VisCoSo

Detection of crisis in socio-material systems
via visual-cognitive-social processes

36 months

	Address	Research Team
	via alla Cascata 56C	Project Manager	Roberta Ferrario
	38123 Trento
		Post-docs	Chiara Bassetti Cinzia Giorgetta Daniele Porello Francesco Setti
		PhDs	Sami Alkindi
		Visitors	Błażej Skrzypulec
		External Collaborators	Marco Cristani
Project Proposal Description Implementation Collaborations Conferences Publications	The project is financed by the Provincia Autonoma di Trento through the scheme Research Unit 2011 for a total amount of 600K euros. For the part concerning the integration of the ontological approach and computer vision, VisCosSo is being carried out in collaboration with the VIPS Laboratory (Vision, Image Processing & Sound) of the Dipartimento di Informatica at the University of Verona, with the supervision of Marco Cristani. The project will concentrate on the detection of critical situations that happens in the entanglement of visual-cognitive-social processes, the process of seeing a scene, forming a belief or an expectation and engaging in an interaction. We argue that, far from being a linear process, this is rather an entanglement, in which vision influences what agents believe, their beliefs determine them to interact in a certain way and sustain their institutions, the rules they are subject to or their previous interactions may let emerge beliefs and expectations, that may influence what they see and how they see it, and so on. Moreover, in a socio-material system, both humans and technical devices participate in and enact such entanglement. Many disciplines, ranging from philosophy to cognitive sciences, artificial intelligence and sociology, have dealt with such phenomena, often focusing only on a part of the entanglement and neglecting the others. The scenario of the project is that of designed socio-material systems, intended as complex systems embedding human and artificial agents, material and institutional artifacts and resources, in which behavior and practices are partly constrained by norms. The complexity of such systems makes it very hard for them to cope with critical situations, as they emerge from the interplay between all participants and cannot be reduced to mere technical malfunctioning or to the negligence/malevolence of the human actors involved, but should be rather dealt with at a systemic level. The challenge addressed by this project is that of enabling a wider and more comprehensive perspective by leveraging on the results of the involved disciplines, thus allowing for the implementation of socio-material systems that are aware of the criticalities that can emerge in the entanglement of visual- cognitive-social processes. An ontological model will be built that captures the insights coming from different disciplines, reached with their own specific paradigms, making use of their own concepts and methodologies. The conceptual clarification resulting from the ontological analysis will enable: scholars of the various disciplines involved to better understand each other, thus enhancing knowledge on their specific field by cross-fertilization of studies designers of socio-material systems to be more aware of the entanglement, thus preventing as much as possible critical situations and preparing a systemic reaction to them agents (human and artificial) that participate in the system to better prevent, recognize, recover and learn from critical situations, and to infer new unexpected critical situations "at run time", while the system evolves, by making it in a sense self-transparent. The expected outcome will be a theoretical artifact, but meant to be practically used within the system and co-evolving with it, and enabling the agents in the system to collectively manage critical situations thanks to an enhanced self- transparency.