생성형 AI 서비스에 대한 이용자의 평가와 만족도 분석

초록

Purpose : This study aims to analyze the causal relationship between users’ evaluation on generative AI service and overall satisfaction. Methods : The input variable dimensions include easiness to use dimension, usefulness dimension, enjoyment dimension, suitability dimension, and reliability dimension, whereas the response is measured by satisfaction dimension. Analysis is done by decision tree model based on Gini algorithm and random forest model. Results : The results of decision tree models reveal that the first splits are mostly done by reliability dimension variables and the next important is usefulness dimension variables. Enjoyment dimension variables and reliability dimension variables are utilized most for the second splits. The random forest model indicates that the first four ranked variables belong to reliability dimension based on the mean_decrease_gini index. Conclusion : The findings of this study suggest that reliability dimension, usefulness dimension, enjoyment dimension, easiness to use dimension, and suitability dimension are crucial for users’ overall satisfaction for generative AI service, all of which are worthwhile to denote for policy makers and business management.

키워드