Zhang, K., Hou, S., Song, J., & Xiao, R. (2025). Construction of an Evaluation Indicator System for Chinese Learning Apps Based on BERT-LDA. Journal of Technology and Chinese Language Teaching, 16(2), 23-47.
[张邝弋, 侯尚余, 宋靖雯, & 肖锐. (2025). 基于BERT-LDA的中文学习APP评价指标体系构建研究. 科技与中文教学 (Journal of Technology and Chinese Language Teaching), 16(2), 23-47.]

Full paper

Abstract/摘要：

This study explores the evaluation standards and quality challenges associated with Chinese learning apps in the context of increasing informatization and intelligence. It introduces a topic clustering algorithm derived from the BERT-LDA model and integrates an expert model for topic extraction utilizing Large Language Models (LLMs). A multidimensional and dynamic evaluation indicator system for Chinese learning apps is developed, focusing on core dimensions such as evaluation content (content quality), evaluation process (user experience), and evaluation outcomes (learning effectiveness). The validity of this system is confirmed through sentiment analysis tasks. Lastly, the study identifies future directions and potential risk challenges for creating evaluation indicator systems in international Chinese education digital resources, emphasizing intelligent, dynamic, and secure approaches.

本研究围绕中文学习APP在信息化及智能化趋势下的评价标准和质量挑战，提出了一种基于BERT-LDA模型的主题聚类算法，并结合LLMs的专家模型主题提取方法，从评价内容（内容质量）、评价过程（用户体验）、评价效果（学习成效）等核心维度构建了中文学习APP的多维度、动态化评价指标体系，并在情感分析任务验证其有效性，最后从智能化、动态化以及安全性等方面指明了未来国际中文教育数字资源评价指标体系构建的未来方向及风险挑战。