[Zurück]

N. TaheriNejad, A. Jantsch, D. Pollreisz:
"Comprehensive Observation and its Role in Self-Awareness; An Emotion Recognition System Example";
Vortrag: The Federated Conference on Computer Science and Information Systems (FedCSIS), Gdansk, Poland; 11.09.2016 - 14.09.2016; in: "Position Papers of the 2016 Federated Conference on Computer Science and Information Systems", (2016), ISBN: 978-83-60810-93-4; S. 117 - 124.

Observation plays a crucial role in self-awareness. In many scenarios, such as the Observe-Decide-Act (ODA) loops, self-awareness is founded upon observations of the system. In other words, observation generates the understanding of the system from the status and behavior of its self and its environment. Although recently more focus has been put on comprehensive and competent observations, we believe that further attention and work is due, especially in the field of cyber-physical systems. Hence, in this paper, we discuss our position on various aspects of observation methods. In a short list, the major aspects are Abstraction, Disambiguation, Desirability, Relevance, Data Reliability, Confidence, Attention, and History. We elaborate and anticipate the potential of these factors in improving the quality of the observation of the system, decreasing the processing load of higher layers, increasing the reliability of decisions, and consequently the overall performance of the system. To put these aspects into perspective, we elaborate them in the context of their potentials in our emotion recognition system under development.

http://publik.tuwien.ac.at/files/publik_254956.pdf