Using Dynamic Knowledge Hypergraphs Toward Proactive AlOps through Digital Twins.

The necessity of minimizing the disruption of offered software­based products and services is becoming progressively prominent in IT operations to maximize quality of service and improve per­ceived value. Nevertheless, the increased operational complexities and uncertainties of modern IT systems overwhelm human compe­tencies and challenge the effectiveness of traditional monitoring systems. Consequently, guaranteeing the expected performance ofIT systems demands augmented autonomic (self-management) capabilities. The introduction and continuous advancement of arti­ficial intelligence for IT operations (i.e., AIOPs) represent a break­through in this regard. Moreover, the integration of AIOPs with the revolutionary concept of digital twins (DTs) might signify a step forward toward the achievement of proactive, AI-led IT op­erations. However, the definition of appropriate mechanisms and infrastructures to realize this integration remains an unexplored field, limiting anticipated benefits and blurring its potential impact. Especially, the modelling of DTs for AIOPs, and in general for soft­ware systems, is an emerging research area yet to be exploited. This short paper describes our vision for applying dynamic knowledge hypergraphs as an instrument to capture and utilize OT-driven information to improve risk mitigation, root cause analysis, and problem resolution in the proactive AIOPs realm.