We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.SE

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Software Engineering

Title: Technical Reports Compilation: Detecting the Fire Drill anti-pattern using Source Code

Abstract: Detecting the presence of project management anti-patterns (AP) currently requires experts on the matter and is an expensive endeavor. Worse, experts may introduce their individual subjectivity or bias. Using the Fire Drill AP, we first introduce a novel way to translate descriptions into detectable AP that are comprised of arbitrary metrics and events such as maintenance activities, which are mined from the underlying source code, thus making the description objective as it becomes data-based. Secondly, we demonstrate a novel method to quantify and score the deviations of real-world projects to data-based AP descriptions. Using nine real-world projects that exhibit a Fire Drill to some degree, we show how to further enhance the translated AP. The ground truth in these projects was extracted from two individual experts and consensus was found between them. Our evaluation spans three kinds of pattern, where the first is purely derived from description, the second type is enhanced by data, and the third kind is derived from data only. The Fire Drill AP as translated from description only shows weak potential of confidently detecting the presence of the anti-pattern in a project. Enriching the AP with data from real-world projects significantly improves the detection. Using patterns derived from data only leads to almost perfect correlations of the scores with the ground truth. Some APs share symptoms with the Fire Drill AP, and we conclude that the presence of similar patterns is most certainly detectable. Furthermore, any pattern that can be characteristically modelled using the proposed approach is potentially well detectable.
Comments: 132 pages
Subjects: Software Engineering (cs.SE); Computation (stat.CO); Machine Learning (stat.ML)
ACM classes: D.2; G.3; I.5; I.6
DOI: 10.13140/RG.2.2.35805.33766/1
Cite as: arXiv:2104.15090 [cs.SE]
  (or arXiv:2104.15090v2 [cs.SE] for this version)

Submission history

From: Sebastian Hönel [view email]
[v1] Fri, 30 Apr 2021 16:16:32 GMT (941kb,D)
[v2] Mon, 3 May 2021 14:52:43 GMT (949kb,D)

Link back to: arXiv, form interface, contact.