Combining spreadsheet smells for improved fault prediction

Patrick Koch, Konstantin Schekotihin, Dietmar Jannach, Birgit Hofer, Franz Wotawa, Thomas Schmitz

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

Spreadsheets are commonly used in organizations as a programming tool for business-related calculations and decision making. Since faults in spreadsheets can have severe business impacts, a number of approaches from general software engineering have been applied to spreadsheets in recent years, among them the concept of code smells. Smells can in particular be used for the task of fault prediction. An analysis of existing spreadsheet smells, however, revealed that the predictive power of individual smells can be limited. In this work we therefore propose a machine learning based approach which combines the predictions of individual smells by using an AdaBoost ensemble classifier. Experiments on two public datasets containing real-world spreadsheet faults show significant improvements in terms of fault prediction accuracy.

Original languageEnglish
Title of host publicationProceedings 2018 ACM/IEEE 40th International Conference on Software Engineering
Subtitle of host publicationNew Ideas and Emerging Results, ICSE-NIER 2018
PublisherIEEE Computer Society, 1998
Pages25-28
Number of pages4
ISBN (Electronic)9781450356626
DOIs
Publication statusPublished - 27 May 2018
Event40th ACM/IEEE International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018 - Gothenburg, Sweden
Duration: 30 May 20181 Jun 2018

Conference

Conference40th ACM/IEEE International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018
CountrySweden
CityGothenburg
Period30/05/181/06/18

Fingerprint

Spreadsheets
Adaptive boosting
Learning systems
Software engineering
Industry
Classifiers
Decision making
Experiments

Keywords

  • Fault Prediction
  • Spreadsheet QA
  • Spreadsheet Smells

ASJC Scopus subject areas

  • Software

Cite this

Koch, P., Schekotihin, K., Jannach, D., Hofer, B., Wotawa, F., & Schmitz, T. (2018). Combining spreadsheet smells for improved fault prediction. In Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018 (pp. 25-28). IEEE Computer Society, 1998. https://doi.org/10.1145/3183399.3183402

Combining spreadsheet smells for improved fault prediction. / Koch, Patrick; Schekotihin, Konstantin; Jannach, Dietmar; Hofer, Birgit; Wotawa, Franz; Schmitz, Thomas.

Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018. IEEE Computer Society, 1998, 2018. p. 25-28.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Koch, P, Schekotihin, K, Jannach, D, Hofer, B, Wotawa, F & Schmitz, T 2018, Combining spreadsheet smells for improved fault prediction. in Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018. IEEE Computer Society, 1998, pp. 25-28, 40th ACM/IEEE International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018, Gothenburg, Sweden, 30/05/18. https://doi.org/10.1145/3183399.3183402
Koch P, Schekotihin K, Jannach D, Hofer B, Wotawa F, Schmitz T. Combining spreadsheet smells for improved fault prediction. In Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018. IEEE Computer Society, 1998. 2018. p. 25-28 https://doi.org/10.1145/3183399.3183402
Koch, Patrick ; Schekotihin, Konstantin ; Jannach, Dietmar ; Hofer, Birgit ; Wotawa, Franz ; Schmitz, Thomas. / Combining spreadsheet smells for improved fault prediction. Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2018. IEEE Computer Society, 1998, 2018. pp. 25-28
@inproceedings{fa2a308764c84143bb9cad975e405840,
title = "Combining spreadsheet smells for improved fault prediction",
abstract = "Spreadsheets are commonly used in organizations as a programming tool for business-related calculations and decision making. Since faults in spreadsheets can have severe business impacts, a number of approaches from general software engineering have been applied to spreadsheets in recent years, among them the concept of code smells. Smells can in particular be used for the task of fault prediction. An analysis of existing spreadsheet smells, however, revealed that the predictive power of individual smells can be limited. In this work we therefore propose a machine learning based approach which combines the predictions of individual smells by using an AdaBoost ensemble classifier. Experiments on two public datasets containing real-world spreadsheet faults show significant improvements in terms of fault prediction accuracy.",
keywords = "Fault Prediction, Spreadsheet QA, Spreadsheet Smells",
author = "Patrick Koch and Konstantin Schekotihin and Dietmar Jannach and Birgit Hofer and Franz Wotawa and Thomas Schmitz",
year = "2018",
month = "5",
day = "27",
doi = "10.1145/3183399.3183402",
language = "English",
pages = "25--28",
booktitle = "Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering",
publisher = "IEEE Computer Society, 1998",

}

TY - GEN

T1 - Combining spreadsheet smells for improved fault prediction

AU - Koch, Patrick

AU - Schekotihin, Konstantin

AU - Jannach, Dietmar

AU - Hofer, Birgit

AU - Wotawa, Franz

AU - Schmitz, Thomas

PY - 2018/5/27

Y1 - 2018/5/27

N2 - Spreadsheets are commonly used in organizations as a programming tool for business-related calculations and decision making. Since faults in spreadsheets can have severe business impacts, a number of approaches from general software engineering have been applied to spreadsheets in recent years, among them the concept of code smells. Smells can in particular be used for the task of fault prediction. An analysis of existing spreadsheet smells, however, revealed that the predictive power of individual smells can be limited. In this work we therefore propose a machine learning based approach which combines the predictions of individual smells by using an AdaBoost ensemble classifier. Experiments on two public datasets containing real-world spreadsheet faults show significant improvements in terms of fault prediction accuracy.

AB - Spreadsheets are commonly used in organizations as a programming tool for business-related calculations and decision making. Since faults in spreadsheets can have severe business impacts, a number of approaches from general software engineering have been applied to spreadsheets in recent years, among them the concept of code smells. Smells can in particular be used for the task of fault prediction. An analysis of existing spreadsheet smells, however, revealed that the predictive power of individual smells can be limited. In this work we therefore propose a machine learning based approach which combines the predictions of individual smells by using an AdaBoost ensemble classifier. Experiments on two public datasets containing real-world spreadsheet faults show significant improvements in terms of fault prediction accuracy.

KW - Fault Prediction

KW - Spreadsheet QA

KW - Spreadsheet Smells

UR - http://www.scopus.com/inward/record.url?scp=85049772205&partnerID=8YFLogxK

U2 - 10.1145/3183399.3183402

DO - 10.1145/3183399.3183402

M3 - Conference contribution

SP - 25

EP - 28

BT - Proceedings 2018 ACM/IEEE 40th International Conference on Software Engineering

PB - IEEE Computer Society, 1998

ER -