Sökning: onr:"swepub:oai:DiVA.org:bth-26101" >
FedCSD :
FedCSD : A Federated Learning Based Approach for Code-Smell Detection
-
- Alawadi, Sadi, 1983- (författare)
- Blekinge Tekniska Högskola,Institutionen för datavetenskap
-
- Alkharabsheh, Khalid (författare)
- Al-Balqa Applied University, Jordan
-
- Alkhabbas, Fahed (författare)
- Malmö University, Internet of Things and People Research Center
-
visa fler...
-
- Kebande, Victor R. (författare)
- Blekinge Tekniska Högskola,Institutionen för datavetenskap
-
- Awaysheh, Feras M. (författare)
- Institute of Computer Science, Estonia
-
- Palomba, Fabio (författare)
- University of Salerno, Italy
-
- Awad, Mohammed (författare)
- Arab American University, Palestine
-
visa färre...
-
(creator_code:org_t)
- Institute of Electrical and Electronics Engineers (IEEE), 2024
- 2024
- Engelska.
-
Ingår i: IEEE Access. - : Institute of Electrical and Electronics Engineers (IEEE). - 2169-3536. ; 12, s. 44888-44904
- Relaterad länk:
-
https://doi.org/10.1...
-
visa fler...
-
https://bth.diva-por... (primary) (Raw object)
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
visa färre...
Abstract
Ämnesord
Stäng
- Software quality is critical, as low quality, or 'Code smell,' increases technical debt and maintenance costs. There is a timely need for a collaborative model that detects and manages code smells by learning from diverse and distributed data sources while respecting privacy and providing a scalable solution for continuously integrating new patterns and practices in code quality management. However, the current literature is still missing such capabilities. This paper addresses the previous challenges by proposing a Federated Learning Code Smell Detection (FedCSD) approach, specifically targeting 'God Class,' to enable organizations to train distributed ML models while safeguarding data privacy collaboratively. We conduct experiments using manually validated datasets to detect and analyze code smell scenarios to validate our approach. Experiment 1, a centralized training experiment, revealed varying accuracies across datasets, with dataset two achieving the lowest accuracy (92.30%) and datasets one and three achieving the highest (98.90% and 99.5%, respectively). Experiment 2, focusing on cross-evaluation, showed a significant drop in accuracy (lowest: 63.80%) when fewer smells were present in the training dataset, reflecting technical debt. Experiment 3 involved splitting the dataset across 10 companies, resulting in a global model accuracy of 98.34%, comparable to the centralized model's highest accuracy. The application of federated ML techniques demonstrates promising performance improvements in code-smell detection, benefiting both software developers and researchers. © 2013 IEEE.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap -- Programvaruteknik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Software Engineering (hsv//eng)
Nyckelord
- code smell detection
- federated learning
- privacy-preserving
- Software quality
- technical debit
- Application programs
- Codes (symbols)
- Computer software maintenance
- Computer software selection and evaluation
- Cryptography
- Object oriented programming
- Odors
- Code
- Code smell
- Ho-momorphic encryptions
- Homomorphic-encryptions
- Object oriented modelling
- Privacy preserving
- Data privacy
Publikations- och innehållstyp
- ref (ämneskategori)
- art (ämneskategori)
Hitta via bibliotek
Till lärosätets databas