A Comparative Quality Assessment Of Human Interpretation And Machine Interpretation Of Expressions With Chinese Characteristics

Based on the theory of quantitative and qualitative evaluation,this study explores the interpretation quality of human interpretation and machine interpretation of expressions with Chinese characteristics from quantitative and qualitative aspects respectively.This research designs indicators for assessment based on Yang Chengshu's Quantitative and Qualitative Assessment Theory,dividing 7 indicators into three categories,namely the fidelity,delivery and language,with the focus on expressions with Chinese characteristics to explore the merits and demerits of machine interpretation and human interpretation.The research selects the excerpts from China 'Internet Plus'&Digital Economy Summit as the material,iFlytek 3.0 as the representative for machine interpretation and 8 interpreters as the representative for human interpretation.After the experiment,data is collected,scored and analyzed.The study draws the following conclusions:First,the overall performance of human interpretation is better than machine interpretation in all three categories,while machine interpretation scores slightly higher in information omission,over interpretation and redundancy.On the premise of correct voice recognition,machine interpretation can ensure a certain degree of fidelity,reaching relatively high score in policy-related expressions but still low in other four types of expressions with Chinese characteristics.Second,the machine has a certain degree of flexibility,and can use interpretation techniques such as de-verbalization.Interpreters use extensive strategies such as information integration and free translation to convey the cultural connotation of Chinese expression.In addition,the study found that speech recognition is still the biggest obstacle affecting the quality of machine interpretation,such as its inability to filter unnecessary information and the instability of online corpus.