On the effectiveness of developer features in code smell prioritization: A replication study

Background Code smells are an indicator that something is misplaced in software systems, which can reduce maintainability and quality of software applications. In the literature, there is much research about code smells where numerous strategies for automating code smell detection have been developed to improve software quality. Objective The purpose of this work is to provide a review of search‐based, heuristic‐based, machine learning‐based, deep learning‐based, and hybrid‐based code smell detection algorithms. Method Concerning the main goals of this research, we have found 38 primary studies. We gathered relevant studies published on this topic between 2017 and 2024. These articles' data were extracted according to some criteria, including code smells, machine learning methods, programming languages, dataset size, evaluation strategy, and statistical tests. Results Machine learning‐based code smell detection methods have been suggested in most empirical investigations. We found that machine learning and deep learning are the most popular approaches for predicting code smells. Researchers typically employ support vector machine and decision tree algorithms. The Arcelli Fontana and Zanoni benchmark dataset was the most frequently investigated dataset. Most of the research community's attention has been focused on a small number of smells, including blob, feature envy, long method, and data class. Researchers also pay more attention to code smells like Long Method and Feature Envy. Deep learning techniques are increasingly used, and most scientists utilize source code metrics as indicators. The standard performance measurements mostly used are F‐measure, recall, and precision. Conclusions The study provides an overview of existing approaches and highlights current research trends in code smell detection, particularly the increasing use of machine learning and deep learning techniques.

Context Code smells in the software systems are indications that usually correspond to deeper problems that can negatively influence software quality characteristics. This review is a part of a R&D project aiming to improve the existing codebeat platform that help developers to avoid code smells and deliver quality code. Objective This study aims to identify and investigate the current state of the art with respect to: (1) predictors used in prediction models to detect code smells, (2) machine learning/artificial intelligence (ML/AI) methods used in prediction models to detect code smells, (3) code smells analyzed in scientific literature. Our secondary objectives were to identify (4) data sets and projects used in research papers to predict code smells, (5) performance measures used to assess prediction models and (6) improvement ideas with regard to code smell detection using ML/AI. Method We conducted a systematic review using a database search in Scopus and evaluated it using the quasi-gold standard procedure to identify relevant studies. In the data sheet used to obtain data from publications we factor research questions into finer-grained ones, which are then answered on a per-publication basis. Those are then merged over a set of publications using an automated script to obtain answers to the posed research questions. Results We have identified 45 primary studies relevant to the primary objectives of this research. The results show the prediction capability of the ML/AI techniques for predicting code smells. Conclusion Only a few smells—Blob, Feature Envy, Long Method and Data Class—have received the vast majority of interest in research community. The usage of deep learning techniques is increasing. Most researchers still use source code metrics as predictors. Precision, recall and F-measure are the go-to performance metrics. There seems to be a need for modern reference data/projects sets that reflect modern constructs of programming languages. We identified various promising paths of research that have the potential to advance the state of the art in the area of code smells prediction.

Code smells are considered symptoms of poor implementation choices, which may hamper the software maintainability. Hence, code smells should be detected as early as possible to avoid software quality degradation. Unfortunately, detecting code smells is not a trivial task. Some preliminary studies investigated and concluded that machine learning (ML) techniques are a promising way to better support smell detection. However, these techniques are hard to be customized to promote an early and accurate detection of specific smell types. Yet, ML techniques usually require numerous code examples to be trained (composing a relevant dataset) in order to achieve satisfactory accuracy. Unfortunately, such a dependency on a large validated dataset is impractical and leads to late detection of code smells. Thus, a prevailing challenge is the early customized detection of code smells taking into account the typical limited training data. In this direction, this paper reports a study in which we collected code smells, from ten active projects, that were actually refactored by developers, differently from studies that rely on code smells inferred by researchers. These smells were used for evaluating the accuracy regarding early detection of code smells by using seven ML techniques. Once we take into account such smells that were considered as important by developers, the ML techniques are able to customize the detection in order to focus on smells observed as relevant in the investigated systems. The results showed that all the analyzed techniques are sensitive to the type of smell and obtained good results for the majority of them, especially JRip and Random Forest. We also observe that the ML techniques did not need a high number of examples to reach their best accuracy results. This finding implies that ML techniques can be successfully used for early detection of smells without depending on the curation of a large dataset.

Code smells regularly cause potential software quality problems in software development. Thus, code smell detection has attracted the attention of many researchers. A number of approaches have been suggested in order to improve the accuracy of code smell detection. Most of these approaches rely solely on structural information (code metrics) extracted from source code and heuristic rules designed by people. In this paper, We propose a method-representation based model to represent the methods in textual code, which can effectively reflect the semantic relationships embedded in textual code. We also propose a deep learning based approach that combines method-representation and a CNN model to detect feature envy. The proposed approach can automatically extract semantic and features from textual code and code metrics, and can also automatically build complex mapping between these features and predictions. Evaluation results on open-source projects demonstrate that our proposed approach achieves better performance than the state-of-the-art in detecting feature envy.

Are you smelling it? Investigating how similar developers detect code smells

Code smells are indicators of potential problems in software. They tend to have a negative impact on software quality. Several studies use machine learning techniques to detect bad smells. The objective of this study is to systematically review and analyze machine learning techniques used to detect code smells to provide interested research community with knowledge about the adopted techniques and practices for code smells detection. We use a systematic literature review approach to review studies that use machine learning techniques to detect code smells. Seventeen primary studies were identified. We found that 27 code smells were used in the identified studies; God Class and Long Method, Feature Envy, and Data Class are the most frequently detected code smells. In addition, we found that 16 machine learning algorithms were employed to detect code smells with acceptable prediction accuracy. Furthermore, we the results also indicate that support vector machine techniques were investigated the most. Moreover, we observed that J48 and Random Forest algorithms outperform the other algorithms. We also noticed that, in some cases, the use of boosting techniques on the models does not always enhance their performance. More studies are needed to consider the use of ensemble learning techniques, multiclassification, and feature selection technique for code smells detection. Thus, the application of machine learning algorithms to detect code smells in systems is still in its infancy and needs more research to facilitate the employment of machine learning algorithms in detecting code smells.

The objective of this paper is to study the relationship between different types of object-oriented software metrics, code smells and actual changes in software code that occur during maintenance period. It is hypothesized that code smells are indicators of maintenance problems. To understand the relationship between code smells and maintenance problems, we extract code smells in a Java based mobile application called MOBAC. Four versions of MOBAC are studied. Machine learning techniques are applied to predict software change-proneness with code smells as predictor variables. The results of this paper indicate that codes smells are more accurate predictors of change-proneness than static code metrics for all machine learning methods. However, class imbalance techniques did not outperform class balance machine learning techniques in change-proneness prediction. The results of this paper are based on accuracy measures such as F-measure and area under ROC curve.

The quality of source code is negatively impacted by code smells. Since the term "code smell" originated, numerous attempts have been made to comprehend it by identifying it using various techniques, such as metric-based, heuristic-based, optimization-based, machine learning (ML)-based, etc. Among these, supervised machine learning (SML) has shown effectiveness in detecting code smells. However, SML techniques have significant limitations, including the dependency on expensive and high-quality labeled data, the need for representative training datasets, and the risk of introducing biases in labeled examples that lead to skewed predictions. To overcome these challenges, this study introduces a method that leverages unsupervised machine learning (UnML) along with feature engineering. Unlike SML, UnML does not require labeled data and minimizes potential biases. The proposed method was evaluated using four datasets containing different types of code smells and was compared with a previous study that used SML techniques. The results indicate that the UnML-based method is effective, achieving outcomes closely aligned with those from the SML approach. This method is especially beneficial in situations where labeled data is scarce or unavailable and can be used to identify new code smells, generate labeled data for SML and detect multiple code smells simultaneously within a codebase.

In software engineering, maintainability is related to investigating the defects and their causes, correcting the defects and modifying the system to meet customer requirements. Maintenance is a time consuming activity within the software life cycle. Therefore, there is a need for efficiently organizing the software resources in terms of time, cost and personnel for maintenance activity. One way of efficiently managing maintenance resources is to predict defects that may occur after the deployment. Many researchers so far have built defect prediction models using different sets of metrics such as churn and static code metrics. However, hidden causes of defects such as code smells have not been investigated thoroughly. In this study we propose using data science and analytics techniques on software data to build defect prediction models. In order to build the prediction model we used code smells metrics, churn metrics and combination of churn and code smells metrics. The results of our experiments on two different software companies show that code smells is a good indicator of defect proneness of the software product. Therefore, we recommend that code smells metrics should be used to train a defect prediction model to guide the software maintenance team.

Code smells are symptoms of poor design and implementation choices that may hinder code comprehension and possibly increase change- and fault-proneness of source code. Several techniques have been proposed in the literature for detecting code smells. These techniques are generally evaluated by comparing their accuracy on a set of detected candidate code smells against a manually-produced oracle. Unfortunately, such comprehensive sets of annotated code smells are not available in the literature with only few exceptions. In this paper we contribute (i) a dataset of 243 instances of five types of code smells identified from 20 open source software projects, (ii) a systematic procedure for validating code smell datasets, (iii) LANDFILL, a Web-based platform for sharing code smell datasets, and (iv) a set of APIs for programmatically accessing LANDFILL's contents. Anyone can contribute to Landfill by (i) improving existing datasets (e.g., Adding missing instances of code smells, flagging possibly incorrectly classified instances), and (ii) sharing and posting new datasets. Landfill is available at www.sesa.unisa.it/landfill/, while the video demonstrating its features in action is available at http://www.sesa.unisa.it/tools/landfill.jsp.

Code smells are specific patterns or characteristics in software code that indicate potential design or implementation problems. Identifying code smells has gained significant attention in software engineering. It is essential to address code smells to maintain high-quality software systems. Machine learning (ML) models, such as Long Short-Term Memory (LSTM), have been to detect code smells automatically based on source code features. However, the imbalanced distribution of code smells within software projects poses a challenge to the accuracy of these models. This study explores the role of data balancing methods in optimizing the accuracy of the LSTM model for code smell detection. We investigate different techniques for addressing the class imbalance problem, including random oversampling and synthetic minority oversampling techniques (SMOTE). We evaluate the performance of the LSTM model with and without data balancing methods using accuracy, precision, recall, f-measure, Matthew’s correlation coefficient (MCC), and the area under a receiver operating characteristic curve (AUC). Our experimental results, conducted on four code smell datasets (God class, data class, feature envy, and long method) extracted from 74 open-source systems, demonstrate the effectiveness of data balancing methods in improving the accuracy of the LSTM model for code smell detection. The results indicate that the use of data balancing methods had a positive effect on the predictive accuracy of the LSTM model. In addition, we compared our proposed method with state-of-the-art code smell detection approaches. The findings from the comparison indicate that our proposed method performs notably better than existing state-of the-art approaches across the majority of datasets.

Code smells can be subjectively interpreted, the results provided by detectors are usually different, the agreement in the results is scarce, and a benchmark for the comparison of these results is not yet available. The main approaches used to detect code smells are based on the computation of a set of metrics. However code smell detectors often use different metrics and/or different thresholds, according to their detection rules. As result of this inconsistency the number of detected smells can increase or decrease accordingly, and this makes hard to understand when, for a specific software, a certain characteristic identifies a code smell or not. In this work, we introduce WekaNose, a tool that allows to perform an experiment to study code smell detection through machine learning techniques. The experiment's purpose is to select rules, and/or obtain trained algorithms, that can classify an instance (method or class) as affected or not by a code smell. These rules have the main advantage of being extracted through an example-based approach, rather then a heuristic-based one.

리팩토링은 내부적으로는 시스템의 품질을 개선하고, 외부적으로는 시스템의 기능을 유지하는 일종의 소프트웨어를 변경하는 과정이다. 이러한 리팩토링을 적용하여 기존 소스코드를 개선하기 위해서는 개선할 사항이 무엇인지를 아는 것이 우선이다. 이를 위해 Martin Fowler와 Kent Beck은 코드속의 나쁜 냄새(코드스멜)를 식별할 수 있는 방법을 제시 하였다. 또한 코드스멜을 탐지하고 어디에 어떤 리팩토링을 적용할 것인가를 결정하는 문제와 관련된 몇몇 연구가 발표되었다. 그러나 이러한 연구들은 코드스멜에 대한 명확한 표현이 부족하거나 한정된 코드스멜만을 탐지하는 단점이 있다. 그리고 리팩토링을 적용할 경우 행위보존을 위한 선행조건들의 표현방법이 리팩토링 절차에 포함되어 있거나 정형화되지 않아 행위보존의 모호함이 발생되는 단점을 가지고 있다. 이에 본 논문에서는 OCL을 이용하여 코드스멜의 정보를 정확히 명세화하고, OCL 번역기를 통해 코드스멜을 자동으로 탐지하여 리팩토링하는 프레임워크를 제안한다. 또한 적용사례를 통하여 자바소스코드속의 코드스멜을 OCL로 명세화하여 자동탐지하고, 리팩토링을 적용해 봄으로써 활용성과 효용성을 검증해본다. Refactoring is a kind of software modification process that improves system qualities internally but maintains system functions externally. What should be improved on the existing source codes should take precedence over the others in such a modification process using this refactoring. Martin Fowler and Kent Beck proposed a method that identifies code smells for this purpose. Also, some studies on determining what refactoring will be applied to which targets through detecting code smells in codes were presented. However, these studies have a lot of disadvantages that show a lack of precise description for such code smells and detect limited code smells only. In addition, these studies showed other disadvantages that generate ambiguity in behavior preservation due to the fact that a description method of pre-conditions for the behavior preservation is included in a refactoring process or unformalized. Thus, our study represents a precise specification of code smells using OCL and proposes a framework that performs a refactoring process through the automatic detection of code smells using an OCL interpreter. Furthermore, we perform the automatic detection in which the code smells are be specified by using OCL to the java program and verify its applicability and effectivity through applying a refactoring process.

Improving and comparing performance of machine learning classifiers optimized by swarm intelligent algorithms for code smell detection

In the realm of software development, detecting code smells is a critical task for ensuring good code quality. God class code smell specifically has a specific nature associated with a great deal of subjectivity due to the levels of coupling and cohesion associated to it. Automated techniques for code smell detection aim to resolve this subjectivity. Machine learning techniques have shown promising results that tend to improve accuracy and reduce the bias associated with other techniques for God class identification. This is due to their pattern recognition capabilities making them more objective in identifying patterns that indicate code smells. However, current results need to be further improved in terms of both accuracy and generalizability. The challenge in the use of machine learning is not only in selecting the most appropriate technique but also lies in effectively representing source code as input patterns fed to Machine Learning (ML) classifier(s). Code representation plays a pivotal role in encoding source code for ML algorithms. This study aims improving the accuracy and generalizability of God class code smell detection via exploring the effect of using various code representation techniques, namely, tree-based, metric-based, code embedding, and token-based code representation techniques on the ML detection results. The study is conducted on the MLCQ dataset, and applies various ML algorithms (specifically: Logistic Regression, Random Forest, SVM, Decision Tree, Naive Bayes, Gradient Boosting, XGBoost). The evaluation results show how different code representation techniques influence ML detection outcomes and the comparative performance of ML algorithms. The study findings reveal that the F1-score achieved outperforms prior studies on the MLCQ dataset, indicating the effectiveness of the proposed approach. The presented results reveal how the code representation technique used makes a significant impact on the ML classifier results. This paves the way for developing intelligent IDE plugins for just in time God Class code smell detection among other code smells.