A Taxonomy of Property Measures to Unify Active Learning and Human-centered Approaches to Data Labeling

Author:

Bernard Jürgen1,Hutter Marco2,Sedlmair Michael3,Zeppelzauer Matthias4,Munzner Tamara1

Affiliation:

1. University of British Columbia, Vancouver, Canada

2. TU Darmstadt, Darmstadt, Germany

3. University of Stuttgart, Stuttgart, Germany

4. St. Pölten University of Applied Sciences, St. Pölten, Austria

Abstract

Strategies for selecting the next data instance to label, in service of generating labeled data for machine learning, have been considered separately in the machine learning literature on active learning and in the visual analytics literature on human-centered approaches. We propose a unified design space for instance selection strategies to support detailed and fine-grained analysis covering both of these perspectives. We identify a concise set of 15 properties, namely measureable characteristics of datasets or of machine learning models applied to them, that cover most of the strategies in these literatures. To quantify these properties, we introduce Property Measures (PM) as fine-grained building blocks that can be used to formalize instance selection strategies. In addition, we present a taxonomy of PMs to support the description, evaluation, and generation of PMs across four dimensions: machine learning (ML) Model Output , Instance Relations , Measure Functionality , and Measure Valence . We also create computational infrastructure to support qualitative visual data analysis: a visual analytics explainer for PMs built around an implementation of PMs using cascades of eight atomic functions. It supports eight analysis tasks, covering the analysis of datasets and ML models using visual comparison within and between PMs and groups of PMs, and over time during the interactive labeling process. We iteratively refined the PM taxonomy, the explainer, and the task abstraction in parallel with each other during a two-year formative process, and show evidence of their utility through a summative evaluation with the same infrastructure. This research builds a formal baseline for the better understanding of the commonalities and differences of instance selection strategies, which can serve as the stepping stone for the synthesis of novel strategies in future work.

Funder

German Academic Exchange Service

Austrian Research Promotion Agency

Lower Austrian Research and Education Company

National Sciences and Engineering Research Council of Canada

Deutsche Forschungsgemeinschaft

Publisher

Association for Computing Machinery (ACM)

Subject

Artificial Intelligence,Human-Computer Interaction

Cited by 9 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Unpacking Human-AI interactions: From Interaction Primitives to a Design Space;ACM Transactions on Interactive Intelligent Systems;2024-08-02

2. Towards an understanding and explanation for mixed-initiative artificial scientific text detection;Information Visualization;2024-04-07

3. Understanding Novice's Annotation Process For 3D Semantic Segmentation Task With Human-In-The-Loop;Proceedings of the 29th International Conference on Intelligent User Interfaces;2024-03-18

4. VideoPro: A Visual Analytics Approach for Interactive Video Programming;IEEE Transactions on Visualization and Computer Graphics;2024-01

5. Label Engineering Methods for ML Systems;Lecture Notes in Networks and Systems;2024

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3