This thesis explores how a computer could facilitate emotional support, focusing on the user group of informal carers. Informal carers are people who provide regular mental or physical assistance to another person, who could not manage without them, without formal payment. They save the UK £132 billion per year. However, many carers find themselves isolated by their caring commitments and may suffer from mental and physical health problems. Good emotional support can help reduce the negative effects of stress. We explore how an Intelligent Virtual Agent (IVA) could provide suitable emotional support to carers; how this emotional support should be adapted to the situation and personality of the carer; and how to add emotional context to support messages. To do this, we create a corpus of emotional support messages of different types and devise an algorithm that selects which type of emotional support to use for different types of stress. We investigate whether to adapt emotional support to personality, developing a novel method of measuring personality using sliders. We explore the identity of the support-giver and find that this affects the perceived supportiveness of an emotional support message. We investigate how emoticons add emotional context to messages, developing a proposed set of emoticons that depict core emotions that people use online. We find that gift emoticons can be used to enhance emotional support messages by representing an effort to 'cheer up' the carer. Finally, we explore how emotional support messages could be used by an IVA in six interviews with carers. Overall, we find that an IVA that helps a carer keep in contact with their personal social network and offers emotional support messages would be well-received by carers, but further work needs to be done to implement it within the framework of existing social media.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:698864 |
| Date | January 2016 |
| Creators | Smith, Kirsten Ailsa |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://digitool.abdn.ac.uk:80/webclient/DeliveryManager?pid=231019 |
Page generated in 0.0023 seconds