Informatica

Due to the complexity and lack of transparency of recent advances in artificial intelligence, Explainable AI (XAI) emerged as a solution to enable the development of causal image-based models. This study examines shadow detection across several fields, including computer vision and visual effects. Three-fold approaches were used to construct a diverse dataset, integrate structural causal models with shadow detection, and apply interventions simultaneously for detection and inferences. While confounding factors have only a minimal impact on cause identification, this study illustrates how shadow detection enhances understanding of both causal inference and confounding variables.

The paper deals with the causality perspective of the Enterprise Architecture (EA) frameworks. The analysis showed that there is a gap between the capabilities of EA frameworks and the behavioural characteristics of the real world domain (enterprise management activities). The contribution of research is bridging the gap between enterprise domain knowledge and EA framework content by the integration of meta-models as part of EA structures. Meta-models that cover not only simple process flows, but also business behaviour, i.e. causality of the domain, have been developed. Meta-models enable to create a layer of knowledge in the EA framework, which ensures smart EA development, allows validation of developer decisions. Two levels of the enterprise causal modelling were obtained. The first level uses the Management Transaction (MT) framework. At the second level, deep knowledge was revealed using a framework called the Elementary Management Cycle (EMC). These two causal frameworks were applied here to justify the causal meta-models of the EA. The new concepts Collapsed Capability, Capability Type and Capability Role which meaningfully complement MODAF with causal knowledge are introduced. Strategic Viewpoint (StV) modelling using causal meta-models is described in detail and illustrated in the case study. The example provided shows a principled way that causal knowledge supports the verification and validation of EA solutions. The presented method provides an opportunity to move the EA development to smart platforms.