Explanatory Debugging: Supporting End-User Debugging of Machine-Learned Programs

By Todd Kulesza, Simone Stumpf, Margaret M. Burnett, Weng-Keen Wong, Yann Riche, Travis Moore, Ian Oberst, Amber Shinsel, Kevin McIntosh. 2010

115 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Software

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Software

Citations in the corpus

Explanation in Human-AI Systems: A Literature Meta-Review, Synopsis of Key Ideas and Publications, and Bibliography for Explainable AI (2019, methodology): “Some studies used "debugging tasks" or "oddball detection" tasks, in which users' corrections were used to modify learned programs (e.g., Kulesza et al., 2010).”
Debugging Machine Learning Models (2016, result): “The work by Kulesza et al. (2010) provided explanations for certain types of errors, and several works have visually represented different model errors as in Brooks et al. (2015).”
Tell me more?: the effects of mental model soundness on personalizing an intelligent agent (2012, methodology): “Recent research has made inroads into supporting this type of functionality [1,11,14,16].”
On Human Intellect and Machine Failures: Troubleshooting Integrative Machine Learning Systems (2016, methodology): “On debugging a single classifier, researchers have developed techniques for a domain expert to interact with the machine learning process [4, 9, 13, 1].”
A Multidisciplinary Survey and Framework for Design and Evaluation of Explainable AI Systems (2018, methodology): “ut the intelligent system’s decision-making process. Analyzing users’ interviews, think-alouds, and self-explanations provides valuable information about the users’ thought processes and mental models[110].Onstudyingusercomprehension,Kuleszaetal.[111]studiedtheimpactofexplanation soundness and compl”
Mini-crowdsourcing end-user assessment of intelligent assistants: A cost-benefit study (2011, background): “These assistants are taking on increasingly critical roles, such as assisting in qualitative research [11].”
Outlining the Design Space of Explainable Intelligent Systems for Medical Diagnosis (2019, background): “In order to make users better understand and explain how the system works, the system should be transparent and show the mental model similar to human’s mental models [25,26].”
You Are the Only Possible Oracle: Effective Test Selection for End Users of Interactive Machine Learning Systems (2014, background): “Work by Lim and Dey has resulted in a toolkit for applications to generate explanations for popular machine learning systems [38], and a few systems add debugging capabilities to explanations [2], [33], [34].”
A comprehensive taxonomy for explainable artificial intelligence: a systematic survey of surveys on methods and concepts (2021, methodology): “ging, like the EluciDebug system [81], – adaption of features as provided by Crayons [44], and Crayons [44]”
Principles of Explanatory Debugging to Personalize Interactive Machine Learning (2015, background): “A risk in enabling users to provide feedback to a machine learning system is that they may actually make its predictions worse [e.g., 24, 38].”
Manifold: A Model-Agnostic Framework for Interpretation and Diagnosis of Machine Learning Models (2018, background): “Other research directions aim to identify important features and instances that are relevant to an issue within the model [7, 14–17].”
Visual Interaction with Deep Learning Models through Collaborative Semantic Inference (2019, background): “The model parameters can be updated through backward interactions [50,69].”
Explanation-Based Human Debugging of NLP Models: A Survey (2021, methodology): “…papers, we started from some pivotal EBHD work, e.g., (Kulesza et al., 2015; Ribeiro et al., 2016; Teso and Kersting, 2019), and added EBHD papers citing or being cited by the pivotal work, e.g., (Stumpf et al., 2009; Kulesza et al., 2010; Lertvittayakumjorn et al., 2020; Yao et al., 2021).”
User Trust in Intelligent Systems: A Journey Over Time (2016, methodology): “The intelligent system used in the case study was the AutoCoder [8].”