Session Description

This special session deals with the problem of developing approaches supporting cognitive situation management. Since intelligent behavior is inherently situational, intelligent artificial agents and systems need to obtain situation awareness and expose situated behavior. Recent advancements in machine learning and artificial intelligence (AI) have mainly focused on models of computational perception of individual local situations, achieving major breakthroughs in the automated detection of individual objects and entities from heterogeneous sources such as images, audio, video, and natural language. Hence, the upcoming challenge is how we can build intelligent agents capable to comprehend the overall observed situation(s), by interrelating these individual situations. As opposed to the clearly defined object recognition tasks, comprehension of situations represents a complex task of fusion of relational, spatial and temporal information, potentially emerging from the aggregation of multiple agents’ views. These challenging characteristics thus call for novel and potentially synergistic approaches from both sub-symbolic and symbolic AI, as well as inspirations from the models of collective (swarm) intelligence in biological agents developed in cognitive science.

Therefore, the goal is to bring together researchers from diverse fields to exchange on recent developments towards cognitive situation management, situation understanding, situation learning and knowledge acquisition, and situated interaction.