Many people who are deaf use American Sign Language (ASL) as their primary mode of communication. ASL is a natural language that is very distinct from English in its structure. Because people who are deaf have such limited access to the English language and are unable to acquire English using the same methods hearing people do, mastering written English poses a unique challenge to people who are deaf.
The ICICLE system (Interactive Computer Identification and Correction of Language Errors) is an intelligent tutoring system under development to tutor deaf users on grammatical principles of written English. ICICLE acts like an intelligent grammar checker and specialized tutor for deaf students who are learning written English. Given an essay that has been marked up with errors this project will generate tutorial instructions that will assist deaf students in understanding and correcting grammatical errors in the English language.
The overall goal of ICICLE is to create effective tutorials individually tailored to each user. The system must handle different errors committed by a user and provide a diversity of tutorials. While current word processing systems do provide feedback to users about grammatical mistakes, they usually only point out the location of the error. They do not offer any information about the type of error made, nor do they provide any instruction on correcting the error. Grammar checkers in word processors are not geared for the specific needs of deaf users; these grammar checkers give a generalization of errors and do not tailor to the specific type of errors usually made by those in the deaf community. The focus of my research is to develop a multilayer user model and multilevel text planner to generate effective tutorials that are individually adapted to the user's learning style and knowledge.