INTELLIGENT AGENT MODEL OF LEARNING GROUP FORMATION IN COMPUTER-SOPPORTED COLLABORATIVE LEARNING (CSCL)

• Main Author: Laksono Putro, Budi
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/62761
• Institution: Institut Teknologi Bandung

ABSTRACT INTELLIGENT AGENT MODEL OF LEARNING GROUP FORMATION IN COMPUTER-SOPPORTED COLLABORATIVE LEARNING (CSCL) by BUDILAKSONOPUTRO NIM:33216001 (Doctoral Program in Electrical Engineering and Informatics) Collaborative learning based on Computer-Supported Collaborative Learning (CSCL) is the core of 21st century learning to build 21st century skills. CSCL is an online learning environment for collaborative learning. CSCL aims to support learning services and social interaction among students of a group or community. There are three types of Intelligent Agent services for collaborative learning in CSCL: Group Formation, Domain Specific Support, and Peer Interaction Support. This study focuses on intelligent agents for group training services. This agent was chosen because group training is early service and has a significant influence on collaborative learning in CSCL. This research problem formulation focuses on developing an intelligent agent model for group formation in the CSCL. The discussion on this topic is based on the scientific domain of service computing based on the A CM 2012 and IEEE 2017 curriculum. An intelligent agent model is a reference for developing an intelligent agent for group formation in CSCL. An intelligent agent model is closely related to the application of computing technology to the learning group formation service functionality in CSCL. There are currently four types of models based on the success factors of group training: User, Group, Domain and Activity (UGDA). The development of an optimal intelligent agent requires the integration of the application of computing technology in group formation services based on the dominant factor of the four success factors (UGDA). The existing model is still based on one type of suc.:cess factor, lac.:king an explanation of how to selec.:t the most dominant success factor and information technology. A method for implementing and evaluating the proposed model for developing intelligent agents was also necessary. The objective of this research is to develop an intelligent agent model. That is a model that is able to integrate group formation services with IT services based on the most dominant success factors (from user, group, domain, and activity), in order to achieve an effective and efficient group formation computing process. This research applies design research methodology (DRM) and strategic development of an intelligent agent model based on the field of service computing. Developing a model that is constructed from the dynamic modeling process of the system. The development stages include: Determine objectives of the model, Identify the components and phases of the model, Engineering & optimization of the group formation servic.:e system in CSCL, and Development of a proposed model. Implementing an intelligent agent based on the proposed model requires a method for developing an intelligent agent. To evaluate the pe,formance of the intelligent agent computation process, a test framework is required. This research produces four outcomes, namely: 1. intelligent agent model, 2. Intelligent agent testing framework, 3. Intelligent agent development method, and 4. Intelligent agent for group formation in CSCL. The Intelligent Agents model is the primary outcome of this study. An intelligent agent model is a reference for intelligent agent development. The development method is a strategy for implementing the proposed model in an intelligent agent for group formation. The pe,formance testing framework is an evaluation method for the computational process of intelligent agents for group formation. An intelligent agent is empirical evidence of the implementation of the intelligent agent model work system, and is built from the development method. Intelligent agents are developed based on case studies of problems. The model successfully integrates information technology services with group training services that are based on the dominant group training/actors. This model succeeded in building intelligent agents with effective and efficient group formation computational processes based on four success factors, namely: compatibility between group members, optimization of group composition, collaboration petformance (CP), and collaborative learning outcomes in CSCL.