Title of article :
Enhancing software artefact traceability recovery processes with link count information
Author/Authors :
Bavota، نويسنده , , Gabriele and De Lucia، نويسنده , , Andrea and Oliveto، نويسنده , , Rocco and Tortora، نويسنده , , Genoveffa، نويسنده ,
Issue Information :
ماهنامه با شماره پیاپی سال 2014
Pages :
20
From page :
163
To page :
182
Abstract :
AbstractContext tensive human effort needed to manually manage traceability information has increased the interest in using semi-automated traceability recovery techniques. In particular, Information Retrieval (IR) techniques have been largely employed in the last ten years to partially automate the traceability recovery process. us studies mainly focused on the analysis of the performances of IR-based traceability recovery methods and several enhancing strategies have been proposed to improve their accuracy. Very few papers investigate how developers (i) use IR-based traceability recovery tools and (ii) analyse the list of suggested links to validate correct links or discard false positives. We focus on this issue and suggest exploiting link count information in IR-based traceability recovery tools to improve the performances of the developers during a traceability recovery process. pirical studies have been conducted to evaluate the usefulness of link count information. The two studies involved 135 University students that had to perform (with and without link count information) traceability recovery tasks on two software project repositories. Then, we evaluated the quality of the recovered traceability links in terms of links correctly and erroneously traced by the students. s sults achieved indicate that the use of link count information significantly increases the number of correct links identified by the participants. sions sults can be used to derive guidelines on how to effectively use traceability recovery approaches and tools proposed in the literature.
Keywords :
Traceability recovery , information retrieval , Link count information , CONTROLLED EXPERIMENTS
Journal title :
Information and Software Technology
Serial Year :
2014
Journal title :
Information and Software Technology
Record number :
2375200
Link To Document :
بازگشت