Focus of the project
We aim to use Artificial Social Agents (ASA), such as chatbots, virtual agents or social robots, to support people who struggle in maintaining a healthy lifestyle and who may suffer from welfare diseases, such as diabetes. We support these people and their social group with ASAs to reduce prejudice and increase pride. For example, a diabetic is upset with a friend who did not accommodate their specific diet at a party, while the friend is embarrassed because they forgot the diet. An ASA can obtain information from a patient and redistribute it strategically in a social group, taking into account what information can and should (not) be shared. In the example, the friend would, without shame, ask the ASA all about the diet and could proudly accommodate the patient’s dietary needs. The patient would be pleased that their diet is accommodated at the party without having to publicly ask for it.
Key research questions
A user will only share (sensitive) information with an ASA when it trusts the information will be used in a responsible, predictable manner (if it is used at all). Therefore, the ASA needs to build a relationship of mutual trust with the user and with users in the social group. To do this, we study the perception of artificial social agents and how to manage the perception in the social context.
Collaborations on ASA technology exist with the Human Media Interaction group of the University of Twente and the European H2020 project Council of Coaches.
Other collaborations within and outside the 4TU partners are being investigated.