CiteULike Search Register Log in
CiteULike is a free online bibliography manager. Register and you can start organising your references online.

Statistical Debugging Using Latent Topic Models

by: David Andrzejewski, Anne Mulhern, Ben Liblit, Xiaojin Zhu
Machine Learning: ECML 2007 (2007), pp. 6-17.

X Abstract

Statistical debugging uses machine learning to model program failures and help identify root causes of bugs. We approach this task using a novel Delta-Latent-Dirichlet-Allocation model. We model execution traces attributed to failed runs of a program as being generated by two types of latent topics: normal usage topics and bug topics. Execution traces attributed to successful runs of the same program, however, are modeled by usage topics only. Joint modeling of both kinds of traces allows us to identify weak bug topics that would otherwise remain undetected. We perform model inference with collapsed Gibbs sampling. In quantitative evaluations on four real programs, our model produces bug topics highly correlated to the true bugs, as measured by the Rand index. Qualitative evaluation by domain experts suggests that our model outperforms existing statistical methods for bug cause identification, and may help support other software tasks not addressed by earlier models.

View the full article here:

DOI

This article has been bookmarked 2 times, initially on 2008-05-21.

2008-10-07 User ldietz , 1 note

"""Where possible, we compare LDA results with corresponding measures from two earlier statistical debugging techniques. PLDI05 refers to earlier work by Liblit et al. [5] that uses an iterative process of selecting and eliminating top-ranked predicates until all failures are explained. The approach bears some resemblance to likelihood ratio testing and biased minimum-set cover problems, but is somewhat ad hoc and highly specialized for debugging. ICML06 refers to earlier work by Zheng et al. [8] that takes inspiration from bi-clustering algorithms. This approach uses graphical models to estimate complete (non-sampled) counts, then applies an iterative collective voting scheme followed by a simple clustering pass to identify and report likely bug causes."""

2008-10-07 15:42:44
2008-05-21 User V
Privacy Statement | Terms & Conditions
CiteULike organises scholarly (or academic) papers or literature and provides bibliographic (which means it makes bibliographies) for universities and higher education establishments. It helps undergraduates and postgraduates. People studying for PhDs or in postdoctoral (postdoc) positions. The service is similar in scope to EndNote or RefWorks or any other reference manager like BibTeX, but it is a social bookmarking service for scientists and humanities researchers.