Incorporating Human Intention into Self-Adaptive Systems
Shihong Huang, Pedro Álvarez de Miranda · 2015
Self-adaptive systems are fed with contextual information from the environments in which the systems operate,from within themselves, and from the users. Traditional self-adaptive systems research has focused on inputs of systems performance, resources, exception, and error recovery that drive systems' reaction to their environments. The intelligent ability ofthese self-adaptive systems is impoverished without knowledge ofa user's covert attention (thoughts, emotions, feelings). As a result, it is difficult to build effective systems that anticipate and react to users' needs as projected by covert behavior. This paperpresents the preliminary research results on capturing users'intention through neural input, and in reaction, commanding actions from software systems (e.g., load an application) based on human intention. Further, systems can self-adapt and refine their behaviors driven by such human covert behavior. The long-term research goal is to incorporate and synergize human neural input.Thus establishing software systems with a self-adaptive capability to "feel" and "anticipate" users intentions and put the human in the loop.