Scholarlink

<p>This study explores the integration of ChatGPT in educational settings, focusing on how students perceive its role in enhancing programming instruction. By combining Learning Analytics (LA) and an innovative feature selection approach, the Improved Binary Exponential Distribution Optimization–Differential Evolution (IBEDO-DE) algorithm, this research aims to streamline data analysis and improve the classification accuracy of student feedback. The IBEDO-DE algorithm, designed with local search and boundary handling techniques, was evaluated through Support Vector Machine (SVM) and <i>k</i>-Nearest Neighbor (<i>k</i>-NN) classifiers using a real-world questionnaire dataset of student perceptions. Results indicate that IBEDO-DE achieved compact feature subsets while maintaining classification accuracy of up to 89.61%. These insights into student perceptions offer actionable guidance for educators aiming to integrate AI-driven tools effectively, with attention to interaction quality, programming support, response efficiency, and platform accessibility.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing educational analytics by analyzing student perceptions of ChatGPT in programming education using enhanced feature selection

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月5日

查看全文
  • Shaymaa E. Sorour,
  • Hosnia. M. M. Ahmed,
  • Mohammed Aljaafari,
  • Hanan E. Abdelkader

摘要

This study explores the integration of ChatGPT in educational settings, focusing on how students perceive its role in enhancing programming instruction. By combining Learning Analytics (LA) and an innovative feature selection approach, the Improved Binary Exponential Distribution Optimization–Differential Evolution (IBEDO-DE) algorithm, this research aims to streamline data analysis and improve the classification accuracy of student feedback. The IBEDO-DE algorithm, designed with local search and boundary handling techniques, was evaluated through Support Vector Machine (SVM) and k-Nearest Neighbor (k-NN) classifiers using a real-world questionnaire dataset of student perceptions. Results indicate that IBEDO-DE achieved compact feature subsets while maintaining classification accuracy of up to 89.61%. These insights into student perceptions offer actionable guidance for educators aiming to integrate AI-driven tools effectively, with attention to interaction quality, programming support, response efficiency, and platform accessibility.