Speaker: Prof. Fernando Buarque de LIMA NETO
Titel: Computational Semiotics and The Semiotic Machine – A Primer
Abstract: Electronic Digital Computers have been the driving force of most technological advances worldwide, for over 40 years now. Despite the huge advances afforded and the high technological content embedded in them, digital processing has still been performed, chiefly, in a von Neumann manner. And even if we consider the foremost advanced hardware and software approaches, namely parallel platforms and intelligent algorithms, respectively, nowadays computers still have a timid grasp of the semantics what they are, in fact, processing. In addition to the computer machinery itself, the current large volume of information to be processed is ubiquitous and is, likewise, a major problem to which users must cope with. Thus, virtually in any decision situation, apt mechanisms/tools and methodologies that can deal appropriately with abundance of data are in high demand. In this talk, I will argue that more appropriate and tailored approaches to tackle more contextually the abundant information can benefit greatly from the adaptability of Artificial / Computational Intelligence when it is hybridized with Computational Semiotics. For that, the mastering of Semiotics applied to computation is seminal. A noticeable consequence of this integration is the possibility of improving the perception capacity of systems so they can output results of better quality, given the input demands (e.g. database queries). Hence, the definition and the main concepts of Semiotics and our proposed Semiotic Machine (SM) will be provided and discussed. SM is a theoretical concept, with almost complete implementation, being developed alongside other researchers at ERCIS. In short, SMs are modular constructs that mimic semeiotic chains and can perform sign-deconstructions, using Peircean trichotomies and triads, by means of Computational and Artificial Intelligent techniques. Finally, I will present, comment and exemplify current on-going "products" of SMs. Most importantly, we will argue and exemplify that SMs, boosted with heuristic methods can indeed improve the appropriateness of some decision support systems, granting more pragmatism to Computational Systems. The examples to be shown may reveal the potential to be of this current research.
FERNANDO BUARQUE DE LIMA NETO holds a doctorate in Artificial Intelligence from the University of London & Diploma of Imperial College London DIC in Artificial Neural Networks (2002). Dr. Fernando Buarque heads the Computational Intelligence Research Group (CIRG) at University of Pernambuco, Brazil. He is fellow of the CNPQ (Brazilian research accreditation agency) since 2010 and Alexander von Humboldt Fellow, Germany since 2015. He has more than 140 scientific papers published and directed about 30 theses and dissertations and 30 undergraduate term papers in Computer Engineering. His three lines of research are (1) Computational Intelligence (Metaheuristics Evolutionary, Social and Hybrid), (2) Modeling / Simulation of Stochastic and complexes problems, and (3) Intelligent Decision Support Systems (including aspects of Computational Semiotics). Regularly he receives foreign visitors from his collaboration network (12 universities, nine of them abroad), and often receives invitations to speak in/outside Brazil. Prof. Fernando Buarque is a full-time Associate Professor (with habilitation), based at the Polytechnic School of Pernambuco POLI, a century-old faculty of the University of Pernambuco UPE. In 2011 was appointed 'Visiting Professor' at the University of Johannesburg. In 2012 he visited as 'Professeur Invité' the Paris-Rocquencourt INRIA, France and was named 'Graduate Faculty' of the Florida Institute of Technology FIT, USA; and in 2016 was appointed Adjunct Professor of Texas A&M. Currently he is Program Chair of IEEE LA-CCI2017, Arequipa, Peru and is leading two task-forces, the first, for conceiving the first Doctoral Program dedicated to Computational Intelligence in Brazil, and the second, for creating a multinational networked Post-Graduate Program in Computer and Systems Engineering in Central America (for Countries that do not possess such kind of Program).