This project aims at developing novel visualisation concepts that facilitate human-centred explanation of formal explications of logic-based approaches. We will conduct visualisation research for formal methods and symbolic AI including the visualisation of a) ontology repair and consequence explications using Description Logics and b) probabilistic model checking. We will work towards our vision of a modular, generalised toolbox of visual model explanations informing future visualisation projects. We will also investigate adaptive visualisations as the core of developing a responsive visualisation framework that allows visual representations to react to various stakeholders, workflows and device contexts, and supporting multiple users in their collaborative work with model visualisations.
Previous Work (2019-2022)
Beside the common textual representation, graph visualisation present both structural information and multivariate data attributes as a basis for improved understanding and explanation of models. This project aims at enabling the exploration and refinement of models at both design and inspection time through interactive visualisation. As a concrete starting point, we will make contributions to the visual analysis of behaviour, causality and compositionality, as well as visualise and facilitate the explanation of consequences as a result of reasoning in knowledge representations using description logics. We will investigate interactive model visualisation in future office setups and multi-display environments (MDE) with novel input and output technologies to improve the exploration process for multiple users and tasks. In the first funding period, we identify, design and evaluate suitable representation of models in different complexity and size and develop novel interaction techniques to explore these models appropriate for the current user, task, context, and display setup. Our focus lies in creating responsive graph visualisations to fit these conditions and developing a framework for automating this adaptation to enable an improved exploration process.