Enhancing the Intelligence of the Adaptive Learning Software through an AI assisted Data Analytics on Students Learning Attributes with Unequal Weight

Abstract

Along with the growing popularity of the adaptive learning platform software in STEM education programs, further enhancing the ability of the adaptive learning platform to support learning effectiveness has become a priority. A common goal of all the adaptive learning platform software is to identify gaps in a student's knowledge, and provide relevant learning materials based on the result of the assessment of the student's exiting knowledge towards to the targeted subject, to increase the learning effectiveness. However, the student's existing knowledge is only one learning attribute, and it cannot reflect all the aspects that have an impact to the student's learning effectiveness. A natural solution to overcome this weakness is to make the adaptive learning decision making algorithm capable to process the dataset of multiple learning attributes of the student efficiently. This paper presents a machine learning algorithm to efficiently process the dataset of a student's multiple learning attributes. The main enabling foundation for this new algorithm is a data structure called “student learning attributes index” which represents every learning attribute as a tuple of three elements: the “learning-attribute-If)”, the “weight” of the learning attribute among all the learning attributes, and the “efficiency” of the contribution to the student's learning effectiveness made by the learning attribute. This study has applied unequal weight to each of the learning attributes, more accurately reflecting that different learning attributes will have different impacts on a student's learning effectiveness. This new algorithm enables various learning support applications to become more practical and accurate in supporting student learning.