Enhancing Network Diagnosis with Reflection in Prolog - Extended Abstract

This conference talk explores how reflection capabilities in Prolog can significantly improve network diagnosis processes, particularly for complex networks "in the wild" like the Internet, enterprise networks, and smart spaces. Learn how meta-programming techniques can help explain unintended network states and adapt reasoning processes based on contextual knowledge. The presentation demonstrates a practical implementation using Prolog's meta-interpreters as both a reflection engine and an injection point for new diagnostic functions. Delivered at the 27th International Symposium on Practical Aspects of Declarative Languages (PADL'25), this talk by Anduo Wang from Temple University addresses the gap between formal methods in clean-slate networks and the more challenging diagnostic needs of everyday network environments.