Interactive Bayesian program analysis enhances static analysis by modeling derivations as probabilistic dependencies, enabling ranking alarms by calculated confidences, proposing highly likely alarms for user inspection, and updating confidences with inspection results. Existing interactive approaches adopt a purely greedy, exploitation-only selection strategy that always inspects the highest-confidence alarm. However, such strategies are prone to local optima, leading to redundant inspections and delayed identification of true alarms.
We propose \textsc{Beer} (Bayesian Exploration-Exploitation Ranker), a framework that systematically integrates the Exploration-Exploitation trade-off into Bayesian program analysis. \textsc{Beer} leverages structural correlations between alarms—derived from shared root causes in the Bayesian model—to estimate expected information gain and guide exploration. When repeated false alarms indicate model stagnation, \textsc{Beer} selects alarms from minimally explored, highly correlated clusters to accelerate learning.
Implemented atop the \textsc{Bingo} framework, \textsc{Beer} achieves up to 32% effectiveness in ranking efficiency over the greedy baseline on datarace, thread-escape, and taint analyses, demonstrating the efficacy of exploration-guided alarm resolution.