GEML: A grammar-based evolutionary machine learning approach for design-pattern detection

Barbudo, Rafael; Ramírez, Aurora; Servant, Francisco; Romero, José Raúl

doi:10.1016/j.jss.2021.110919

Cited by 11 publications

(8 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These unbalanced results are caused by the number of instances of singleton existing in each corpus, which have a great impact on the learning of the classifier. Results made by [13] and [35] do not refer exactly to the performance of both approaches to recover singleton pattern instances because other patterns are included. Therefore, we have tested the DPDf and GEML with Singleton specific data.…”

Section: Dpdf Reported Resultsmentioning

confidence: 99%

“…Another recent work proposed in [35] presents a novel machine learning-based approach for DPD named GEML. Like other work, the used method explores the ML capacity, but Barbudo et al [35] addressed their limitations by using G3P as a basis of the proposed approach.…”

Section: Related Workmentioning

confidence: 99%

“…To position our work against state-of-the-art studies, we select two recent relevant approaches working also with Machine Learning. The first is proposed by Rafael Barbudo et al [35] named GEML and the second is proposed by Najam Nazar et al [13] named DPDf.…”

Section: Comparison With Similar Existing Approachesmentioning

confidence: 99%

“…In [35] a comparison between GEML and other DPD methods, including both ML and non-ML is realized. GEML outperform MARPLE techniques [26] with high values in term of accuracy and f1 score.…”

Section: The Benchmark Dp Detection Approachesmentioning

confidence: 99%

“…Recently, Najam Nazar et al in [13] selected 15 feature codes and use machine learning classifiers to automatically train a design pattern detector. Another recent work proposed in [35] presents a novel machine learning-based approach for DPD named GEML. Like other work, the used method explores the ML capacity, but Barbudo et al [35] addressed their limitations by using G3P as a basis of the proposed approach.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Features and Supervised Machine Learning-Based Method for Singleton Design Pattern Variants Detection

Nacef

Bahroun

Khalfallah

et al. 2022

Preprint

View full text Add to dashboard Cite

Design patterns codify standard solutions to common problems in software design and architecture. Given their importance in improving software quality and facilitating code reuse, many types of research are proposed on their automatic detection. In this paper, we focus on singleton pattern recovery by proposing a method that can identify orthodox implementations and non-standard variants. The recovery process is based on specific data created using a set of relevant features. These features are specific information defining each variant which is extracted from the java program by structural and semantic analysis. We are based on the singleton analysis and different proposed features presented by Nacef et al. in [1] to create structured data. This data contains a combination of feature values defining each singleton variant to train a supervised Machine Learning (ML) algorithm. The goal is not limited to detecting the singleton pattern but also the specification of the implemented variant as so as the incorrect structure that is incoherent with the pattern intent. We use different ML algorithms to create the Singleton Detector (SD) to compare their performance. The empirical results demonstrate that our method for automatically training the (SD) based on features and supervised ML, can identify any singleton implementation with the specific variant’s name achieving 99% of precision, and recall. We have compared the proposed approach to similar studies namely DPDf and GEML. The results show that the SD outperforms the state-of-the-art approaches by more than 20% on evaluated data constructed from different repositories; PMART, DPB and DPDf corpus in terms of precision.

show abstract

Section: Dpdf Reported Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%