You Are the Only Possible Oracle: Effective Test Selection for End Users of Interactive Machine Learning Systems
The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
How do you test a program when only a single user, with no expertise in software testing, is able to determine if the program is performing correctly? Such programs are common today in the form of machine-learned classifiers. We consider the problem of testing this common kind of machine-generated program when the only oracle is an end user: e.g., only you can determine if your email is properly filed. We present test selection methods that provide very good failure rates even for small test suites, and show that these methods work in both large-scale random experiments using a “gold standard” and in studies with real users. Our methods are inexpensive and largely algorithm-independent. Key to our methods is an exploitation of properties of classifiers that is not possible in traditional software testing. Our results suggest that it is plausible for time-pressured end users to interactively detect failures-even very hard-to-find failures-without wading through a large number of successful (and thus less useful) tests. We additionally show that some methods are able to find the arguably most difficult-to-detect faults of classifiers: cases where machine learning algorithms have high confidence in an incorrect result.
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.
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
Cortes, C., & Vapnik, V. (1995) · Machine Learning · doi:10.1007/bf00994018 · Cited by 59,421
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
A Sequential Algorithm for Training Text Classifiers
Lewis, D. D., & Gale, W. A. (1994) · SIGIR ’94 · doi:10.1007/978-1-4471-2099-5_1 · Cited by 1,208
All of Statistics: A Concise Course in Statistical Inference
Wasserman, L. (2004) · Springer Texts in Statistics · doi:10.1007/978-0-387-21736-9 · Cited by 1,090 · cited 1×
“Confidence intervals were computed over all runs for each training set size, in standard statistical practice [59].”
MethodAutomatic Indexing: An Experimental Inquiry
Maron, M. E. (1961) · Journal of the ACM · doi:10.1145/321075.321084 · Cited by 570 · cited 1×
“To produce multiple classifiers C to test, we trained a Naive Bayes (NB) classifier [40] and a support vector machine [12] on each training set.”
MethodWhy 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
What we know about spreadsheet errors
Panko, R. R. (1998) · Journal of Organizational and End User Computing · doi:10.4018/joeuc.1998040102 · Cited by 464
Toolkit to support intelligibility in context-aware applications
Lim, B. Y., & Dey, A. K. (2010) · Proceedings of the 12th ACM international conference on Ubiquitous computing · doi:10.1145/1864349.1864353 · Cited by 143 · cited 1×
“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].”
BackgroundFixing the program my computer learned: barriers for end users, challenges for the machine
Kulesza, T., Wong, W.-K., Stumpf, S., Perona, S., White, R., Burnett, M. M., Oberst, I., & Ko, A. J. (2009) · Proceedings of the 14th international conference on Intelligent user interfaces · doi:10.1145/1502650.1502678 · Cited by 83 · cited 1×
“In our experiments, we use information gain (over the entire universe of instances) to measure feature relevance; another possibility is for the user to tell the classifier which features are most relevant [35].”
MethodSelf-citation
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
Ali, S., Abuhmed, T., El-Sappagh, S., Muhammad, K., Alonso-Moral, J. M., Confalonieri, R., Guidotti, R., Del Ser, J., Díaz-Rodríguez, N., & Herrera, F. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.101805 · Cited by 1,762
A Multidisciplinary Survey and Framework for Design and Evaluation of Explainable AI Systems
Mohseni, S., Zarei, N., & Ragan, E. D. (2021) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3387166 · Cited by 785
A Multidisciplinary Survey and Framework for Design and Evaluation of Explainable AI Systems
Mohseni, S., Zarei, N., & Ragan, E. D. (2018) · arXiv · doi:10.48550/arxiv.1811.11839 · Cited by 783
“with the agent’s output, as well as on users’ confidence and experience with the system. Another use case for machine learning explanations is to help users to justify the correctness of system output [39,61,116]. Explanations also assist users in debugging interactive machine learning programs for their needs [63,65].…”
BackgroundPrinciples of Explanatory Debugging to Personalize Interactive Machine Learning
Kulesza, T., Burnett, M., Wong, W.-K., & Stumpf, S. (2015) · Proceedings of the 20th International Conference on Intelligent User Interfaces · doi:10.1145/2678025.2701399 · Cited by 576
Self-citationSix Human-Centered Artificial Intelligence Grand Challenges
Ozmen Garibay, O., Winslow, B., Andolina, S., Antona, M., Bodenschatz, A., Coursaris, C., Falco, G., Fiore, S. M., Garibay, I., Grieman, K., et al. (2023) · International Journal of Human–Computer Interaction · doi:10.1080/10447318.2022.2153320 · Cited by 501
“This may result in an inappropriate shift in responsibility where end users are expected to evaluate a tool ’ s performance without a desire or realistic capability of doing so (Amershi et al., 2014; Groce et al., 2014).”
BackgroundA Review of User Interface Design for Interactive Machine Learning
Dudley, J. J., & Kristensson, P. O. (2018) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3185517 · Cited by 343
Testing machine learning based systems: a systematic mapping
Riccio, V., Jahangirova, G., Stocco, A., Humbatova, N., Weiss, M., & Tonella, P. (2020) · Empirical Software Engineering · doi:10.1007/s10664-020-09881-0 · Cited by 251
Giray, G. (2021) · Journal of Systems and Software · doi:10.1016/j.jss.2021.111031 · Cited by 187
Deng, W. H., Guo, B., Devrio, A., Shen, H., Eslami, M., & Holstein, K. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581026 · Cited by 103
Toward involving end-users in interactive human-in-the-loop AI fairness
Nakao, Y., Stumpf, S., Ahmed, S., Naseer, A., & Strappelli, L. (2022) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3514258 · Cited by 69
Self-citationThe Machine Learning Algorithm as Creative Musical Tool
Fiebrink, R., & Caramiaux, B. (2016) · arXiv · doi:10.48550/arxiv.1611.00379 · Cited by 69
“…Olsen originally proposed this approach in the context of computational image analysis, but it has since been applied to a variety of other problems, such as webpage analysis [Amershi et al., 2015], social network group creation [Amershi et al., 2012] and system “debugging” [Groce et al., 2014].”
MethodPersonalising Explainable Recommendations: Literature and Conceptualisation
Naiseh, M., Jiang, N., Ma, J., & Ali, R. (2020) · Advances in Intelligent Systems and Computing · doi:10.1007/978-3-030-45691-7_49 · Cited by 42
“This style is defined in [22, 30, 12] where users are given control around how personalisation should be conducted.”
Background"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · arXiv
Metadata
- Authors
- Groce, A., Kulesza, T., Zhang, C., Shamasunder, S., Burnett, M., Wong, W.-K., Stumpf, S., Das, S., Shinsel, A., Bice, F., & McIntosh, K.
- Venue
- IEEE Transactions on Software Engineering
- Published
- 2014-03
- Volume
- 40
- Issue
- 3
- Pages
- 307-323
- ISSN
- 0098-5589 (ISSN-L)
- 1939-3520
- OpenAlex
- W2036779186
- Cited by
- 66
- References
- 9
- FWCI
- 18.70
- Type
- journal article
- Language
- en
- Open access
- closed
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.