| dc.description.abstract | According to the World Health Organization, non-communicable diseases accounted for 68% of global deaths in 2012. Care for this type of diseases transcends patient engagement, extending to their family, friends and, acquaintances, who may influence their treatment posi- tively or negatively. Social support can be understood as the ability that social networks have in alleviating the harmful effects of stress and other health risks. Social support in the care of non- communicable diseases is assisted by computing mainly through information systems such as discussion forums, blogs, wikis and video sharing tools, which aim for increasing patients’ and families’ prevention and care of these diseases. However, few studies explore the big amount of data that are currently generated by smartphones, wearables, smart devices, and social media. Such data, if properly explored, can serve as a means for improving social support in the preven- tion and care of non-communicable diseases by recommending new beneficial health contacts, or by presenting the influence that individuals make on the health of others as a way of aware them about their habits and the impact of these habits on the lives of their peers. In other words, make computer-aided social support computer aware of the social context. This thesis presents the Pompilos model, which provides social aware preventive care for non-communicable dis- eases. Pompilos explores the data generated by the use of social media, smartphones, wearables and smart devices to infer users’ profiles and social networks. Profiles and social networks are used to account the influence that users exert on others in aspects related to the care and preven- tion of non-communicable diseases, such as food consumption, physical activity or smoking. This influence is used to recommend information that can improve users’ engagement in the prevention of non-communicable diseases, as well as to show which behaviors influence the health of their peers as a mean of social awareness. A prototype of the Pompilos model was developed and tested in two random experiments by 61 users. A health assistant application which implemented the model was provided to the users who were divided into two groups control and intervention. This application had regular features of noncommunicable diseases prevention as diet, physical activity practice, weight management and, history charts. To the intervention users, an extra feature was provided. This extra feature allowed users to receive messages related to noncommunicable diseases prevention collected from 15 monitored Twitter profiles. The intervention groups on both experiments had improved the use of the history charts when compared to the control group, indicating that they were more concerned in following up their care when receiving messages relating to noncommunicable diseases prevention. Finally, two monitored profiles changed posting behaviors when aware of the use of their messages to aid user on their care. | en |
