Gerhard Brewka, Gerhard Lakemeyer
Hybrid Reasoning for Intelligent
Systems: A Focus of KR Research
For quite a while, research in knowledge representation
and reasoning (KR) had a strong focus on purely symbolic methods, for instance in subareas like description logics, logic programming and nonmonotonic reasoning,
belief revision, reasoning about action, and the like.
Although the motivation for studying symbolic methods
is still valid, it becomes more and more apparent that
numerical methods are needed to handle certain aspects
— such as time, resources, uncertainty, and preferences
— of realistic problems. To model realistic problems, one
has to enhance symbolic reasoning methods with capabilities to exploit and process numerical information. In
other words, one has to rely on hybrid methods that
combine qualitative and quantitative reasoning in one
way or the other.
With this in mind, a group of German KR researchers
decided to start a research initiative focusing on hybrid
reasoning. The primary funding agency for basic research
in Germany is the German Research Foundation (DFG).
One of DFG’s funding instruments are Research Units
(Forschergruppen in German), which consist of a collection of closely related projects, typically around five, that
investigate different aspects of a fundamental joint
research question (figure 1). Funding is granted for up to
six years, with an intermediate evaluation after the first
phase of three years.
The application for the Hybrid Reasoning for Intelli-
gent Systems (Hybris) Research Unit was successful and
Hybris began its work in 2012. Since then, Hybris has also
passed the intermediate evaluation and so funding will
be available until 2019. Seven principal investigators
from six different German universities are involved in
the current, second phase of the project: Franz Baader,
TU Dresden; Gerhard Brewka, Leipzig University; Wol-
fram Burgard, University of Freiburg; Gabriele Kern-
Isberner, TU Dortmund; Gerhard Lakemeyer, RWTH
Aachen (coordinator); Bernhard Nebel, University of
Copyright © 2017, Association for the Advancement of Artificial Intelligence. All rights reserved. ISSN 0738-4602
; This column describes the efforts of a
group of German knowledge representation and reasoning researchers to
pursue a research initiative focusing
on hybrid reasoning.