Research Front 2 aims at designing engaging user interfaces, consisting in interactive intelligent Human-System (IIHS) that engages inhabitants of residential buildings towards sober energy management through user interaction and that helps them to maintain their behavior change over time. According to J. Grudin {tooltip}1{end-texte}J. Grudin. 2017. From Tool to Partner: The Evolution of Human-Computer Interaction. Morgan Claypool publishers, 183 pages. https://doi.org/10.2200/S00745ED1V01Y201612HCI035.{end-tooltip}, the future of Human Computer Interaction (HCI) are smart digital partners. Thus, the goal is to investigate mixed initiative through symmetrical co-learning interactions: both parties will inform, explain, ask, suggest and learn from the other. It is a new paradigm for IIHS, which fits well the unicity of each home where knowledge raises up from confrontation of parties. Our hypothesis is that co-learning will leverage user engagement as it puts users back in the decision loop by letting them to control the system boundaries. The inhabitants could also better understand their consumption behavior thanks to the IHEMAS capacities. Precisely, the collection of activity labels/preferences will make it possible to design a prototype enabling inhabitants (1) to commit with the system in a symmetrical co-construction of knowledge; and, thanks to this acquired knowledge and incentive mechanisms, (2) to help inhabitants to commit in a long-term change of practices. Thus, co-learning should lead to more accurate profiles and consequently to more accurate and personalized incentives, contributing to increase trust and credibility. It will consequently integrate tools for interactive labelling, exploration, and generation of explanations.
However, it raises several issues: out-of-user context requests or suggestions, interactional and/or cognitive overload of requests, system’s requests or system’s reasoning understandability. To address these issues, the approach is twofold. First, the targeted IHEMAS will be designed as a plastic user interface, e.g. adaptation capability of the user interaction depending on the context of use. Second, in order to put the user in the loop, this IHEMAS will need to be able to explain what leads it to a result, and the user will need to be able to correct reasoning. The user interfaces will be designed to make understandable and modifiable the model learned by the system. Thanks its plasticity capability, user interaction offered by the IHEMAS will evolve from simple interactions at the beginning to more expert interactions over time with the ability to refute an IHEMAS analysis. Therefore, in this research front 2, the design of minimalistic and context-dependent User Interfaces (Uis) will be followed, consisting in reducing the nudges and information as much as possible, while supporting a semantic zoom, thereby taking benefit from advances in plasticity. The IHEMAS prototype will be compared to approaches that are easier to implement in research front 3.
Regarding PhD2, the research work has started with the PhD thesis {tooltip}2{end-texte} V. B. Nguyen, Conception de systèmes interactifs persuasifs : application au domaine de l’énergie, PhD thesis in Computer Sciences of the University Grenoble Alpes, supervised by G. Calvary and Y. Laurillau, 2019{end-tooltip} (ANR INVOLVED). It has investigated a novel interaction technique (TOP-sliders widget) {tooltip}3{end-texte}Y. Laurillau, V.-B. Nguyen, J. Coutaz, G. Calvary, N. Mandran, F. Camara, and R. Balzarini. 2018. The TOP-slider for Multi-criteria Decision Making by Non-specialists. In Proceedings of the 10th Nordic Conference on Human-Computer Interaction (NordiCHI ’18), 642–653. https://doi.org/10.1145/3240167.3240185 {end-tooltip} designed to make understandable for non-experts the Pareto-based optimization model used by the system to generate explanations and action plans. A second novel interaction technique (Plan4Actions widget) allows users to control and modify the action plan suggested by the system. In {tooltip}4{end-texte}Rikke Hagensby Jensen, Jesper Kjeldskov, and Mikael B. Skov. 2016. HeatDial: Beyond User Scheduling in Eco-Interaction. In Proceedings of the 9th Nordic Conference on Human-Computer Interaction – NordiCHI ’16, 1–10. https://doi.org/10.1145/2971485.2971525{end-tooltip}, householders were able to both benefit non-intrusive automatic features in the HeatDial system and to control the system’s reasoning system. In {tooltip}5{end-texte} Alper T. Alan, Enrico Costanza, Sarvapali D. Ramchurn, Joel Fischer, Tom Rodden, and Nicholas R. Jennings. 2016. Tariff Agent: Interacting with a Future Smart Energy System at Home. ACM Transactions on Computer-Human Interaction 23, 4: 1–28. https://doi.org/10.1145/2943770{end-tooltip}, the system was designed to understand and adjust results calculated by the system (e.g. best energy price) by temporarily overriding the learning. The challenge of this research work is to extend previous work by (1) considering adaptation capabilities to adapt depending on the expertise of the users; (2) by investigating novel interactions techniques to support co-learning.
1.. J. Grudin. 2017. From Tool to Partner: The Evolution of Human-Computer Interaction. Morgan Claypool publishers, 183 pages. https://doi.org/10.2200/S00745ED1V01Y201612HCI035.↩
2.. V. B. Nguyen, Conception de systèmes interactifs persuasifs : application au domaine de l’énergie, PhD thesis in Computer Sciences of the University Grenoble Alpes, supervised by G. Calvary and Y. Laurillau, 2019↩
3.. Y. Laurillau, V.-B. Nguyen, J. Coutaz, G. Calvary, N. Mandran, F. Camara, and R. Balzarini. 2018. The TOP-slider for Multi-criteria Decision Making by Non-specialists. In Proceedings of the 10th Nordic Conference on Human-Computer Interaction (NordiCHI ’18), 642–653. https://doi.org/10.1145/3240167.3240185. ↩
4.. Rikke Hagensby Jensen, Jesper Kjeldskov, and Mikael B. Skov. 2016. HeatDial: Beyond User Scheduling in Eco-Interaction. In Proceedings of the 9th Nordic Conference on Human-Computer Interaction – NordiCHI ’16, 1–10. https://doi.org/10.1145/2971485.2971525↩
5.. Alper T. Alan, Enrico Costanza, Sarvapali D. Ramchurn, Joel Fischer, Tom Rodden, and Nicholas R. Jennings. 2016. Tariff Agent: Interacting with a Future Smart Energy System at Home. ACM Transactions on Computer-Human Interaction 23, 4: 1–28. https://doi.org/10.1145/2943770. ↩