Abstract
This research adopts the logistic regression model fitted on richly annotated corpus data to investigate the constraining factors of “X BUBI Y · Z”“X MEIYOU Y · Z”and their effect sizes. Conditional inference tree and random forest models are also utilized to analyze the multiple interaction effects between constraints and their relative importance. The results show that the factors such as the positive and negative meanings of the conclusion, the semantic class of the compared item, the position of the comparison point, the type of the conclusion, the word length, register and the semantic class of the compared item significantly influence the constructional choice of “X BUBI Y · Z”and “X MEIYOU Y · Z”.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call