A Comparative Study of Linguistic Style between AI Chatbot and Human in Chinese Civil Servant Interview

Structured interviews are a crucial method for talent selection, tailored to the specific requirements of a job and the competencies of candidates. These interviews are conducted using a set protocol, with standardized questions and evaluations designed to determine a candidate's suitability. In China, civil service exams commonly employ this method to assess candidates' behavioral strategies in public service roles. The performance of AI in such tasks could significantly indicate its potential to assume responsibilities traditionally held by civil servants. In this study, we utilized a lexical analysis approach based on corpus research to compare the vocabulary characteristics used by an AI chatbot (Large Language Model, GLM-4) and human experts in structured interview training. This comparison focused on their responses to simulated questions from civil service examinations. Our preliminary findings highlight distinct linguistic differences in word usage, thematic keywords, and readability between the human experts and the GLM-4 bot. This research not only examines the role of AI in enhancing interview training skills but also explores the wider implications of integrating AI into public administration. The insights gained could inform future strategies on the incorporation of AI technologies in public sector operations, potentially transforming aspects of state governance.