Application of improved neighbor propagation algorithm in international communication and cooperation to promote internationalization of higher education

Abstract

AbstractMost educational institutions use web 3.0 and blended learning technologies to improve student learning. Because not all students are intrinsically motivated to learn, the instructor in an e‐learning environment has significant challenges, lowering student performance. As a result, by merging the global dimension, intercultural, and international education functions, the Internationalization Standard for Higher Education (ISHE) is introduced. The Internationalization standard higher education (IoHE) process minimizes the difficulties in education resources because of the effective utilization of Information and Communication Technology (ICT). The ICT process improves innovative learning and boosts the e‐learning platform. To achieve the effectiveness of web 3.0 with a blended learning process, an Improved Neighbor Propagation Algorithm (INPA) with a concept‐mind mapping tool is introduced in this study. Student data are classified based on machine learning in an online environment. Classification is the process of predicting the class of given student data points. The relationship between each data point is examined here, and a tree structure is formed according to the concept. Mind Mapping is a suitable method that helps learn, increases information recording, shows how diverse facts and ideas are related, and improves problem‐solving. The concept and mind map are linked with different ideas, and the respective Collaborative Online International Learning (COIL) program is conducted to enhance the student's performance. The discussed blended learning‐based internationalization higher education process is evaluated with the case studies. The numerical results of the proposed INPA algorithm improve the student performance by 98.9%, satisfaction ratio by 97.5%, engagement level by 95.6%, accuracy ratio by 98.1%, the precision ratio by 96.2%, and recall ratio by 95.4% compared with other existing methods.