Cloud-Based Adaptive Systems

The term cloud computing describes a large number of computers on the Internet sharing computing resources that can be dynamically allocated to a computational task. The network of computers sharing their resources then becomes a vast resource that serves thousands of users. The virtual sharing of computer power is a concept that has its origin in traditional client-server architecture. In the early 2000s, Amazon played a key role in creating a scale deployment of an e-commerce site serving millions of users. Cloud computing has grown in a very short time to become the standard as a scalable architecture for large-scale computer applications. Amazon Web Services (AWS) has set standards now followed by numerous cloud service providers.

Adaptive learning has emerged as a standard promising approach for online education. Adaptive learning, which refers to meeting the needs of the individual learner, requires dynamic rendering of the content and hence needs much higher computational power than static Web applications. Cloud computing is a perfect match for computer-intensive applications, such as adaptive learning. Massive open online courses (MOOCs) are another case where cloud computing is appropriate as MOOCs may have thousands of learners connecting to a course; if the MOOC is to adapt to individual learners, then cloud-based resources are one way to implement an adaptive MOOC. Adaptive learning systems that customize content to the needs of individual learners have the potential to provide a personalized learning experience to the large number of learners in a MOOC. Cloud-based adaptive learning systems can deliver individual customization of content delivery on a massive scale, allowing each learner to reach the necessary competency.

Adaptive learning requires the development of differentiated learning strategies and intelligent feedback for building learner competency. The system elaborated in this entry presents content using five distinct learning strategies. Differentiated learning strategies and dynamic rendering of the content are provided to each learner, building competency for each individual.

[Page 104]The pedagogical framework that supports dynamic rendering requires appropriate cloud-based computational and Web-services architecture that can accommodate delivery of webpages within an acceptable response time. Some MOOCs deployed for worldwide audiences are expected to serve 50,000 to 150,000 users. The problem becomes intractable for conventionally hosted adaptive learning systems. A scalable cloud architecture that marries AWS with an adaptive mobile learning platform (AMOL) is one promising approach for an adaptive cloud-based educational system.

The origin of the adaptive learning systems can be traced to the early research conducted in the area of artificial intelligence, intelligent tutors, expert systems, and adaptive controls. These systems are known to change the output signal based on the input parameters and hence are termed adaptive systems. In the context of education, adaptive systems are able to change the content and its presentation based on the learner’s preferences.

Educational adaptive systems need four essential elements to qualify as adaptive learning systems. First, they must have a pedagogical framework to offer differentiated instruction. Second, the content must be presented using various learning strategies. Third, there has to be a mechanism to conduct diagnostic formative assessment to identify the success of learner in using each learning strategy. Fourth, there must be the ability to dynamically generate a revision trajectory for the learner to improve his or her performance.

...