Abstract Background The advent of the globalized economy has promoted the rapid development of cross-border educational exchanges, and the scale of the international student group has continued to expand. In an environment where there is a language barrier, daily communication becomes the main bottleneck for international students to integrate into the local society, which intensifies their social anxiety and sense of loneliness. For this reason, real-time speech recognition and translation technology have become potential intervention tools. To alleviate the social anxiety of international students, this study combines speech recognition real-time translation technology to intervene in the language barriers of international students in real social scenarios, and records and analyzes the changes in their anxiety levels. Methods The study selected 46 international students from different countries as subjects, whose ages ranged from 18 to 30. The study divided it into two groups, Group A and Group B, and implemented different intervention measures for the students in the two groups. Group A students participated in an eight-week daily conversation training using real-time translation devices based on speech recognition, while Group B received traditional language courses without using any translation AIDS. During the experiment, the social situations and anxiety levels of the two groups of students were recorded every week. The frequency of social interaction, the duration of conversations and the number of communication partners are used as assessment indicators of social ability. Teachers conduct blind evaluations of their social performance, with scores ranging from 1 to 5. The higher the score, the better the social performance. Anxiety levels are evaluated using the Hamilton Anxiety Scale (HAMA). Results The changes in social performance and anxiety levels of the two groups of students during the experiment are shown in Fig. 1. As can be seen from Fig. 1 (a), the social performance scores of students in group A steadily increased over time. The average score in the eighth week reached 4.3 points, which was significantly higher than that of students in group B at the same period, which was 2.8 points. The difference was statistically significant (p.05). Fig. 1 (b) shows that the HAMA score of group A decreased from 21.2 points at baseline to 12.6 points at week 8, while that of group B only dropped from 20.9 points to 17.3 points. The anxiety level in group A was relieved more significantly, and the difference between the groups was statistically significant (p.05). Discussion The research results show that the real-time translation technology of speech recognition can effectively alleviate the anxiety of international students in cross-language social interaction and significantly enhance their social participation ability. Current research only focuses on short-term intervention effects and has not yet examined the possible long-term impacts of technological dependence. Future research can extend the intervention period and include subjects from more cultural backgrounds to verify the universality of this technology in different language environments.
Chaoying Liu (Sun,) studied this question.