Recording Daily Health Status with Chatbot on Mobile Phone - A Preliminary Study

Abstract

To support in-home long-term care, we are studying techniques of mind sensing, which externalizes internal states of elderly people as words through conversations with agents or robots. We previously developed the memory-aid service, where a chatbot on a mobile phone autonomously talks to elderly people, to record their conditions, events, and memorandums. During experiments with healthy elders, we found that they regularly talked to the chatbot about health status, such as weight and blood pressure. This motivated us to use the mind sensing as an affordable and practical means to record daily health status. In this paper, we present a method where individual users can declare health metrics of their interests, and record them through the mind sensing. Specifically, for each user-defined metric, the chatbot asks the user the current value of the metric at the designated time. The text conversations are then put in a data mining process to extract time-series values of the metric. The time-series data is finally visualized as a graph, with which the user can review the health status. Our preliminary experiment shows that individual health metrics can be recorded and visualized successfully even without “connected” measuring instruments.