Assessment Model Construction of International Students’ Intercultural Adaptation Based on Dual Attention Graph Convolutional Network

Abstract

Intercultural adaptability predicated on sojourners' sociocultural adaptation to their new surroundings. Globalization and the development of communication technology have led to a sharp rise in cross-border migration. However, there are significant obstacles to intercultural adaption, including student variations in traditional English cross-cultural instruction, developing students' cross-cultural communication skills, and improving the caliber of instruction. This manuscript proposes a Dual Attention Graph Convolutional Network (DAGCN) optimized with the Cheetah Optimization Algorithm (COA) for assessment model construction of international students’ intercultural adaptation. The data are collected from Pakistani students at Huazhong University of Science and Technology, P.R. China.  Afterward, the data’s are fed to pre-processing. In pre-processing segment; it removes the noise and enhances the input data’s utilizing Adaptive Robust Cubature Kalman Filtering (ARCKF). The outcome from the pre-processing data is transferred to the DAGCN. The language proficiency, cultural awareness, communication skill, and emotional resilience are successfully classified by using DAGCN. The COA is used to optimize the weight parameter of DAGCN. The proposed DAGCN-COA is utilized within the MATLAB/Simulink software environment. Performance metrics including accuracy, precision, sensitivity, F1-score, calculation time, and recall were examined in order to determine the proposed strategy. The proposed DAGCN-COA technique yields higher precision (14.89%, 16.89%, and 18.23%), higher sensitivity (16.34%, 12.23%, and 18.54%), shorter computation times (82.37%, 94.47%, and 87.76%), and higher accuracy (16.65%, 18.85%, and 17.89%).The proposed ISIA-DAGCN-COA method is compared with the existing methods such as ISIA-ANN, ISIA-CNN, and ISIA-RBF, respectively.