Course Quality Evaluation of Higher Integrated Distance Education Based on Group Decision-making Algorithm

Abstract

Higher integrated distance education courses involve an educational process where students with and without disabilities learn together. The course quality evaluation methods of these courses require indicators that fully consider the conditions of different students. Due to the significant differences among students, the methods primarily based on the superposition of personal evaluations cannot be applied to these two special student groups. It leads to a lack of consensus and balance in the course quality evaluation and lower quality and accuracy in decision-making. To solve this problem, this paper proposes a group decision-making algorithm for the quality evaluation of higher integrated distance education courses. In a large, differentiated student group, the decision-making results are formed by collecting, integrating, and weighing group members' opinions, knowledge, experience, and other information. Firstly, the algorithm proposed uses the K-means algorithm to cluster and collect the learning behavior data of students into predefined K clusters based on feature similarity. Secondly, this algorithm constructs course quality evaluation indicators from multiple perspectives under the support of multiple evaluation principles. Finally, this algorithm brings together the personal preferences of each decision-maker into a collective preference, allowing these decision-makers to rank the decision-making algorithm by their preference in this paper. Students are regarded as decision-makers, the higher integrated education courses are regarded as decision-making options, the learning behaviors and preferences of students are regarded as attribute decision variables. Thus, the quality evaluation of higher integrated distance education courses can be regarded as a group decision-making process according to the problem description above. This process makes decisions through iterative optimization. Experimental results show that this algorithm performs excellently in the metrics consistency ratio and comprehensive score and has practical application value.