ABSTRACT

Intelligent computer-assisted language learning (ICALL) presents challenges along several dimensions. First, there are many design and implementation issues related to the use of artificial intelligence and natural language processing (NLP) technologies in instructional systems. ICALL programs must be able to respond to input noisier than that used by other NLP programs. Second, there are pedagogical and associated cognitive issues. ICALL systems must be able to model the learning situation to some degree of granularity-at least to be able to offer an analysis of what might go wrong or what might constitute progress toward a goal. Finally, there are practical issues in the actual use and reuse of systems. The designers are not typically the instructors who use the system, but they must anticipate the changing needs of the curriculum and students. They must also embed sophisticated technology in a computing environment that is usable by those who do not understand the technology. Furthermore, because of the long development time of NLP software, ICALL systems must be built for reuse and reapplication, without having to repeat the initial effort.