Human-level Ordinal Maintainability Prediction Based on Static Code Metrics

Schnappinger, Markus; Fietzke, Arnaud; Pretschner, Alexander

doi:10.1145/3463274.3463315

Cited by 5 publications

(6 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…So far, we did not investigate combinations of metrics, e.g. using machine learning [7], [31]. One of the first steps in building machine-learned models is to identify adequate feature candidates.…”

Section: B Threats To Validitymentioning

confidence: 99%

See 1 more Smart Citation

Revisiting Inter-Class Maintainability Indicators

Gregor,

Schnappinger,

Pretschner

2023

2023 IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER)

Self Cite

View full text Add to dashboard Cite

Over the last few decades, a variety of static code metrics have been published and promoted to measure the maintainability of software systems.This study evaluates 12 common static code metrics for their correlation with observed maintenance efforts. Leveraging modern repository mining techniques, we examine the historical data of three large open-source software systems with a combined size of over 1M LOC and over 10k classes. We automatically identify maintenance activities and measure the effort needed to perform them through revised lines of code. Then, we investigate if the state of the system as captured by these metrics is an indicator for the required maintenance effort.In contrast to earlier research, our results could not validate a general correlation between any of the examined metrics and maintainability. Instead, all evaluated metrics showed positive and negative correlations with maintenance effort depending on the considered time interval. Strong correlations only hold for specific projects, and within these projects, only for limited time spans. Across the project history, however, all metrics showed moderate correlations at most.As no metric was found to be a good indicator for high maintenance efforts in all contexts, we advocate against using any of the evaluated metrics without project-specific validation. If metrics are to be used to monitor the maintainability of a system, either directly or through models based on these metrics, engineers have to validate their applicability not just for the project at hand, but also for the current time span.

show abstract

Section: B Threats To Validitymentioning

confidence: 99%

“…One promising approach to finding a universal predictor for maintainability may be through aggregating code metrics based on machine learning as in [7]. However, their models predict the maintainability judgment of analysts that focuses on the quality of individual code files.…”

Section: Introductionmentioning

confidence: 99%

Revisiting Inter-Class Maintainability Indicators

Gregor,

Schnappinger,

Pretschner

2023

2023 IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER)

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the same context, Schnappinger et al [23] resorted to engaging experts in order to manually label a set of data regarding their maintainability degree and made use of a set of various metrics to evaluate maintainability. Other recent research works examine various metrics and approaches for evaluating maintainability, such as the examination of open source projects and their maintainability degree [24], the combination of machine learning techniques such as the Bayesian networks and association rules [25], the use of soft computing techniques such as the neuro-fuzzy model [26] or employing ensembles to predict maintainability on imbalanced data [27].…”

Section: Related Workmentioning

confidence: 99%

Analyzing Static Analysis Metric Trends towards Early Identification of Non-Maintainable Software Components

Karanikiotis

Papamichail²,

Symeonidis

2021

Sustainability

View full text Add to dashboard Cite

Nowadays, agile software development is considered a mainstream approach for software with fast release cycles and frequent changes in requirements. Most of the time, high velocity in software development implies poor software quality, especially when it comes to maintainability. In this work, we argue that ensuring the maintainability of a software component is not the result of a one-time only (or few-times only) set of fixes that eliminate technical debt, but the result of a continuous process across the software’s life cycle. We propose a maintainability evaluation methodology, where data residing in code hosting platforms are being used in order to identify non-maintainable software classes. Upon detecting classes that have been dropped from their project, we examine the progressing behavior of their static analysis metrics and evaluate maintainability upon the four primary source code properties: complexity, cohesion, inheritance and coupling. The evaluation of our methodology upon various axes, both qualitative and quantitative, indicates that our approach can provide actionable and interpretable maintainability evaluation at class level and identify non-maintainable components around 50% ahead of the software life cycle. Based on these results, we argue that the progressing behavior of static analysis metrics at a class level can provide valuable information about the maintainability degree of the component in time.

show abstract

“…Our research project concerns software maintainability assessment working in collaboration with the development team [13]. As part of this effort, we began to reproduce a study from Schnappinger et al [33], because their work is based on a recent, high quality software maintainability dataset [32]. This introduction presents what maintainability and metrics are, before continuing on the importance of the program structure and having clear and unambiguous metrics.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, while reproducing the study from Schnappinger et al [33], we encountered many problems collecting metrics. Many tools exist and many metrics have been studied [12,28].…”

Section: Introductionmentioning

confidence: 99%

Building an Operable Graph Representation of a Java Program as a Basis for Automatic Software Maintainability Analysis

Bertrand

Favier

André

2022

The International Conference on Evaluation and Assessment in Software Engineering 2022

View full text Add to dashboard Cite

As a part of a research project concerning software maintainability assessment in collaboration with the development team, we were interested in the frequent use of metrics as predictors. Many metrics exist, often with opaque and arguable implementations. We claim metrics mix the assessment of presentation, structure and model. In order to focus on true detectable maintainability defects, we computed metrics solely based on the structure of the program. Our approach was to parse the source code of Java programs as a graph, and to compute metrics in a declarative query language. To this end, we developed Javanalyser and implemented 34 metrics using Spoon to parse Java programs and Neo4j as graph database. We will show that the program graph constitutes a steady basis to compute met-rics and conduct future machine-learning studies to assess maintainability.

show abstract

Human-level Ordinal Maintainability Prediction Based on Static Code Metrics

Cited by 5 publications

References 43 publications

Revisiting Inter-Class Maintainability Indicators

Revisiting Inter-Class Maintainability Indicators

Analyzing Static Analysis Metric Trends towards Early Identification of Non-Maintainable Software Components

Building an Operable Graph Representation of a Java Program as a Basis for Automatic Software Maintainability Analysis

Contact Info

Product

Resources

About