As the current education scenario has transformed itself to online mode, all learning and assessment activities including quizzes, report submissions, problem solving, peer assessment are done online. Identification of students' characteristics in terms of their academic performance and attitude is the need of the hour for personal tutoring. Also, collaborative learning, which forms an integral part of learning, has group formation as an influential activity for the success of learning. This work proposes an intelligent solution to group learners based on their outcomes and participation in various online assessment activities. This chapter considers the online assessment results of the learners and uses Kohonen self-organizing map neural network (SOM) to group the learners. The proposed method is experimented with a student set in the course “Digital Systems” (n=84). MATLAB is used for implementing SOM and the results obtained from simulations confirm the efficacy of the proposed network with 93.33% performance metric.
TopIntroduction
Smart education requires smart tutoring and peer learning. Facilitating good learning opportunities among the students is a major role of a teacher. Active learning and collaborative learning are always considered appreciable in improving student engagement. The entire world suffers from pandemic for years. The situation transforms the education sector from physical classes to virtual classes. Due to the sudden transformation of learning from face-to-face classes to online classes, the inclusion of these activities such as active learning and collaborative learning got disrupted a while. A wide range of online tools for implementing active learning and collaborative learning has been identified then and introduced in the teaching community. The pedagogical training has been given to the faculty to handle the new change. Lots and lots of challenges are faced by teaching community to adapt to this massive change. The online tools have increased the opportunities to provide better learning amidst the challenges faced during online classes. Online quiz tools and feedback tools are identified and used for formative and summative assessments. Conventionally, the group formation shall be done by the students based on their preference or by the instructor based on the students’ academic performance, attitude or random methods. There are lot of online activities and results which makes the team formation task of the instructor to be tedious. With a rich set of online activities available now, the group formation can be done without much effort from the instructor with intelligent automatic techniques.
Collaborative learning refers to doing certain tasks as a group and the same needs intellectual efforts among the group of students jointly (Chu & Kennedy, 2011). The major activities of Collaborative learning include group formation, conduct of collaborative tasks and assessment of the tasks. Group formation is a crucial factor for the success of any collaborative activity. There are many methods of group formation including random formation, any specific order of number or name, specific methods based on academic performance and voluntary. Groups may be composed as heterogeneous or homogeneous with respect to few parameters such as students’ cognitive ability, culture, attitude and gender. Grouping the students according to their cognitive ability shall help the teachers to understand the potential of the students for further measures like assigning challenging problems, counselling and peer assistance. The group composition may have its impact on the performance and comfort level of the individual and the whole team as well. It becomes a tough job for the teacher to identify the characteristics of the students in a large classroom for appropriate grouping of students. Especially with lot of online activities, the identification of cognitive ability tends to be tougher. There is a need of looking into suitable strategies for group formation in a learning environment which comprises of many online activities. As every online activity of a learner is now available in the online environments, intelligent techniques shall be used for group formation with the learner data. Clusters of students can be identified with their performance and collaborative groups shall be framed based on the intended learning outcomes.
Clustering is an unsupervised artificial intelligence technique that groups any data without labels. Clustering helps to group data based on the similarity between them. The clusters so formed have many applications which are dealt in literatures vastly. A SOM is an artificial neural network(ANN) and it is basically an unsupervised technique used in machine learning and uses competitive learning for clustering data. Kohonen, T. conducted many simulation experiments and explained the self-organized map clearly and also suggested that architecture for artificial neural networks along with practical applications. It is also stated that these SOMs have the property of creating internal representations of various features of input data and their concepts and are spatially organized, thereby discovering sematic relationships effectively.
Normally, Student academic counselling is done based on performance of the students. It requires interest of the students to provide additional learning materials, one-to-one discussions and learning opportunities for the students.
This research work uses this clustering technique with appropriate learner data and attempts to use the results for the group formation in collaborative activities and special counselling of students.