A Multi-Paradigm Interpretability Framework for Case Outcome Prediction with Causal Reasoning
This conference paper proposes a multi-paradigm interpretability framework for legal case outcome prediction, emphasizing not only predictive accuracy but also explanation validity and causal reasoning. The work combines statistical pattern recognition, knowledge graph modeling, attention mechanisms, and causal inference to explore explainable AI for legal decision-support scenarios.
Presented as a poster at the 2026 4th IEEE International Conference on Pattern Recognition, Machine Vision and Artificial Intelligence (PRMVAI 2026). Authors visible on the poster include Ning Mou, Shuo Han, Yi-Chi Zhang, and Fan Wu.
面向案件結果預測的多範式可解釋性框架：結合因果推理
本研究提出一種面向法律案件結果預測的多範式可解釋性框架，不僅關注預測準確率，也強調解釋有效性與因果合理性。該框架結合統計模式識別、知識圖譜、注意力機制與因果推理，用於探索法律判斷輔助場景中可解釋 AI 模型的實際應用價值。
該論文於2026年第四屆IEEE模式識別、機器視覺與人工智慧國際會議（PRMVAI 2026）以海報形式發表。海報列出的作者包括 Ning Mou、Shuo Han、Yi-Chi Zhang、Fan Wu。