a community we are not only
	improving existing capabilities, but identifying and generating
	innovative and creative ways to use NLP in applications for writing,
	reading, speaking, and critical thinking."
		[from Workshop Description, 4th Workshop on Innovative Use of
		 NLP for Building Educational Applications, 2009]

This course will both introduce the foundational areas of Speech and
Natural Language Processing, and of Artificial Intelligence in
Education, then present recent work addressing their integration.
The course will be a seminar-style course.  Students will lead one or
more class discussions, participate in the other discussions, and
complete a course project.

Topics will be drawn from the following areas:

- Automated Scoring/Evaluation for oral and written student responses
   o Content analysis, grammatical error detection, and discourse analysis
   o Machine translation for assessment, instruction, and curriculum development

- Intelligent Tutoring Systems
   o Tutorial dialogue systems
   o Multi-modal communication

- Learner Meta-Cognition
   o Systems that detect and adapt to learners' meta-cognitive or emotional states

- Empirical Methods
   o  Annotation standards
   o  Data mining of corpora

- Classroom Tools
   o  Automatic identification or generation of materials at a particular readability or grade level
   o  Automatic generation of test questions, e.g., multiple choice or short answer
   o  Processing of and access to online lecture materials (slides, audio)
   o  Language-based educational games

- Deployed systems (both commercial products, and research prototypes in classrooms)

Prerequisites: Natural Language Processing, OR Artificial Intelligence, OR consent of the instructor.

**NOTE: Because 3710 is a "Topics" Course, it is different every
 semester. Students can thus take CS 3710/ISSP 3565 multiple times.**

If you have questions about the course, please feel free to contact
me.  If you plan to take the course, I would appreciate an email
saying so (for planning purposes).

Diane Litman
litman@cs.pitt.edu
http://www.cs.pitt.edu/~litman