Bi-Fact: A Bidirectional Factorization-based Evaluation of Intent Extraction from UI Trajectories
This work addresses the problem of accurately evaluating intent extraction from GUIs for researchers and developers in the field of human-computer interaction.
The authors tackled the problem of evaluating intent extraction from GUIs and achieved a more robust evaluation framework with their proposed method, Bi-Fact, which showed superior correlation with human judgments. The exact numbers are not provided, but the results demonstrate improved precision and recall.
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.