Device Interaction Graph: Directed Decision Graph for Settings Auto-Completion

Knowledge generated from appliances usage data can be used for personalizing and reducing the number of user interactions. Appliances such as Washing Machines (WM) are capable of storing the settings data selected by the user, during operation. Before the operation of a WM, the user has to enable the options for these settings for a preferred mode of operation. We have proposed a Device Interaction Graph solution that automatically recommends and auto-completes the user's preferred settings in real time. For direct manual interactions on the WM, the auto-completion solution is applied when the user incrementally changes the settings on the user interface. For voice based command interactions with the WM, the auto-completion solution would be applied after understanding any incomplete settings requirements mentioned by the user. This settings auto-completion solution improves the user experience by reducing the number of manual or verbal interactions required to enable the operation of the device. We have applied the auto-completion solution on the data obtained from 158,213 connected WMs and achieved recommendation accuracy of 89.79% with an average 3.132 interactions.