Reciprocal human machine learning
Reciprocal Human Machine Learning (RHML) is an interdisciplinary approach to designing human-AI interaction systems.[1] RHML aims to enable continual learning between humans and machine learning models by having them learn from each other. This approach keeps the human expert "in the loop" to oversee and enhance machine learning performance and simultaneously support the human expert continue learning.
Background[edit]
RHML emerged in the context of the rise of big data analytics and artificial intelligence for intelligent tasks like sense-making and decision-making.[2] As machine learning advanced to take on more roles, researchers realized fully autonomous systems had limitations and needed human guidance.[3][4][5][6]
RHML extends the concept of human-in-the-loop systems by promoting reciprocal learning. Humans learn from their interactions with machine learning models, staying up-to-date on evolving technology.[7] The models also learn from human feedback and oversight. This amplification of learning on both sides is a key focus of RHML.
The approach draws on theories of learning in dyads from education and psychology. It also builds on human-computer interaction and human-centered design principles. Implementing RHML requires developing specialized tools and interfaces tailored to the application[8]
Applications[edit]
RHML has been explored across diverse domains including:
- Cybersecurity - Software to enable reciprocal learning between experts and AI models for social media threat detection.[8]
- Organizational decision-making - RHML to structure collaboration between humans and AI systems.[9]
- Workplace training - Using RHML for workers to learn from AI technologies on the job.[10]
- Open science - Using human and AI collaboration to promote open science.[5]
- Production and logistics - turning workers and intelligent machines into teammates.[6]
RHML maintains human oversight and control over AI systems, while enabling cutting-edge machine learning performance. This collaborative approach highlights the importance of keeping the human expert involved in the loop.
References[edit]
- ^ Te'eni, Dov; Yahav, Inbal; Zagalsky, Alexey; Schwartz, David G.; Silverman, Gahl; Cohen, Daniel; Mann, Yossi; Lewinsky, Dafna (2023-11-14). "Reciprocal Human-Machine Learning: A Theory and an Instantiation for the Case of Message Classification". Management Science.
- ^ University of Virginia; Abbasi, Ahmed; Zhou, Yilu; Fordham University; Deng, Shasha; Shanghai International Studies University; Zhang, Pengzhu; Shanghai Jiaotong University (2018-02-02). "Text Analytics to Support Sense-Making in Social Media: A Language-Action Perspective" (PDF). MIS Quarterly. 42 (2): 427–464. doi:10.25300/MISQ/2018/13239.
- ^ van den Broek, Elmira; Sergeeva, Anastasia; Huysman, Marleen (2021-09-01). "When the Machine Meets the Expert: An Ethnography of Developing AI for Hiring". Management Information Systems Quarterly. 45 (3): 1557–1580. doi:10.25300/MISQ/2021/16559. ISSN 0276-7783. S2CID 238222671.
- ^ Sturm, Timo; Gerlach, Jin; Pumplun, Luisa; Mesbah, Neda; Peters, Felix; Tauchert, Christoph; Nan, Ning; Buxmann, Peter (2021-09-01). "Coordinating Human and Machine Learning for Effective Organization Learning". Management Information Systems Quarterly. 45 (3): 1581–1602. doi:10.25300/MISQ/2021/16543. ISSN 0276-7783. S2CID 238222756.
- ^ a b Elizabeth, Abigail (2023-09-08). "Reciprocal, human-machine learning could transform how researchers open science". OSF.
- ^ a b Nixdorf, Steffen; Zhang, Minqi; Ansari, Fazel; Grosse, Eric H. (2022-01-01). "Reciprocal Learning in Production and Logistics". IFAC-PapersOnLine. 10th IFAC Conference on Manufacturing Modelling, Management and Control MIM 2022. 55 (10): 854–859. doi:10.1016/j.ifacol.2022.09.519. ISSN 2405-8963. S2CID 253195774.
- ^ So, Chaehan (2020). "Human-in-the-Loop Design Cycles – A Process Framework that Integrates Design Sprints, Agile Processes, and Machine Learning with Humans". In Degen, Helmut; Reinerman-Jones, Lauren (eds.). Artificial Intelligence in HCI. Lecture Notes in Computer Science. Vol. 12217. Cham: Springer International Publishing. pp. 136–145. arXiv:2003.05268. doi:10.1007/978-3-030-50334-5_9. ISBN 978-3-030-50334-5. S2CID 212657395.
- ^ a b Zagalsky, Alexey; Te'eni, Dov; Yahav, Inbal; Schwartz, David G.; Silverman, Gahl; Cohen, Daniel; Mann, Yossi; Lewinsky, Dafna (2021-10-18). "The Design of Reciprocal Learning Between Human and Artificial Intelligence". Proceedings of the ACM on Human-Computer Interaction. 5 (CSCW2): 443:1–443:36. doi:10.1145/3479587. S2CID 239020698.
- ^ Shrestha, Yash Raj; Ben-Menahem, Shiko M.; von Krogh, Georg (August 2019). "Organizational Decision-Making Structures in the Age of Artificial Intelligence". California Management Review. 61 (4): 66–83. doi:10.1177/0008125619862257. S2CID 199310764.
- ^ Grønsund, Tor; Aanestad, Margunn (2020-06-01). "Augmenting the algorithm: Emerging human-in-the-loop work configurations". The Journal of Strategic Information Systems. 29 (2): 101614. doi:10.1016/j.jsis.2020.101614. hdl:10852/78524. S2CID 224771474.