Learning Tractable Probabilistic Models for Fault Localization

TL;DR

This paper introduces Tractable Fault Localization Models (TFLMs), which learn from data across multiple buggy programs to improve bug localization by modeling dependencies between code lines, outperforming existing methods.

Contribution

The paper proposes TFLMs that leverage recent tractable probabilistic models to generalize fault localization across programs, incorporating multiple features and dependencies.

Findings

TFLMs outperform previous statistical debugging methods.

Incorporating TARANTULA scores improves bug localization.

TFLMs effectively model dependencies between code lines.

Abstract

In recent years, several probabilistic techniques have been applied to various debugging problems. However, most existing probabilistic debugging systems use relatively simple statistical models, and fail to generalize across multiple programs. In this work, we propose Tractable Fault Localization Models (TFLMs) that can be learned from data, and probabilistically infer the location of the bug. While most previous statistical debugging methods generalize over many executions of a single program, TFLMs are trained on a corpus of previously seen buggy programs, and learn to identify recurring patterns of bugs. Widely-used fault localization techniques such as TARANTULA evaluate the suspiciousness of each line in isolation; in contrast, a TFLM defines a joint probability distribution over buggy indicator variables for each line. Joint distributions with rich dependency structure are often computationally intractable; TFLMs avoid this by exploiting recent developments in tractable probabilistic models (specifically, Relational SPNs). Further, TFLMs can incorporate additional sources of information, including coverage-based features such as TARANTULA. We evaluate the fault localization performance of TFLMs that include TARANTULA scores as features in the probabilistic model. Our study shows that the learned TFLMs isolate bugs more effectively than previous statistical methods or using TARANTULA directly.