The Peripatetic conferences on Cognitive Systems Modeling are organized since 2010 by and for those who appreciate thinking and exchanging ideas while walking and those who, like many before, believe that the rarefied mountain air and the physical exertion bring our brains to another level of functioning.
Cognitive systems are defined through their ability to adapt to the ever-changing conditions of the environment. We observe such processes of adaptation on many different levels: from single cells and tissues, through regulation systems such as the immune system, to people operating on conscious level and whole social groups. The aim of the conference is to present recent research on cognitive systems from multitude of perspectives and methodologies, and to bridge gaps between different disciplines in order to develop common language as well as new methods and paradigms for modeling cognitive phenomena.
Among the participants of the previous editions of the conference were psychologists, cognitive scientists, computer scientists, linguists, physicists, biologists, mathematicians, anthropologists, sociologists (see previous abstract book)… No matter what is your discipline, what counts is your interest in cognitive systems, your openness to talk to others and your love for the mountains.
One would think that if a domain defines itself with a certain concept, this concept should be understandable in a roughly uniform way by the sub-disciplines, creating a common ground rather than instigating lively debates. Cognitive science is a curious discipline: it is often defined as a study of how systems get, process and use information. Yet the concept of information still, after more than half century of the field’s successful development, in important aspects remains vague.
Shannon’s information theory proposes how to measure the amount of information based on the difference in states that we consider informative. However, what should count as a difference in states (a “difference that makes a difference”) is far from clear, both on the theoretical and operational level. What is it that the intelligent systems “pick up” from the environment, what does it mean that organisms “process” it? What are the time-scales of relevant events? What are the bits (information units) and how the “bits” cohere in complex flexible arrangements? One of the central questions is if these arrangements are due to the individual and internal brain processes or rather the information around is already rich, in some sense “preprocessed” by the virtue of how our bodies are built and by the construction of the environmental (physical and social) niches around the organisms.
Thus this year’s topic of the 7th Peripatetic Conference revolves around “information” and “information around” as the unifying thread of research on various levels of cognitive systems’ organization: from molecular to neural to social. Even though a certain prominence will be given to language and communicative processes, the above questions emerge at any of those levels. In many current research domains within the cognitive science they seem to be backgrounded or avoided entirely for pragmatic reasons, under the assumption that the established methodologies and operationalization can be relied on. We hope that the informal nature of the Peripatetic conference will create an atmosphere where even such basic questions can be boldly asked, answers proposed and searched for together, and lack of answers admitted.
Thea Cameron-Faulkner, PhD, CPsychol., Senior Lecturer in Linguistics, University of Manchester, Manchester, UK
Thea is an interdisciplinary researcher working in the area of infant development and has a passion for naturalistic data. Over the years she has spent more hours than she cares to remember collecting and analysing audio and video data of naturally occurring parent-child interaction. This experience has led her to adopt a process relational, ecological, dynamic (this list is not exhaustive…) account of infant development and unleashed the energy for discussion of exciting theoretical approaches (see, e.g., her paper in Infancy).
Angelo Cangelosi, PhD, Professor of Machine Learning and Robotics, University of Manchester, Manchester, UK
Angelo Cangelosi's main research expertise is on language grounding and embodiment in humanoid robots, developmental robotics, human-robot interaction, and on the application of neuromorphic systems for robot learning. He has produced more than 250 scientific publications, and has been general/bridging chair of numerous workshops and conferences including the IEEE ICDL-EpiRob Conferences (Frankfurt 2011, Osaka 2013, Lisbon 2017, Tokyo 2018). His latest book “Developmental Robotics: From Babies to Robots” (MIT Press; co-authored with Matt Schlesinger) was published in January 2015, and recently translated in Chinese and Japanese.
Bruno Galantucci, PhD, associate professor, John Cabot University, Rome, Italy
Bruno Galantucci is an Associate Professor in the Department of Economics and Social Sciences at John Cabot University. His domain of expertise is human communication, which he has investigated from a number of perspectives. His earlier research focused on the psychology of language, including speech perception, word recognition, and sentence processing. He then focused on the emergence and evolution of other forms of human communication, contributing to the development of a new scientific discipline called Experimental Semiotics. Currently, his research interests have extended to the scientific investigation of everyday acts of communication. If you want to know more about him, please check out his personal website.
Ingar Brinck, PhD, Professor of Theoretical Philosophy, Lund University, Lund, Sweden
Interdisciplinary research on embodied and embedded dynamic cognition from the developmental, philosophical or evolutionary perspective. Currently working on cognition, making, and improvisation in the arts and crafts (research grant SAB17/RJ). Publications in journals such as Cognitive Processing, Phenomenology and the Cognitive Sciences, Theory & Psychology, Infant and Child Development, Behavioral and Brain Sciences, on intersubjectivity and the We, social norms, social robots, cooperation, joint attention and pointing, metacognition, change blindness. One monograph, The Indexical ’I’ (Kluwer adademic publ. 1997). Referee for Plos One, Consciousness Studies. On the editorial board of Theoria.
Marek Kuś, PhD, full professor, Center for Theoretical Physics of the Polish Academy of Sciences, Warsaw, Poland
Talk: Quantum vs. Classical Machine Learning
An unparalleled development of quantum information theory we have witnessed for nearly twenty five years has not lead, up to now, to groundbreaking achievements in the area where promises were most attractive, namely quantum computing. Employing principles of quantum physics to perform information-theoretic task can, in principle, revolutionize computing making it faster. Alas! Even putting aside technical problems of constructing a quantum computer we are confronted with a modest (to put it mildly) number of interesting quantum algorithms. (In fact, only two such algorithms can be honestly quoted: the quantum Fast Fourier Transform that is a basis of the Shore algorithm potentially useful in breaking almost all public-key cryptography systems and the Grover algorithm speeding up databases searches). Recently however various ideas using quantum principles to achieve such tasks like machine learning or finding patterns in huge amounts of data were proposed. In my short talk I will address these topics and prospects of attacking such an ambitious tasks by quantum physical methods.
Aleksandra Przegalińska-Skierkowska, PhD, assistant professor, Koźmiński University, Warsaw, Poland
In the shades of the uncanny valley: An experimental study of human–chatbot interaction
Talk: Human-chatbot interaction
Chatbots are increasingly frequently used in business to facilitate various processes, particularly those related to customer service and personalization. In this paper we are proposing novel methods of tracking human-chatbot interactions and measuring chatbot performance that take into consideration ethical concerns. Our proposed methodology links neuroscientific methods, text mining and deep learning with the issue of trust allocated in chatbots and their overall transparency. Moreover, we argue that trust is the focal point of successful human-chatbot interaction and assess how trust as a category is redefined with the advent of deep-learning supported chatbots. We have previously conducted the first of its kind study of human-chatbot online interaction (Ciechanowski et al. 2018; Ciechanowski, Przegalinska, and Wegner 2018), using both subjective (questionnaires) and objective (psychophysiology) measures. In this article we propose a novel method of analysing the content of messages produced in human-chatbot interactions, using the Condor Tribefinder system developed by us for text-mining (De Oliveira and Gloor 2018; Gloor et al., n.d.), which is based on machine learning classification engine. Our results may be of significant use in the domain of building better social bots for interaction in business or commercial environments.
The conference will take place on 18-21 October 2018 in pension “Gościniec nad potokiem” in Małe Ciche near Zakopane. We start on Thursday with a dinner and an evening session. We finish all official sessions on Saturday, Sunday is reserved solely for peripatetic activity.
Conference fee: 50 PLN (students and PhD students), 150 PLN (researchers)
Costs of accomodation (stay + meals): approx 90 PLN a day (depends on meal options)
Conference program consists of talks given by participants and discussions in subgroups during walks in the nearby mountain valleys (peripatetic part, at least 4-5 hours trips). Important: Take your weather jacket and trekking boots!
Getting to the venue from Zakopane (by bus):
Take a bus from the stop “Plac PKP” (near the railway station, usually from “stanowisko 4”) with the destination “Małe Ciche”. Regular buses operate between 7AM and 8PM, there is a bus every 1 hour (approximately).
“Gościniec nad potokiem”
Małe Ciche 32 B
Come and share your research results, doubts, questions, ideas, collaboration opportunities. During the stationary meetings and peripatetic ones (from Ancient Greek peripatētikós – “given to walking around”) senior researchers, post-docs, graduate and undergraduate students share their knowledge and draw inspirations for further work.
We invite talks and posters in any field concerning (broadly understood) cognition and ways of its modeling. Just to give some examples, your talk may concern:
We all are active participants on this conference. There are three possible types of participation:
If you are an undergraduate student, and are just starting in the field, you may present either a poster (concerning your ongoing research or an idea for a future project) or a short book review. Some suggestions for books are here, but if you do not find anything interesting there you can suggest your own book to review (be sure to have it approved by the organizers though). Book review should focus on your own impressions and questions inspired by the lecture.
Abstract submission is closed. Here is the registration form for accepted participants.
Joanna Rączaszek-Leonardi, program chair
Agnieszka Kulesza, local arrangements
Ewa Machnacz, local arrangements
Konrad Zieliński, local arrangements
Julian Zubek, website