A major challenge in computer science education is to improve both instructional productivity and learning quality for a large and diverse population of students under real-world constraints such as limited financial resources and insufficient qualified instructors. The literature in education suggests that students who are actively engaged in the learning process will be more likely to achieve success [12]. The approach of active learning emphasizes on engaging students in the learning process [6,22,23], where learning activities involve some kind of experience or some kind of dialogue. The two main kinds of dialogue are dialogue with self (think reflectively) and dialogue with others. The two main kinds of experience are observing and doing. There are several ways to incorporate more active learning into teaching:
(1) Expand learning experiences. The most traditional teaching consists of little more than having students read a text and listen to a lecture, providing a very limited form of dialogue with others. Examples of more dynamic forms are creating small groups of students and having them make a decision or answer a focused question periodically, or finding ways for students to engage in authentic dialogue with people other than fellow classmates who know something about the subject (on the Web, by email, or live).
(2) Take advantage of the ``power of interaction.'' There are four modes of learning when the two kinds of dialogue are coupled with the two kinds of experience. Each of the four modes has its own value, and just using more of them should add variety and thereby be more interesting for the learners. However, when properly connected, the various learning activities can have an impact that is more than additive or cumulative; they can be interactive and multiply the educational impact. For example, if students write their own thoughts on a topic (dialogue with self) before they engage in small group discussion (dialogue with others), the group discussion should be richer and more engaging. If they can do both of these and then observe the phenomena or action (observation), the observation should be richer and again more engaging. Then, if this is followed by having the students engage in the action itself (doing), they will have a better sense of what they need to do and what they need to learn during doing. Finally if, after doing, the learners process this experience by writing about it (dialogue with self) and/or discussing it with others (dialogue with others), this will add further insight. Such a sequence of learning activities will give the teacher and learners the advantage of the Power of Interaction. Alternatively, advocates of problem-based learning would suggest that a teacher start with doing by posing a real problem for students to work on, and then having students consult with each other (dialogue with others) on how best to proceed in order to find a solution to the problem. The learners will likely use a variety of learning options, including observing and dialogue with self.
(3) Create a dialectic between experience and dialogue. New experiences (whether of doing or observing) have the potential to give learners a new perspective on what is true (beliefs) and/or what is good (values) in the world. Dialogue (whether with self or with others) has the potential to help learners construct the many possible meanings of experience and the insights that come from them. People learn faster when new concepts are useful in their present as well as future lives. The roles of an educator is to assess the audience's interest, current skills, and aims. This information then guides the structuring of a learning atmosphere and selection of methods most satisfying and effective for the learners.
The growing emphasis on student-centered active learning has yielded a veritable revolution in educational theory and practice. A number of current theories of learning and pedagogy revolve around constructivism, which emphasizes the student's knowledge construction process. In student-centered education, students should be engaged in active exploration, be intrinsically motivated, and develop an understanding of a domain through challenging and enjoyable problem-solving activities. Systems and procedures that support active learning have been intensively investigated and developed [7,8,9,11,18].
In this paper, we present an intelligent agent assisted system to support student-centered, self-paced, and highly interactive learning, a first step in building an effective active learning environment. The system provides a rich set of online contents and around the clock information access, maximizes the interactivity between the intelligent learning system and the students, and customizes the learning process to the needs of individual students. In the system, students' learning-related profiles, such as learning styles and background knowledge, are used for selecting, organizing, and presenting the learning materials to individual students and in supporting active learning. It supports personalized and more pleasant interaction between the users and the learning systems, enables adaptive delivery of IT education content, facilitates automatic evaluation of learning outcomes, and provides easy-to-use authoring tools. The system also incorporates a new approach to course content organization and delivery, which is developed based on smart instructional components, called lecturelets. Lecturelets are designed for customized interactive presentation of subjects. They are self-contained, autonomous, and can be easily integrated into a wide range of courses. The intelligent distributed environment for active learning (IDEAL) is implemented using the prevalent Internet, Web, digital library, and multi-agent technologies. IDEAL adopts an open system architecture supporting open standards in information technology and can scale to large-scale, distributed operations.