Explainable AI for Sentiment Analysis of Human Metapneumovirus (HMPV) Using XLNet

2025-02-01

AI Summary

Public emotional responses to human metapneumovirus (HMPV) outbreaks are difficult to model with interpretability using conventional approaches. Method: We propose the first explainable sentiment analysis framework integrating XLNet—a Transformer-based language model—with SHAP (Shapley Additive Explanations) for fine-grained analysis of HMPV-related social media texts. Our approach employs an XLNet classifier for high-accuracy sentiment prediction and leverages SHAP to generate user-level, causally grounded feature attributions. Contribution/Results: This work pioneers the synergistic application of XLNet and SHAP to respiratory virus-related舆情 analysis, overcoming the opacity of traditional black-box models. Evaluated on a real-world HMPV social media dataset, our framework achieves 93.50% accuracy—significantly outperforming BERT and LSTM baselines—while delivering intuitive, trustworthy, instance-level explanations. The framework thus enables transparent, evidence-informed public health surveillance and intervention decision-making.

Technology Category

• Humans and AI: Explainable AI (XAI) for Human Understanding
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Transparent, Interpretable, Explainable ML

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

In 2024, the outbreak of Human Metapneumovirus (HMPV) in China, which later spread to the UK and other countries, raised significant public concern. While HMPV typically causes mild symptoms, its effects on vulnerable individuals prompted health authorities to emphasize preventive measures. This paper explores how sentiment analysis can enhance our understanding of public reactions to HMPV by analyzing social media data. We apply transformer models, particularly XLNet, achieving 93.50% accuracy in sentiment classification. Additionally, we use explainable AI (XAI) through SHAP to improve model transparency.

Problem

• Analyze public sentiment on HMPV outbreak
• Apply XLNet for sentiment classification
• Enhance model transparency using explainable AI

Innovation

• XLNet for sentiment analysis
• Explainable AI with SHAP
• 93.50% accuracy achieved