Artifact for Evaluating the Impact of Experimental Assumptions in Automated Fault Localization
This document describes the artifact for the ICSE ’23 paper titled “Evaluating the Impact of Experimental Assumptions in Automated Fault Localization”. The original paper studies the impact of three (3) major experimental assumptions commonly made by researchers when conducting debugging evaluation, namely (a) perfect bug understanding (PBU), (b) using fix location(s) as root cause bug diagnosis and (c) assuming a single fault location. This artifact includes the documentation of (1) the prevalence analysis of the studied assumptions in debugging literature and bug datasets, (2) the experiments evaluating the effect of these assumptions on the measured effectiveness of Automated Fault Localization (AFL) methods and Automated Program Repair (APR) tools, and (3) the user study examining the effect of the assumptions on the debugging productivity of developers. We provide technical information about each component of the artifact and describe the use of the artifact for future research.
Abstract for the Artifact (icse23-debugging-artifact-abstract.pdf) | 85KiB |