CAREER: Activity Discovery for Programmable & Adaptive Personalized Environments

本研究项目的目的是支持适应性和个性化的信息环境。这一愿景的核心是需要改进用户行为的活动发现和表示。这将需要开发新的机器学习算法来建立该活动的模型。(提出了支持循环预测后缀树的ABL代理语言的扩展。)这种方法将应用于两个创新领域，即用户与智能家电的交互和在线游戏环境中的半自主代理。

The intent of this research project is to support adaptive and personalized information environments. Central to this vision is the need to improve activity discovery and representation of user behavior. This will necessitate the development of novel machine learning algorithms for building models of that activity. (An extension of the ABL agent language with support for looping prediction suffix trees is proposed.) This approach will be applied to two innovative domains, user interactions with smart appliances and semi-autonomous agents in online gaming environments.