Simultaneous Isolation of Software Faults for Effective Fault Localization

摘要

Due to fault-to-failure complexity in the existence of multiple faults, debugging faults is extremely hard. Many studies were done to improve localization effectiveness in the existence of multiple faults. Some studies attempt to isolate faults into separate fault-focused clusters that target single faults. However, isolating failures to their causative faults is still an issue and needs improvement. In this paper, we propose the use of a network community clustering algorithm to isolate faults into separate fault-focused communities, each targeting a single fault. These fault-focused communities will be given to developers to debug the faults simultaneously in parallel. The method is evaluated on 5 well-known multiple-fault subject programs from the Siemens test suite benchmark. The experimental results show that the network community clustering algorithm is relatively effective in isolating different faults into distinct fault-focused communities with improvements in faults localization effectiveness. The result also shows improvement in terms of reducing the expense to produce a failure-free program.