Uncertain event-based model for egocentric context sensing

Abstract

Calm Technology characterizes those technologies that move forth and back between the center and periphery of our attention. That is, while center denotes what one is currently focused on, periphery denotes what one is attuned to without focusing on it explicitly. Context-aware computing exploits such a concept by allowing applications to adapt their behavior (i.e. the center) in response to the context sensed within the environment (i.e. the periphery). An application in this setting should have minimal assumptions in order to operate while being able to dynamically adapt to and learn what the surrounding context offers.In this paper we discuss ongoing work in designing an event-based model that allows applications to egocentrically perceive the periphery and evaluate its relevance and uncertainty with respect to the center of the application attention. In particular we discuss our ongoing work in designing and developing ECHOES, an uncertain event-model for Egocentric computing. Characteristics of ECHOES are (a) departing from usual conjunctive pattern-matching algorithms implemented in event-notification models and, (b) achieving event correlation through the use of complex filters defined by means of Fuzzy Logic formulas. The paper introduces the specification of ECHOES as well as the design of an early prototype developed as a modified implementation of the SIENA Publish/Subscribe Middleware.