Personalizing and Adapting Behaviour Change Systems

旨在激励人们采取对他们和他们的社区有益的行为和行动的互动系统正在吸引政府、从业者和研究人员的注意。劝导技术（PTs）是一种交互系统，旨在激发理想的行为改变。尽管对PT的兴趣和投资不断增加，但当前的PT设计框架存在一个主要的限制：它们基于一刀切的方法，因此产生的PT不能个性化以适合目标用户。说服技术中的个性化是具有挑战性的，因为激励用户进行特定的行为或行动需要考虑到广泛的个人和环境相关因素。本研究的主要目标是发展、评估和完善新的多维机制（同时考虑多个因素），以个性化和适应说服技术，使其适合个人用户，并提高其促进积极行为的功效。首先，我们将使用通过用户研究收集的显性和隐性用户行为数据，使用交互日志和各种行为背景下的生理反应，开发用户对不同说服策略反应的多维个性化模型。其次，为了评估模型在各种说服环境中的有效性，我们将应用这些模型来设计、开发和评估新的模型驱动的个性化和自适应说服系统，目标是三个技术领域（数字游戏、移动和社交计算）和两个应用领域（健康/健康和安全/保安）。说服系统将针对不同的受众，包括处于危险和服务不足的加拿大人口：非洲新斯科舍省人和土著居民，老年人和年轻人。本研究的主要成果将是：1)开发新的个性化和适应说服系统的多维机制；2)将这些机制应用于设计和评估几种个性化说服系统，以激励和帮助用户做出有益的行为。总的来说，这项研究的成功完成将导致自适应系统设计领域的技术进步。我们的创新在加拿大电脑游戏产业中具有潜在的经济效益，能够帮助我们设计出更好的游戏。最后，我们的工作体现在社会公益的适应性和智能系统的更广泛框架中：它可以通过针对几个问题领域（包括安全/保障和健康/保健）来改善加拿大人的生活。这项工作可以帮助每个人，但也可以针对特定的，有风险的加拿大人和服务不足的人群（例如，非洲新斯科舍省人，土著居民，老年人）。

Interactive systems that aim to motivate people to adopt behaviours and actions that are beneficial for them and their community are attracting the attention of governments, practitioners, and researchers. Persuasive Technologies (PTs) are interactive systems designed to motivate desirable behaviour change. Despite the growing interest and investments into PTs, current PT design frameworks suffer a major limitation: They are based on the one-size-fits-all approach and hence resulting PTs are not personalized to be appropriate for the target user(s). Personalization in persuasive technologies is challenging, since motivating users for a particular behaviour or action needs to take into account a wide range of individual and context-dependent factors. The primary objective of this research is to develop, evaluate, and refine novel multi-dimensional mechanisms (that considers multiple factors simultaneously) for personalizing and adapting persuasive technologies to make them appropriate for individual users and increase their efficacy at promoting positive behaviour. First, we will develop multidimensional personalization models of user's responses to distinct persuasive strategies using both explicit and implicit user behaviour data gathered via user studies, using interaction log, and physiological responses across various behaviour contexts. Second, to evaluate the efficacy of the models across various persuasion contexts, we will apply the models to design, develop, and evaluate new model-driven personalized and adaptive persuasive systems targeting three technical domains (digital games, mobile, and social computing) and two application areas (health/wellness and safety/security). The persuasive systems will be targeted at diverse audience including at-risk and underserved Canadian populations: African Nova Scotians and Aboriginals, older adults, and youths.The primary outcomes of this research will be: 1) The development of novel multi-dimensional mechanisms for personalizing and adapting persuasive systems, and 2) applications of these mechanisms in designing and evaluating several personalized persuasive systems that motivate and assist users towards beneficial behaviours. Overall, successful completion of this research will lead to technological advances in the area of adaptive systems design. Our innovations have the potential for economic benefits within the Canadian computer games industry, to inform better game designs. Finally, our work inscribes in the broader framework of adaptive and intelligent systems for social good: it can improve the lives of Canadians by targeting several problem domains (including safety/security and health/wellness). This work can help everyone, but can also target specific, at-risk Canadians and under-served populations (e.g., African Nova Scotians, Aboriginals, older adults).