As an important means of simultaneous interpretation, prediction has been widely discussed by scholars. However, the existing research is either too broad or too narrow. In order to explore this problem, this paper attempts to combine the simultaneous prediction research with the compound case auxiliary word research, taking the compound case auxiliary words ‘demotte’, ‘niyotte’, ‘wotsujite’ as the object of research, by analyzing their Semantic features and grammatical functions. This paper draws the correlation and difference between the three, and provide a certain range for the interpreter to predict the subsequent information of compound case auxiliary words based on this kind of language schema.