Todd Kulesza
Margaret Burnett
Simone Stumpf
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.
- Amershi, S., Cakmak, M., Knox, W. B., and Kulesza, T. Power to the people: The role of humans in interactive machine learning. AI Magazine (in press).Google Scholar
- Amershi, S., Fogarty, J., Kapoor, A., and Tan, D. Examining multiple potential models in end-user interactive concept learning. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2010), 1357--1360. Google ScholarDigital Library
- Blackwell, A. F. First steps in programming: A rationale for attention investment models. In Proceedings of the IEEE 2002 Symposia on Human Centric Computing Languages and Environments, IEEE (2002), 2--10. Google ScholarDigital Library
- Bostandjiev, S., O'Donovan, J., and Höllerer, T. TasteWeights: A visual interactive hybrid recommender system. In Proceedings of the 6th ACM Conference on Recommender Systems (2012), 35--42. Google ScholarDigital Library
- Bryan, N. J., Mysore, G. J., and Wang, G. ISSE: An interactive source separation editor. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2014), 257--266. Google ScholarDigital Library
- Bunt, A., Lount, M., and Lauzon, C. Are explanations always important? A study of deployed, low-cost intelligent interactive systems. In Proceedings of the 2012 ACM International Conference on Intelligent User Interfaces (2012), 169--178. Google ScholarDigital Library
- Cakmak, M., Chao, C., and Thomaz, A. L. Designing interactions for robot active learners. IEEE Transactions on Autonomous Mental Development 2, 2 (2010), 108--118. Google ScholarDigital Library
- Cramer, H., Evers, V., Ramlal, S., van Someren, M., Rutledge, L., Stash, N., Aroyo, L., and Wielinga, B. The effects of transparency on trust in and acceptance of a content-based art recommender. User Modeling and User-Adapted Interaction 18, 5 (2008), 455--496. Google ScholarDigital Library
- Craven, M. W., and Shavlik, J. W. Using neural networks for data mining. Future generation computer systems 13 (1997), 211--229. Google ScholarDigital Library
- Dzindolet, M. T., Peterson, S. A., Pomranky, R. A., Pierce, L. G., and Beck, H. P. The role of trust in automation reliance. International Journal of Human-Computer Studies 58, 6 (2003), 697--718. Google ScholarDigital Library
- Fails, J. A., Olsen, D. R., and Jr. Interactive machine learning. In Proceedings of the 8th International Conference on Intelligent User Interfaces (2003), 39--45. Google ScholarDigital Library
- Fiebrink, R., Cook, P. R., and Trueman, D. Human model evaluation in interactive supervised learning. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2011), 147--156. Google ScholarDigital Library
- Fogarty, J., Tan, D., Kapoor, A., and Winder, S. CueFlik: Interactive concept learning in image search. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2008), 29--38. Google ScholarDigital Library
- Groce, A., Kulesza, T., Zhang, C., Shamasunder, S., Burnett, M. M., Wong, W.-K., Stumpf, S., Das, S., Shinsel, A., Bice, F., and McIntosh, K. You are the only possible oracle: Effective test selection for end users of interactive machine learning systems. IEEE Transactions on Software Engineering 40, 3 (2014), 307--323. Google ScholarDigital Library
- Hart, S. G., and Staveland, L. E. Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. Advances in Psychology 52 (1988), 139--183.Google ScholarCross Ref
- Herlocker, J. L., Konstan, J. A., and Riedl, J. Explaining collaborative filtering recommendations. In Proceedings of the 2000 ACM Conference on Computer Supported Cooperative Work (2000), 241--250. Google ScholarDigital Library
- Johnson-Laird, P. N. Mental models: Towards a cognitive science of language, inference, and consciousness. Harvard University Press, 1983. Google ScholarDigital Library
- Kapoor, A., Lee, B., Tan, D., and Horvitz, E. Interactive optimization for steering machine classification. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2010), 1343--1352. Google ScholarDigital Library
- Kempton, W. Two theories of home heat control. Cognitive Science 10 (1986), 75--90.Google ScholarCross Ref
- Kibriya, A. M., Frank, E., Pfahringer, B., and Holmes, G. Multinomial naive Bayes for text categorization revisited. In AI 2004: Advances in Artificial Intelligence. Springer Berlin Heidelberg, Berlin, Heidelberg, 2004, 488--499. Google ScholarDigital Library
- Knox, W. B., and Stone, P. Reinforcement learning from human reward: Discounting in episodic tasks. In Proceedings of the 21st IEEE International Symposium on Robot and Human Interactive Communication (2012), 878--885.Google ScholarCross Ref
- Ko, A. J., and Myers, B. A. Debugging reinvented. the 13th international conference (2008), 301--310. Google ScholarDigital Library
- Kulesza, T., Stumpf, S., Burnett, M. M., and Kwan, I. Tell me more? The effects of mental model soundness on personalizing an intelligent agent. In Proceedings of the ACM Conference on Human Factors in Computing Systems (2012), 1--10. Google ScholarDigital Library
- Kulesza, T., Stumpf, S., Burnett, M. M., Wong, W.-K., Riche, Y., Moore, T., Oberst, I., Shinsel, A., and McIntosh, K. Explanatory debugging: Supporting end-user debugging of machine-learned programs. In Proceedings of the 2010 IEEE Symposium on Visual Languages and Human-Centric Computing (2010), 41--48. Google ScholarDigital Library
- Kulesza, T., Stumpf, S., Burnett, M. M., and Yang, S. Too much, too little, or just right? Ways explanations impact end users' mental models. Proceedings of the 2013 IEEE Symposium on Visual Languages and Human-Centric Computing (2013), 3--10.Google ScholarCross Ref
- Kulesza, T., Stumpf, S., Wong, W.-K., Burnett, M. M., Perona, S., Ko, A. J., and Oberst, I. Why-oriented end-user debugging of naive Bayes text classification. ACM Transactions on Interactive Intelligent Systems 1, 1 (2011). Google ScholarDigital Library
- Lacave, C., and Díez, F. J. A review of explanation methods for Bayesian networks. The Knowledge Engineering Review 17, 2 (2002), 107--127. Google ScholarDigital Library
Index Terms
- Principles of Explanatory Debugging to Personalize Interactive Machine Learning
Recommendations
Recent research advances on interactive machine learning
Interactive machine learning (IML) is an iterative learning process that tightly couples a human with a machine learner, which is widely used by researchers and practitioners to effectively solve a wide variety of real-world application problems. ...
A Review of User Interface Design for Interactive Machine Learning
Special Issue on Human-Centered Machine LearningInteractive Machine Learning (IML) seeks to complement human perception and intelligence by tightly integrating these strengths with the computational power and speed of computers. The interactive process is designed to involve input from the user but ...
Marcelle: Composing Interactive Machine Learning Workflows and Interfaces
UIST '21: The 34th Annual ACM Symposium on User Interface Software and TechnologyHuman-centered approaches to machine learning have established theoretical foundations, design principles and interaction techniques to facilitate end-user interaction with machine learning systems. Yet, general-purpose toolkits supporting the design of ...
Comments