Scholarlink

In order to extract emotion features more efficiently, enhance the model’s ability to understand the emotions of each category. In this paper, we mine the complex relationship between each emotion in a sentence through a multi-level interactive attention mechanism. we use emotion-guided constraint optimization methods to strengthen the representation capability of emotion query vectors and enhance the model’s feature capture for specific emotion categories. the experimental results show that this integrated approach effectively enhances the model’s performance in fine-grained sentiment analysis when dealing with data with imbalanced categories.

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

Fine-Grained Sentiment Analysis of Microblogs Based on Multidimensional Interactive Attention Sentiment Guidance

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chuanlei Zhang,
  • Tong Ye,
  • Hui Ma,
  • Di Sun,
  • Haifeng Fan,
  • Jianrong Li

摘要

In order to extract emotion features more efficiently, enhance the model’s ability to understand the emotions of each category. In this paper, we mine the complex relationship between each emotion in a sentence through a multi-level interactive attention mechanism. we use emotion-guided constraint optimization methods to strengthen the representation capability of emotion query vectors and enhance the model’s feature capture for specific emotion categories. the experimental results show that this integrated approach effectively enhances the model’s performance in fine-grained sentiment analysis when dealing with data with imbalanced categories.