TITLE: An Analysis of the Relationship between Conditional Entropy and Failed Error Propagation in Software Testing
SPEAKER: David Clark (UCL)

ABSTRACT:

Failed error propagation is known to hamper software testing, yet it remains poorly understood.
We introduce an information theoretic formulation of failed error propagation that computes approximate measures of conditional entropy associated with statements and paths.
We demonstrate, experimentally, that our metrics are well-correlated with failed error propagation.
Our empirical study involved 20 programs, for which we executed a total of 7,140,000 test cases.
The results reveal a Spearman rank correlation to failed error propagation 
of 0.699 for conditional entropy associated with a statement and 0.956 for conditional entropy associated with a program path.