In this paper, we propose a technique for leveraging historical field failure records in conjunction with automated static analysis alerts to determine which alerts or sets of alerts are predictive of a field failure. Our technique uses singular value decomposition to generate groupings of static analysis alert types, which we call alert signatures, that have been historically linked to field failure-prone files in previous releases of a software system. The signatures can be applied to sets of alerts from a current build of a software system. Files that have a matching alert signature are identified as having similar static analysis alert characteristics to files with known field failures in a previous release of the system. We performed a case study involving an industrial software system at IBM and found three distinct alert signatures that could be applied to the system. We found that 50% of the field failures reported since the last static analysis run could be discovered by examining the 10% of the files and static analysis alerts indicated by these three alert signatures. The remaining failures were either not detected by a signature which could be an indication of a new type of error in the field, or they were on areas of the code where no static analysis alerts were detected.

We present prediction by grammatical match (PGM), a new general purpose adaptive text compression framework successfully blending finite-context and general context-free models. A PGM compressor operates incrementally by parsing a prefix of the input text, generating a set of analyses; these analyses are scored according to encoding cost, the cheapest is selected, and sent through an order k PPM encoder. PGM's primary innovations include the use of a generalized PPM in selection and coding; the simultaneous use of multiple context-free grammars; the use of lexical left-corner derivations (LLCD); and an aggressive algorithm for constructing an LR(O) parsable metalanguage for LLCD. LLCD are a hybrid of bottom-up and top-down descriptions that represent grammatical information implicitly with each lexeme. The constructed metalanguage extends this to include explicit top-down steps to resolve local ambiguities in at most one strictly grammatical symbol. These properties combine to deliver excellent compression. On a test corpus of about 1 MB of Scheme program text, PGM with a generic Scheme grammar required about 26% fewer bits than PPM to represent the entire corpus, with reductions on individual files reaching as high as 55%. In addition, PGM enriches the time-compression-memory tradeoff options, since a low order PGM can achieve bpc rates comparable to high order PPM at considerable savings in space. PGM compression operates in expected linear time and space for many kinds of grammars. PGM decompression operates in guaranteed linear time and space.