Principles of Explanatory Debugging to Personalize Interactive Machine Learning
The DOI matches its Crossref record; 47 of the 48 references it cites were found in a registry or an index.
Abstract
How can end users efficiently influence the predictions that machine learning systems make on their behalf? This paper presents Explanatory Debugging, an approach in which the system explains to users how it made each of its predictions, and the user then explains any necessary corrections back to the learning system. We present the principles underlying this approach and a prototype instantiating it. An empirical evaluation shows that Explanatory Debugging increased participants' understanding of the learning system by 52% and allowed participants to correct its mistakes up to twice as efficiently as participants using a traditional learning system.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“we identified patient-specific, salient features and utilized only the top three salient features with the highest scores to avoid overwhelming users [35, 41].”
From Understanding the Effect of Counterfactual Explanations on Trust and Reliance on AI for Human-AI Collaborative Clinical Decision Making (2023) · cites this paper as [35]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It can also promote user learning and insight from complex data, as humans increasingly work with complex inferential systems for analytic purposes [44,45,73].”
From Progressive Disclosure: When, Why, and How Do Users Want Algorithmic Transparency Information? (2020) · cites this paper as [44]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In some settings there are expected benefits: Transparency improves algorithmic perceptions, because users better understand system behavior [43,44,51].”
From Progressive Disclosure: When, Why, and How Do Users Want Algorithmic Transparency Information? (2020) · cites this paper as [44]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“There are many other ways to illustrate how an algorithm operates including verbal explanations, concrete user exploration, and so forth [44,45,51,73,88].”
From Progressive Disclosure: When, Why, and How Do Users Want Algorithmic Transparency Information? (2020) · cites this paper as [44]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It can also promote user learning and insight from complex data, which is important as humans increasingly work with complex inferential systems for analytic purposes [43,71].”
From Progressive disclosure (2019) · cites this paper as [43]
Experimental — accuracy not yet measured
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Development of NASA-TLX (Task Load Index): Results of Empirical and Theoretical Research
(1988) · Advances in Psychology · doi:10.1016/s0166-4115(08)62386-9 · Cited by 15,003
Norman, D. (2016) · Vahlen eBooks · doi:10.15358/9783800648108 · Cited by 6,374
(2009) · Minds at UW (University of Wisconsin) · Cited by 4,791
A Comparative Study on Feature Selection in Text Categorization
Yang, Y., & Pedersen, J. (1997) · Cited by 4,790
Mental Models: Towards a Cognitive Science of Language, Inference, and Consciousness
Ford, M., & Johnson-Laird, P. N. (1985) · Language · doi:10.2307/414498 · Cited by 4,419
(2013) · Cited by 1,951
Explaining collaborative filtering recommendations
Herlocker, J. L., Konstan, J. A., & Riedl, J. (2000) · Proceedings of the 2000 ACM conference on Computer supported cooperative work · doi:10.1145/358916.358995 · Cited by 1,731 · cited 1×
“An alternative to the “black box” model is a “white box” model [16], in which the machine learning system is made more transparent to the user.”
BackgroundThe role of trust in automation reliance
Dzindolet, M. T., Peterson, S. A., Pomranky, R. A., Pierce, L. G., & Beck, H. P. (2003) · International Journal of Human-Computer Studies · doi:10.1016/s1071-5819(03)00038-7 · Cited by 1,276 · cited 1×
“Increased transparency has been associated with many benefits, including increased user acceptance of (or satisfaction with) recommendations and predictions [4, 8, 42], improved training of intelligent agents [41], and increased trust in the system’s predictions [10].”
BackgroundPower to the People: The Role of Humans in Interactive Machine Learning
Amershi, S., Cakmak, M., Knox, W. B., & Kulesza, T. (2014) · AI Magazine · doi:10.1609/aimag.v35i4.2513 · Cited by 1,084
Self-citationFails, J. A., & Olsen, D. R. (2003) · Proceedings of the 8th international conference on Intelligent user interfaces · doi:10.1145/604045.604056 · Cited by 633
Why and why not explanations improve the intelligibility of context-aware intelligent systems
Lim, B. Y., Dey, A. K., & Avrahami, D. (2009) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1518701.1519023 · Cited by 552 · cited 3×
“Other work has studied the impact of machine-generated explanations on users’ mental models, but did not explore whether this helped participants’ personalize their learning systems [8, 25, 30].”
BackgroundWhich is the correct statistical test to use?
McCrum-Gardner, E. (2008) · British Journal of Oral and Maxillofacial Surgery · doi:10.1016/j.bjoms.2007.09.002 · Cited by 479
The effects of transparency on trust in and acceptance of a content-based art recommender
Cramer, H., Evers, V., Ramlal, S., van Someren, M., Rutledge, L., Stash, N., Aroyo, L., & Wielinga, B. (2008) · User Modeling and User-Adapted Interaction · doi:10.1007/s11257-008-9051-3 · Cited by 448 · cited 2×
“Other work has studied the impact of machine-generated explanations on users’ mental models, but did not explore whether this helped participants’ personalize their learning systems [8, 25, 30].”
BackgroundMultinomial Naive Bayes for Text Categorization Revisited
Kibriya, A. M., Frank, E., Pfahringer, B., & Holmes, G. (2004) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-30549-1_43 · Cited by 427
Too much, too little, or just right? Ways explanations impact end users' mental models
Kulesza, T., Stumpf, S., Burnett, M., Yang, S., Kwan, I., & Wong, W.-K. (2013) · 2013 IEEE Symposium on Visual Languages and Human Centric Computing · doi:10.1109/vlhcc.2013.6645235 · Cited by 388 · cited 10×
“Principle 1.3: Be Complete Kulesza et al. describe completeness as “the extent to which all of the underlying system is described by the explanation” [25], so a complete explanation does not omit important information about the model.”
BackgroundInfluentialSelf-citationTintarev, N., & Masthoff, J. (2012) · User Modeling and User-Adapted Interaction · doi:10.1007/s11257-011-9117-5 · Cited by 334
Two Theories of Home Heat Control*
Kempton, W. (1986) · Cognitive Science · doi:10.1207/s15516709cog1001_3 · Cited by 327
Debugging reinvented: asking and answering why and why not questions about program behavior
Ko, A. J., & Myers, B. A. (2008) · Proceedings of the 30th international conference on Software engineering · doi:10.1145/1368088.1368130 · Cited by 296
TasteWeights: a visual interactive hybrid recommender system
Bostandjiev, S., O'Donovan, J., & Höllerer, T. (2012) · Proceedings of the sixth ACM conference on Recommender systems · doi:10.1145/2365952.2365964 · Cited by 274
Learning from a learning thermostat: lessons for intelligent systems for the home
Yang, R., & Newman, M. W. (2013) · Proceedings of the 2013 ACM international joint conference on Pervasive and ubiquitous computing · doi:10.1145/2493432.2493489 · Cited by 268 · cited 1×
“As Yang and Newman found when studying users of learning thermostats [46], a system that appears to disregard user feedback deters users from continuing to provide feedback.”
BackgroundUsing neural networks for data mining
Craven, M. W., & Shavlik, J. W. (1997) · Future Generation Computer Systems · doi:10.1016/s0167-739x(97)00022-8 · Cited by 257
Assessing demand for intelligibility in context-aware applications
Lim, B. Y., & Dey, A. K. (2009) · Proceedings of the 11th international conference on Ubiquitous computing · doi:10.1145/1620545.1620576 · Cited by 246 · cited 1×
“One method for evaluating completeness is via Lim and Dey’s intelligibility types [29], with more complete explanations including more of these intelligibility types.”
MethodA review of explanation methods for Bayesian networks
LACAVE, C., & DÍEZ, F. J. (2002) · The Knowledge Engineering Review · doi:10.1017/s026988890200019x · Cited by 237
First steps in programming: a rationale for attention investment models
Blackwell, A. F. (2003) · Proceedings IEEE 2002 Symposia on Human Centric Computing Languages and Environments · doi:10.1109/hcc.2002.1046334 · Cited by 234
EnsembleMatrix: interactive visualization to support machine learning with multiple classifiers
Talbot, J., Lee, B., Kapoor, A., & Tan, D. S. (2009) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1518701.1518895 · Cited by 226
Tell me more?: the effects of mental model soundness on personalizing an intelligent agent
Kulesza, T., Stumpf, S., Burnett, M., & Kwan, I. (2012) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/2207676.2207678 · Cited by 225 · cited 7×
“To help users build useful mental models of the learning system, these explanations should observe the following principles: Research has suggested that users may best personalize a learning system if they build their mental models while interacting with it [23].”
BackgroundInfluentialSelf-citationHuman model evaluation in interactive supervised learning
Fiebrink, R., Cook, P. R., & Trueman, D. (2011) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1978942.1978965 · Cited by 207
Interacting meaningfully with machine learning systems: Three experiments
Stumpf, S., Rajaram, V., Li, L., Wong, W.-K., Burnett, M., Dietterich, T., Sullivan, E., & Herlocker, J. (2009) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2009.03.004 · Cited by 184
Self-citationCueFlik: interactive concept learning in image search
Fogarty, J., Tan, D., Kapoor, A., & Winder, S. (2008) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1357054.1357061 · Cited by 178
Principles and Heuristics for Designing Minimalist Instruction
van der Meij, H., & Carroll, J. M. (1998) · The MIT Press eBooks · doi:10.7551/mitpress/4616.003.0003 · Cited by 158
Designing Interactions for Robot Active Learners
Cakmak, M., Chao, C., & Thomaz, A. L. (2010) · IEEE Transactions on Autonomous Mental Development · doi:10.1109/tamd.2010.2051030 · Cited by 151
How it works: a field study of non-technical users interacting with an intelligent system
Tullio, J., Dey, A. K., Chalecki, J., & Fogarty, J. (2007) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1240624.1240630 · Cited by 137
Interactive optimization for steering machine classification
Kapoor, A., Lee, B., Tan, D., & Horvitz, E. (2010) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1753326.1753529 · Cited by 133
Some Observations on Mental Models
(2014) · Mental Models · doi:10.4324/9781315802725-5 · Cited by 128
Are explanations always important?: a study of deployed, low-cost intelligent interactive systems
Bunt, A., Lount, M., & Lauzon, C. (2012) · Proceedings of the 2012 ACM international conference on Intelligent User Interfaces · doi:10.1145/2166966.2166996 · Cited by 106 · cited 1×
“Research suggests that to elicit corrections from users, this feedback mechanism should embody the following principles: Principle 2.1: Be Actionable Both theory [3] and prior empirical findings [6, 23, 25] suggest end users will ignore explanations when the benefit of attending to them is unclear.”
BackgroundInfluentialExplanatory Debugging: Supporting End-User Debugging of Machine-Learned Programs
Kulesza, T., Stumpf, S., Burnett, M., Wong, W.-K., Riche, Y., Moore, T., Oberst, I., Shinsel, A., & McIntosh, K. (2010) · 2010 IEEE Symposium on Visual Languages and Human-Centric Computing · doi:10.1109/vlhcc.2010.15 · Cited by 93
Self-citationTeaching science as continuous inquiry
Rowe, M. B. (1973) · Medical Entomology and Zoology · Cited by 83
Why-oriented end-user debugging of naive Bayes text classification
Kulesza, T., Stumpf, S., Wong, W.-K., Burnett, M. M., Perona, S., Ko, A. J., & Oberst, I. (2011) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/2030365.2030367 · Cited by 71
Self-citationGroce, A., Kulesza, T., Zhang, C., Shamasunder, S., Burnett, M., Wong, W.-K., Stumpf, S., Das, S., Shinsel, A., Bice, F., et al. (2014) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2013.59 · Cited by 66
Self-citationReinforcement learning from human reward: Discounting in episodic tasks
Knox, W. B., & Stone, P. (2012) · 2012 IEEE RO-MAN: The 21st IEEE International Symposium on Robot and Human Interactive Communication · doi:10.1109/roman.2012.6343862 · Cited by 65
Vig, J., Sen, S., & Riedl, J. (2011) · Proceedings of the 16th international conference on Intelligent user interfaces · doi:10.1145/1943403.1943418 · Cited by 46
Examining multiple potential models in end-user interactive concept learning
Amershi, S., Fogarty, J., Kapoor, A., & Tan, D. (2010) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1753326.1753531 · Cited by 41
Strategies and behaviors of end-user programmers with interactive fault localization
Prabhakararao, S., Cook, C., Ruthruff, J., Creswick, E., Main, M., Durham, M., & Burnett, M. (2004) · IEEE Symposium on Human Centric Computing Languages and Environments, 2003. Proceedings. 2003 · doi:10.1109/hcc.2003.1260197 · Cited by 40
Self-citationISSE: an interactive source separation editor
Bryan, N. J., Mysore, G. J., & Wang, G. (2014) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/2556288.2557253 · Cited by 36
Improving Understanding and Trust with Intelligibility in Context-Aware Applications
Lim, B. Y. (2012) · KiltHub Repository · doi:10.1184/r1/6720179 · Cited by 29
GAIA: A CAD Environment for Model-Based Adaptation of Game-Playing Software Agents
Rugaber, S., Goel, A. K., & Martie, L. (2013) · Procedia Computer Science · doi:10.1016/j.procs.2013.01.004 · Cited by 20
Explaining Naive Bayes Classifications
Greiner, R., Poulin, B., Lu, P., Anvik, J., Lu, Z., Macdonell, C., Wishart, D. S., Eisner, R., & Szafron, D. (2003) · doi:10.7939/r36d5ph6n · Cited by 12
Research works printed but not linked: 2
- 38. Thomaz, A. L., and Breazeal, C. Transparency and socially guided machine learning. In Proceedings of the 5th International Conference on Development and Learning (2006).
- 39. Tintarev, N., and Mastho , J. Evaluating the e ectiveness of explanations for recommender systems. User Modeling and User-Adapted Interaction 22 , 4-5 (2012), 399-439.
Cited by
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555
Explanation in artificial intelligence: Insights from the social sciences
Miller, T. (2019) · Artificial Intelligence · doi:10.1016/j.artint.2018.07.007 · Cited by 5,302
What is AI Literacy? Competencies and Design Considerations
Long, D., & Magerko, B. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376727 · Cited by 3,241
Algorithms at Work: The New Contested Terrain of Control
Kellogg, K. C., Valentine, M. A., & Christin, A. (2020) · Academy of Management Annals · doi:10.5465/annals.2018.0174 · Cited by 2,732
Guidelines for Human-AI Interaction
Amershi, S., Weld, D., Vorvoreanu, M., Fourney, A., Nushi, B., Collisson, P., Suh, J., Iqbal, S., Bennett, P. N., Inkpen, K., et al. (2019) · Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3290605.3300233 · Cited by 2,439
“For example, a large body of work exists and continues to grow around how to increase transparency or explain the behaviors of AI systems (e.g., [14, 21, 23, 36, 38, 44], to name a few).”
Background
Metadata
- Authors
- Kulesza, T., Burnett, M., Wong, W.-K., & Stumpf, S.
- Affiliations
- Oregon State University (US)
- City, University of London (GB)
- Venue
- Proceedings of the 20th International Conference on Intelligent User Interfaces
- Publisher
- ACM
- Published
- 2015-03-18
- Pages
- 126-137
- OpenAlex
- W2053075547
- Cited by
- 576
- References
- 48
- Citations by year
- 2026: 59 · 2025: 47 · 2024: 58 · 2023: 64 · 2022: 50 · 2021: 76 · 2020: 74 · 2019: 60 · 2018: 48 · 2017: 17 · 2016: 18 · 2015: 5
- FWCI
- 16.53
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM International Conference on Intelligent User Interfaces (IUI) (publishedVersion)
- City Research Online (City University London) (acceptedVersion)
- ENLIGHTEN (Jurnal Bimbingan dan Konseling Islam) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Physical Sciences › Computer Science › Human-Computer Interaction › Usability and User Interface Design
- Keywords
- interactive machine learning, personalized machine learning, explainable machine learning, user feedback, prediction correction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.