The objective of this paper is to develop an educational framework, using data mining technologies, with the help of dynamic web technologies that will be used by teachers to organize the course contents on the web according to existing infrastructure, experience, needs, reorganizing it later on if necessary, depending upon the performance of students. The approach to organizing the lecture contents is based on adaptive learning theory, incorporating a Problem Based Learning (PBL) strategy. Presently, course syllabus and handouts on web sites provided to the student are static in nature. Once distributed, these documents cannot be changed or modified, and lack depth. When course materials are placed on the web, students can select a topic in the course outline and look at the description of a topic, and required reading assignments. Instructors can easily change schedules in these on-line documents and inform the students via e-mail. Students can also submit assignments, projects and take-home exams electronically. A course home page comprises a syllabus, assignments, projects and exams, readings and references, class presentation charts and student handouts. Students on a course are mostly assessed based on questions such as Why, How, What, etc. In this way, a student can be graded and ranked, which in turn provides the feedback to the student for future improvement and challenges. Most such web sites are implemented on the theory of constructivism. Constructivists propose that the construction for new knowledge starts from one’s observations of events through previous experiences. Hence, learning is the integration of new knowledge and behaviors into a framework and subsequently recalling relevant events in the appropriate situation. This theory is also applied in our educational framework.