Tutorial on Cognitive Logics:
Mechanisms Predicting Human Inference
Systems and methods for Artificial Intelligence (AI) applicable in the real
world require to represent and reason about uncertain knowledge. While
this is a limitation of classical first-order logic, there is a large number of
so-called non-monotonic logics, i.e., logics that aim to draw inferences only
cautiously, allowing for revising them if new information becomes available.
Cognitive analysis have shown that human inferential behavior can be better
described employing such logics.
In this tutorial we introduce the cognitive and formal foundations of cognitive logics based on nonmonotonic logics, relevant benchmark problems
from psychology that are dealt with by formal AI methods, and current
challenges in modeling cognitive reasoning. The tutorial addresses a joint
view on characteristics of human reasoning from the perspective of computer
science and cognitive science.
The tutorial will be held at the 24th European Conference on Artificial Intelligence (ECAI 2020),
which takes place on August 29-September 2 in Santiago de Compostela, Spain.
Presenters: Marco Ragni, Kai Sauerwald and Gabriele Kern-Isberner
Length of the Tutorial: 90 minutes (spotlight tutorial)
Outline of the Tutorial
Topics to be dealt with in the tutorial:
- Introduction to nonmonotonic and defeasible reasoning from a commonsense perspective
- Formal aspects of Cognitive Logics
- Nonmonotonic logics, formal models of commonsense reasoning
- Cognitive aspects of Cognitive Logics
- Cognitive theories and principles, benchmark examples
- Cognitive modelling of human reasoning
- Current challenge: syllogistic reasoning
- Logical incorrectness of human reasoning and how to escape the criticism of irrationality
- The role of background knowledge in psychological experiments
Target audience and prerequisite knowledge
Any AI researcher interested in the fields of knowledge representation and
reasoning, data science, logic, uncertain reasoning, commonsense sense rea-
soning, and their relationships to cognition and cognitive psychology may
benefit from this tutorial.
Only basic knowledge of propositional and first-order logic is required to
follow the tutorial.
, Department of Computer Science, Technical Faculty, University of Freiburg, 79110 Freiburg, Germany.
Marco Ragni revceived his PhD in Artificial Intelligence in 2008 from the
Technical Faculty and his PhD in Cognitive Science in 2013 in Cognitive
Science from the Center for Cognitive Science at the University Freiburg.
He received his habilitations in Computer Science in 2014 and in Cognitive
Science and General Psychology in 2015, and is now an Associate Professor
(apl. Prof.) at the Technical Faculty of the University Freiburg and a DFG-Heisenbergfellow. His research interests focus on computational models of high-level cognition, both from a cognitive, computational, and neuroscience perspective.
, Department of Computer Science, TU Dortmund, 44227 Dortmund, Germany.
Gabriele Kern-Isberner received her diploma in mathematics in 1979, and
her doctoral degree in mathematics in 1985, both from the University of
Dortmund. In 2000, she did her habilitation in computer science at the
FernUniversität in Hagen, the German Open University, and got the Venia
legendi for computer science. She worked as a research assistant and as a
lecturer at the universities of Dortmund, Hagen, and Leipzig. Since 2004,
she has been a Professor for Information Engineering at the department of
computer science at the University of Technology Dortmund.
Her scientific work focuses on qualitative and quantitative approaches to
knowledge representation such as default and non-monotonic logics, uncertain reasoning, belief revision, and argumentation. Her research interests
include in particular the development of methods that help integrate approaches from different fields, such as the combination of first-order logic and
probabilities, or building bridges between uncertain reasoning and learning.
Some of her works also deal with the cognitive aspects of formal reasoning
models. She has been involved in the organization of major conferences in
AI, was co-chair of ECSQARU 2019, was co-chair of FoIKS 2020, and she
currently co-chairs the steering committee of NMR workshops.
, Knowledge Based Systems, Faculty of Mathematics and Computer Science, FernUniversität in Hagen, 58097 Hagen, Germany.
Kai Sauerwald recived his master degree in computer science from the TU Dortmund, and is currently a resarcher at the Knowledge Base Systems group at the FernUniversität in Hagen.
Recently he was one of the two local-organisers of FoIKS 2020.
His research interests focus on belief change and its applications and connections to other areas from knowledge representation and psychology.