Beschreibung
In recent years, intelligent agents in the contexts of open environments and multi agent systems have become a leading paradigm in AI. Acting successfully in such environments that are uncertain, only partially accessible, and dynamic, requires sophisticated knowledge representation and reasoning techniques for the modelling of the epistemic state of the agent. In particular, in evolving environments, the agent must continuously react to new observations and to any unforeseen changes that occur. Its epistemic state must undergo corresponding changes to provide the agent with a suitable world view at any time. Thus, modern knowledge representation methods have to deal with the evolution of knowledge and belief, due to uncertain or incomplete information, or to changes in the environment.
This is the 3rd Workshop on "Dynamics of Knowledge and Belief" organized by the GI-Fachgrupppe "Wissensrepräsentation und Schließen", following two previous workshops at KI-2007 in Osnabrueck and at KI-2009 in Paderborn. The focus of the workshop is on any topics of knowledge representation and reasoning that address the epistemic modelling of agents in open environments, and in particular on processes concerning evolving knowledge and belief both in theory and in applications. We welcome papers on the following and any related topics:
- Knowledge representation in theory and practice
- Belief revision and belief update
- Nonmonotonic and uncertain reasoning
- Learning and knowledge discovery in data
- Argumentation theories
- Decision theory and preferences
- Ontologies and description logics
- Agents and multiagent systems
- Action and change
Programm
09:00 - 10:30 Modelling and Reasoning in Probabilistics
- On Prototypical Indifference and Lifted Inference in Relational Probabilistic Conditional Logic
Matthias Thimm - Knowledge Engineering with Markov Logic Networks: A Review
Dominik Jain - Analyzing Inconsistencies in Probabilistic Conditional Knowledge Bases using Continuous Inconsistency Measures
Matthias Thimm
10:30 -11:00 Coffee
11:00 - 12:30 Relational Probabilistic Learning
- Learning Scenarios under Relational Probabilistic Semantics and ME Reasoning
Marc Finthammer, Nico Potyka - Statistical Relational Learning in Dynamic Environments - An Agent-Based Approach to Traffic Navigation Using Bayesian Logic Networks
Daan Apeldoorn - On efficient algorithms for minimal ME-learning
Nico Potyka