Font Size: a A A

Web-based And Data Mining Research And Development Of Intelligent Tutoring System

Posted on:2006-04-18Degree:MasterType:Thesis
Country:ChinaCandidate:X J QiaoFull Text:PDF
GTID:2207360152986798Subject:Curriculum and pedagogy
Abstract/Summary:PDF Full Text Request
The traditional CAI software lacks reasoning mechanism and the surport of student model, it can't tutor students pertinently and heuriscticly. Also, when developing the individual teaching method, it exists many deficiencies on carrying out the thoughts and methods of teaching students in accordance of their aptitude. Because of these shortages, CAI can't adapt to the need of current education. While, the popularizing and developing of AI technology and Internet provide a new opportunity for CAI's breakthrough. As the application of web in the ITS, so much useful instructing information will come forth at the same time. So we need aneffective technology to utilize these sources-datamining provides a new idea todeeply research in this area.In this paper, as analysing the elementary princeple, charater and relative models of ITS, it applies the data-mining technology into the system. Then, it presents a theoretic opinion on a web and data-mining based ITS. It provides the general design and implement scheme as well. First, we mainly research the classifation algorithm and the application of its scalable methods. We use the ID3 algorithm to classify the students into different difficult level successfully. And the method of tightly-coupled with DBMS makes the system be able to adapt large information mining. As a result, applying the datamining technology into ITS to improve the system's performance is a new challenge in this field. Second, a cognitive student modle is constructed, which uses the algorithm of fuzzy evaluation. Based on the fuzzy evaluation, we build the effective instruction regular which is more up to the teachers' language. Third, on the research of knowledge representation and reasoning methodology, we construct the production system based on certain reasoning. In the system, we use the natural reasoning as well as the reversal reasoning, so as to realize the magistral teaching procedure and individual learning process.This system utilizes dynamic HTML, JSP and newly Internet technology. Under the instruction of constructivism learning theory, we set up an intelligent, long-distance instruction platform. This system can teach many subjects. It also can realize individual teaching, provide learning suggestions and tutoring, reflect students' cognitive ability. So it is a universal long-distance teaching platform.The ITS in this paper has already debugged in the school network and can function high performance. It proves that the research of this intelligent, long-distance teaching system in this paper is of theoretic and applicational value.
Keywords/Search Tags:Intelligent Tutoring System, data mining, cognitive student model, individual instruction, production system
PDF Full Text Request
Related items