Learning through Negotiation

• Main Authors: Chung-Hsien Lan, 藍中賢
• Other Authors: K. Robert Lai
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/68978617812750385093

博士 === 元智大學 === 資訊工程學系 === 97 === This dissertation presents a theory of learning through agent negotiation. In a computer-supported learning environment, interaction plays a critical role for developing instructional strategies and promoting learning effectiveness. Especially, learners, peers and instructors often bring different ideas and perspectives to the learning process. Thus, how to resolve these differences and without creating unnecessary discontentment could be of paramount important for the effectiveness of learning. To that end, a learning environment is viewed as a multi-agent system, in which agents can support effective interactions between each other to create a social community with rich social behaviors. Then, conflicts and discontentment can be reconciled through agent negotiation mechanism. Accordingly, a fuzzy constraint-directed agent negotiation is proposed and then implemented in supporting three learning scenarios, including peer assessment, game-based problem-based learning and adaptive learning, to illustrate how to incorporate the notion of agent negotiation into a learning environment and, more importantly, to improve the effectiveness of learning. In peer assessment, the proposed model facilitates learners in grading peer work with fuzzy constraints and, to reduce the bias, assessments can be adjusted according to personal characteristics. Through iterative negotiation, learners can reach an agreement with peers for the assessment. Experimental results suggest that learners are more willing to accept the assessment and able to acquire richer feedback to reflect upon their work seriously and enhance learning effectiveness. Moreover, instructors can understand learners’ performances to appropriately adjust instructional strategies by observing learners’ participation and assessment result. In problem-based learning, a game-based collaborative business simulator is developed to provide learners an environment for learning strategic planning and decision making in a beer game. In this agent-based business simulator, instructors are responsible to describe the problems of scenarios, and learners can play multi-roles to mimic more closely to the real-world business scenario and negotiate with either peers or virtual agents to solve the problems. During the process of negotiation, learners can acquire necessary information while attempting to develop the solutions. Moreover, the business simulator can also guide the learners to explore the decision alternatives. Through the process of iterative negotiation, learners can reflect upon self-cognition and also acquire necessary knowledge to achieve the learning goals. In adaptive learning, breaking from traditional instructional approaches, the proposed framework considers learners’ learning goals and personal characteristics to provide adaptive learning materials and allow learners to negotiate mutual cognitive differences in knowledge levels and learning concepts with virtual tutor to decide next learning materials. Experimental results show that learners can represent self-cognition by using fuzzy constraints and discover better learning sequences through negotiation with virtual tutor. In summary, all these learning scenarios reveal that the notion of negotiation can be a valid approach for promoting learning effectiveness.