abstract:
We introduce in this paper an itemset mining approach to tackle the fault localization problem, which is one of the most difficult processes in software debugging. We formalize the problem of fault localization as finding the k best patterns satisfying a set of constraints modelling the suspicious statements. We use a Constraint Programming(CP) approach to model and to solve our itemset based fault localization problem. Our approach consists of two steps: (i) mining top-k suspicious suites of statements; (ii) fault localization by processing top-k patterns. Experiments performed on standard benchmark programs show that our approach enables to propose a more precise localization than a standard approach.
ps. The second step aims at ranking in a more accurate way the whole top-k statements by taking benefit of two main observations:
(1) where faults are introduced in a program can be seen as a pattern (set of statements), which is more frequent in failing executions than passing ones;
(2) the difference beween a more suspicious pattern and a less suspicious one is a set of statements that appears/disappears in one or other; this difference helps us to know more about the location of the fault.
We have shown how these two properties can be exploited in an ad-hoc ranking algorithm producing accuate localization.
As future works, we plan to experiment our approach on programs with complex faults (more than one faulty statement). We also plan to explore other observations on the behavior of a faulty program and adding them as contraint for mining the location of faults.