Situation awareness via abductive reasoning from Semantic Sensor data: A preliminary report
Year: 2009
Type: conference-paper
Abstract: Semantic sensor Web enhances raw sensor data with spatial, temporal, and thematic annotations to enable high-level reasoning. In this paper, we explore how abductive reasoning framework can benefit formalization and interpretation of sensor data to garner situation awareness. Specifically, we show how abductive logic programming techniques, in conjunction with symbolic knowledge rules, can be used to detect inconsistent sensor data and to generate human accessible description of the state of the world from consistent subset of the sensor data. We also show how trust/belief information can be incorporated into the interpreter to enhance reliability. For concreteness, we formalize weather domain and develop a meta-interpreter in Prolog to explain weather data. This preliminary work illustrates synthesis of high-level, reliable information for situation awareness by querying low-level sensor data.
Institution: Wright State University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2109326677