← 研踪 PaperTrailarXiv:2502.13149 · v3
Bi-Fact: A Bidirectional Factorization-based Evaluation of Intent Extraction from UI Trajectories
首次提交 2025-02-11 21:16
版本更新 2025-03-05 17:54
雷达观测 2026-11-01 22:46
摘要
Evaluating intent extraction from GUIs demands accurate, fine-grained metrics. This paper introduces Bi-Fact, a novel method that decomposes intents into atomic facts and performs bidirectional comparisons to assess precision and recall. Experiments demonstrate Bi-Fact's superior correlation with human judgments compared to existing metrics, establishing a more robust evaluation framework for UI-driven intent understanding.
论文发现出处
这些页面提供论文线索，不代表同行评审或质量认证。
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v3 · 2025-03-05 17:54 · 最新观测版本
Bi-Fact: A Bidirectional Factorization-based Evaluation of Intent Extraction from UI Trajectories
Evaluating intent extraction from GUIs demands accurate, fine-grained metrics. This paper introduces Bi-Fact, a novel method that decomposes intents into atomic facts and performs bidirectional comparisons to assess precision and recall. Experiments demonstrate Bi-Fact's superior correlation with human judgments compared to existing metrics, establishing a more robust evaluation framework for UI-driven intent understanding.
查看此版本解读 →查看此版本原文 ↗