Recognition of fuzzy contexts from temporal data under uncertainty case study: Activity recognition in smart homes

Intelligence within a system causes non-linearly behavior of system to achieve its goals, which is defined as desired world states of the intelligent system. To achieve the intended goals, an intelligent system actuates the world by realizing of strategies, scenarios, actions, activities and operations. To assist an intelligent system, we need information and knowledge from world to reason in normality of the world states and to evaluate how much the intelligent system succeeds. The non-linear behavior of intelligent systems makes it rather difficult to reason in normality of the world state, because there is no clear border or frontier to isolate the normal and abnormal world states. To do this judgment two criteria are considered. One criterion is the possible context that an activity can be accomplished in it and the other criterion is verification of the correct realization of activities. In this paper, we propose a temporal data-driven artificial intelligence technique to recognize contexts of scenarios and verify if a scenario is done in an appropriate context.