An Automatic Scoring System for Chinese Verbal Creativity Based on Pre-Trained Language Models: Construction and Verification

This data set is the data set and supplementary materials for the study of "An Automatic Scoring System for Chinese Verbal Creativity Based on Pre-Trained Language Models: Construction and Verification".This study aims to develop an automatic scoring system for Chinese creativity based on dynamic pre-trained language models (BERT, S-BERT, SimCSE) to provide automatic and reliable scoring for Chinese creativity.

This study consisted of two sub-studies, in which a total of 691 participants were recruited to complete four AUT tasks. The main purpose of the study is to investigate whether the model-rated scores can replace the traditional human-rated scores, that is, whether they are highly correlated and whether they have similar correlations with external criterion variables.