Scholarlink

<p>Nowadays, Natural Language Processing (NLP) is making increasingly significant contributions, particularly in the field of education, where the development of intelligent NLP processing systems is receiving growing attention. This is an inevitable trend as technology continues to evolve, requiring both learners and educators to adapt proactively in order to move toward Education 5.0. This shift aims to enhance teaching quality and effectively capture learners’ trends and needs. This study focuses on leveraging the power of Transformer architectures in combination with traditional models, such as Convolutional Neural Networks and Recurrent Neural Network, to improve accuracy in tasks of sentiment and emotion classification, as well as determining cognitive levels based on Bloom’s taxonomy. In the task of classifying sentiment from student feedback, the proposed model achieved outstanding performance, with an accuracy of 95.71%. For the task of level classification based on Bloom’s taxonomy, the model achieved even higher performance, with an accuracy of 93.44%.</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

A Hybrid Model for Sentiment, Emotion, and Bloom’s Taxonomy Analysis

  • 期刊论文

  • 发表日期：

    2025年7月24日

查看全文
  • Hai T. Nguyen,
  • Anh K. Su,
  • Tinh N. Vo,
  • Loi Q. Duong,
  • Nathalie Guilbert,
  • Phuong Le

摘要

Nowadays, Natural Language Processing (NLP) is making increasingly significant contributions, particularly in the field of education, where the development of intelligent NLP processing systems is receiving growing attention. This is an inevitable trend as technology continues to evolve, requiring both learners and educators to adapt proactively in order to move toward Education 5.0. This shift aims to enhance teaching quality and effectively capture learners’ trends and needs. This study focuses on leveraging the power of Transformer architectures in combination with traditional models, such as Convolutional Neural Networks and Recurrent Neural Network, to improve accuracy in tasks of sentiment and emotion classification, as well as determining cognitive levels based on Bloom’s taxonomy. In the task of classifying sentiment from student feedback, the proposed model achieved outstanding performance, with an accuracy of 95.71%. For the task of level classification based on Bloom’s taxonomy, the model achieved even higher performance, with an accuracy of 93.44%.