Abstract: To generate suitable learning sequence according to the student’s individual characteristic is a key to build an intelligent tutoring system,and is also an exhibition of intelligence for the system．The difficulty is that both the relationship between the knowledge items should be considered,and the personalized learning characteristics of students should be adapted．In this paper,a learning sequence generation algorithm,which is under the guidance of relationships between knowledge,was proposed．The computation of learning level and the modeling of student personality were described．Based on these,the learning sequence can be optimized.And the detailed adjust and optimization methods were discussed in detail．Learning window was proposed to describe the learning contents and to adapt to the students with different learning abilities,and the learning model was used to distinguish the organization of contents and the corresponding learning styles．The application of the prototype indicates the effectiveness of the proposal,and the prototype system is adaptive to meet students’ individual learning needs.

Key words: Learning sequence,Adaptability,Personality,Intelligent tutoring system