Article Preview
TopIntroduction
With the increasing of science and technology, more and more technologies are applied in the field of education. A lot of education software that claimed to be able to help children study has appeared on the market. Does the education software can help students? What functions of good education software that teachers and students need should have?
Students in general need the software to solve difficult problems in their study or work with them to explore problems, rather than the software only providing some fixed contents of problem sets. Students who learn better wish the software could provide the environment of training their innovation abilities, instead of simply imparting knowledge. Teachers hope the software could also be powerful teaching assistants to help them answer the general questions raised by students, so that they have time and energy to do more creative work.
So far, many scholars and research institutions have made a big effort, and have developed many actual intelligent tutoring systems. In 1970s, the research of intelligent tutoring systems focused on the problem generation, particularly at the microcosmic level. The typical intelligent tutoring systems include Scholar (Carbonell, 1970), WHY (Stevens & Collins, 1977), SOPHE (Brown & Burton, 1975), WEST (Brown & Burton, 1976), BUGGY (Brown & Burton, 1978), GUIDON (Clancey, 1979), etc. In 1980s, the research was emphasized particularly on the learners and the learning process. The research of intelligent tutoring systems emphasized model tracing. The typical intelligent tutoring systems of this period included LISP Tutor (Anderson, Boyle, & Reiser, 1985), Geometry Tutor (Anderson, Boyle, & Yost, 1985), PROUST (Johnson, 1986), PIXIE (Sleeman, 1982), etc. In 1990s, the research focused on learner control. The typical intelligent tutoring systems of this period included Smithtown (Shute & Glaser, 1990), Bridge (Shute, 1991), Stat Lady (Shute & Gawlick-Grendell, 1996), SQL-Tutor (Mitrovic, 2003), Auto-Tutor (Graesser, 2004), etc. In the 21st century, the research emphasis turned to be the adaptive learning support, focusing on the results of cognitive science research (especially on learning mechanism research) and the design of learning process and learning environment. VC Prolog Tutor (Peylo, Thelen, Rollinger, & Gust, 2000), SCoT-DC (Clark, Fry, Ginzton, Peters, Pon-Barry, & Thomsen-Gray, 2001), Slide Tutor (Crowley & Medvedeva, 2006), AHP-Tutor (Ishizaka & Lust, 2004), MATHEMA (Alexandros, Maria, & Georgios, 2009), Mathtutor (Aleven, McLaren, & Sewall, 2009) are the typical intelligent tutoring systems of this period.
The successful development of these systems provided the tutoring of various fields, and promoted the progress of education technology. On one hand, an intelligent tutoring system serving as an intelligent system is the result of the comprehensive application of artificial intelligence technologies. It reflects the development level of artificial intelligent technologies. It is unimaginable that a tutoring system can tutor students without intelligence. On the other hand, an intelligent tutoring system acting as a tutoring system must be able to tutor. It must have good tutoring functions, so it can serve students better. Consequently, how to improve the intelligence and how to provide good tutoring functions both are the important research directions for establishing intelligent tutoring systems.