ShortcutLens: A Visual Analytics Approach for Exploring Shortcuts in Natural Language Understanding Dataset

Zhihua Jin, Xingbo Wang, Furui Cheng, Chunhui Sun, Qun Liu, Huamin Qu

Room: 109

2023-10-25T22:36:00ZGMT-0600Change your timezone on the schedule page
Exemplar figure, described by caption below
ShortcutLens is a visual analytics tool that assists NLU experts in conducting the multi-level exploration of shortcuts in NLU benchmark datasets. ShortcutLens consists of three visualization components. The Statistics View (b) helps users inspect the statistics about the benchmark dataset and shortcuts. It also allows users to conduct what-if analysis on shortcuts of interest. The Template View (c) enables users to check the relationship of shortcuts and inspect the statistics about individual shortcuts. The Instance View (d) displays the instances covered by selected shortcuts from the Template View. They enable users to gain a better understanding of benchmark dataset issues and inspire the creation of more challenging and pertinent benchmark datasets.
Fast forward
Full Video

Visual Analytics;Natural Language Understanding;Shortcut


Benchmark datasets play an important role in evaluating Natural Language Understanding (NLU) models. However, shortcuts—unwanted biases in the benchmark datasets—can damage the effectiveness of benchmark datasets in revealing models’ real capabilities. Since shortcuts vary in coverage, productivity, and semantic meaning, it is challenging for NLU experts to systematically understand and avoid them when creating benchmark datasets. In this paper, we develop a visual analytics system, ShortcutLens, to help NLU experts explore shortcuts in NLU benchmark datasets. The system allows users to conduct multi-level exploration of shortcuts. Specifically, Statistics View helps users grasp the statistics such as coverage and productivity of shortcuts in the benchmark dataset. Template View employs hierarchical and interpretable templates to summarize different types of shortcuts. Instance View allows users to check the corresponding instances covered by the shortcuts. We conduct case studies and expert interviews to evaluate the effectiveness and usability of the system. The results demonstrate that ShortcutLens supports users in gaining a better understanding of benchmark dataset issues through shortcuts, inspiring them to create challenging and pertinent benchmark datasets.