A large empirical assessment of the role of data balancing in machine-learning-based code smell detection

Journal of Systems and Software, Journal Year: 2020, Volume and Issue: 169, P. 110693 - 110693

Published: June 8, 2020

Language: Английский

---

Code smell detection using multi-label classification approach

Software Quality Journal, Journal Year: 2020, Volume and Issue: 28(3), P. 1063 - 1086

Published: April 4, 2020

Language: Английский

---

Code Smell Detection Using Ensemble Machine Learning Algorithms

Applied Sciences, Journal Year: 2022, Volume and Issue: 12(20), P. 10321 - 10321

Published: Oct. 13, 2022

Code smells are the result of not following software engineering principles during development, especially in design and coding phase. It leads to low maintainability. To evaluate quality its maintainability, code smell detection can be helpful. Many machine learning algorithms being used detect smells. In this study, we applied five ensemble two deep Four datasets were analyzed: Data class, God Feature-envy, Long-method datasets. previous works, stacking dataset results found acceptable, but there is scope improvement. A class balancing technique (SMOTE) was handle imbalance problem The Chi-square feature extraction select more relevant features each dataset. All obtained highest accuracy—100% for with different selected sets metrics, poorest accuracy, 91.45%, achieved by Max voting method Feature-envy twelve metrics.

Language: Английский

---

Leveraging Prompt-Based Large Language Models for Code Smell Detection: A Comparative Study on the MLCQ Dataset

Lecture notes on data engineering and communications technologies, Journal Year: 2025, Volume and Issue: unknown, P. 444 - 454

Published: Jan. 1, 2025

Language: Английский

---

Developer-Driven Code Smell Prioritization

Published: June 29, 2020

Code smells are symptoms of poor implementation choices applied during software evolution. While previous research has devoted effort in the definition automated solutions to detect them, still little is known on how support developers when prioritizing them. Some works attempted deliver that can rank smell instances based their severity, computed basis metrics. However, this may not be enough since it been shown recommendations provided by current approaches do take developer's perception design issues into account. In paper, we perform a first step toward concept developer-driven code prioritization and propose an approach machine learning able according perceived criticality assign We evaluate our technique empirical study investigate its accuracy features more relevant for classifying perception. Finally, compare with state-of-the-art technique. Key findings show solution F-Measure up 85% outperforms baseline approach.

Language: Английский

---

Improving performance with hybrid feature selection and ensemble machine learning techniques for code smell detection

Science of Computer Programming, Journal Year: 2021, Volume and Issue: 212, P. 102713 - 102713

Published: Aug. 17, 2021

Language: Английский

---

Predicting Code Smells and Analysis of Predictions: Using Machine Learning Techniques and Software Metrics

Journal of Computer Science and Technology, Journal Year: 2020, Volume and Issue: 35(6), P. 1428 - 1445

Published: Nov. 1, 2020

Language: Английский

---

A Novel Approach for Code Smell Detection: An Empirical Study

IEEE Access, Journal Year: 2021, Volume and Issue: 9, P. 162869 - 162883

Published: Jan. 1, 2021

Code smells detection helps in improving understandability and maintainability of software while reducing the chances system failure. In this study, six machine learning algorithms have been applied to predict code smells. For purpose, four smell datasets (God-class, Data-class, Feature-envy, Long-method) are considered which generated from 74 open-source systems. To evaluate performance on these datasets, 10-fold cross validation technique is that predicts model by partitioning original dataset into a training set train test it. Two feature selection techniques enhance our prediction accuracy. The Chi-squared Wrapper-based used improve accuracy total methods choosing top metrics each dataset. Results obtained applying two compared. algorithms, grid search-based parameter optimization applied. 100% was for Long-method using Logistic Regression algorithm with all features worst 95.20 % Naive Bayes chi-square technique.

Language: Английский

---

Machine/Deep Learning for Software Engineering: A Systematic Literature Review

IEEE Transactions on Software Engineering, Journal Year: 2022, Volume and Issue: 49(3), P. 1188 - 1231

Published: May 10, 2022

Since 2009, the deep learning revolution, which was triggered by introduction of ImageNet, has stimulated synergy between Software Engineering (SE) and Machine Learning (ML)/Deep (DL). Meanwhile, critical reviews have emerged that suggest ML/DL should be used cautiously. To improve applicability generalizability ML/DL-related SE studies, we conducted a 12-year Systematic Literature Review (SLR) on 1,428 papers published 2009 2020. Our trend analysis demonstrated impacts brought to SE. We examined complexity applying solutions problems how such led issues concerning reproducibility replicability studies in Specifically, investigated ML DL differ data preprocessing, model training, evaluation when applied tasks, what details need provided ensure study can reproduced or replicated. By categorizing rationales behind selection techniques into five themes, analyzed performance, robustness, interpretability, complexity, simplicity affected choices models.

Language: Английский

---

Predictive Models in Software Engineering: Challenges and Opportunities

ACM Transactions on Software Engineering and Methodology, Journal Year: 2022, Volume and Issue: 31(3), P. 1 - 72

Published: Jan. 31, 2022

Predictive models are one of the most important techniques that widely applied in many areas software engineering. There have been a large number primary studies apply predictive and present well-performed various research domains, including requirements, design development, testing debugging, maintenance. This article is first attempt to systematically organize knowledge this area by surveying body 421 papers on published between 2009 2020. We describe key approaches used, classify different models, summarize range application areas, analyze results. Based our findings, we also propose set current challenges still need be addressed future work provide proposed road map for these opportunities.

Language: Английский

---