Software systems succeed or fail based on their alignment with stakeholder expectations. Yet, while verification rigorously checks conformance to \emph{specifications}, the crucial validation of implicit, nuanced stakeholder expectations remains a pre-formal, ad-hoc activity. This creates a critical validation gap: systems can be formally correct yet functionally misaligned, leading to integration failures and poor user experience. This paper introduces the core idea of Semantic Expectation Logic (SEL), a novel formalism that \textbf{elevates stakeholder expectations to become the central, formally-reasoned artifact in the validation process}. SEL provides a logic and framework to systematically elicit, model, and quantify the alignment between what stakeholders expect, what the domain requires, and what the system delivers. Through a motivating case study on REST API conventions, we demonstrate how this expectation-centric approach uncovers validation failures that specification-based methods cannot see. SEL charts a new course for validation, reframing it as a quantifiable engineering discipline of expectation alignment.
Program Display Configuration
Fri 17 Apr
Displayed time zone: Brasilia, Distrito Federal, Brazilchange
Zehong Yu KLISS, BNRist, School of Software, Tsinghua University, Zhuo Su Beihang University, Rui Wang Capital Normal University, Beijing, China, Yu Jiang Tsinghua University
Tao Liu Harbin Institute of Technology, Shenzhen; PengCheng Laboratory;, Zhao Liu, Ning Hu PengCheng Laboratory, Jingrun Liu Harbin Institute of Technology, Shenzhen, Binxing Fang Harbin Institute of Technology, Shenzhen, Qing Liao Harbin Institute of Technology