Design patterns in performance prediction

Verkamo, A. Inkeri; Gustafsson, Janne; Nenonen, L.; Paakki, Jukka

doi:10.1145/350391.350422

Cited by 9 publications

(6 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both can be used in many domains for many different purposes, such as performance prediction [19], multi-agent systems [20], Java enterprise systems [21], and many more. Each pattern or antipattern includes a problem description and an abstraction of the solution that allows it to be reused in various settings.…”

Section: B Design Patterns and Antipatternsmentioning

confidence: 99%

Identifying Instances of Model Design Patterns and Antipatterns Using Model Clone Detection

Stephan

Cordy

2015

2015 IEEE/ACM 7th International Workshop on Modeling in Software Engineering

View full text Add to dashboard Cite

Abstract-A hurdle in the growth of model driven software engineering is our ability to evaluate the quality of models automatically. One perspective is that software quality is a function of the existence, or lack thereof, of good and bad properties, also known as patterns and antipatterns, respectively. In this paper, we introduce the notion of using model clone detection to detect model pattern and antipattern instances by looking for models that are cross clones of pattern models. By detecting patterns at the model level, analysis is accomplished earlier in the engineering process, can be applied to primarily modelbased projects, and remains at the same level of abstraction that engineers are used to. We outline the process of using model clone detection for this purpose, including representing the patterns and detection of instances. We present some Simulink examples of pattern representations and discuss future work and research in the area. I. INTRODUCTIONModel-driven Engineering (MDE) is a relatively new approach to software development that entails higher-level abstractions, or models, being used as the primary artifacts in system development and management. This includes incorporating modeling in all four phases of Software Engineering: requirements, design, implementation, and testing. By utilizing artifacts at an abstraction level closer to the problem space and hiding lower-level details, MDE can yield higher quality systems, facilitate better communication among stakeholders, and make project teams adaptable and flexible. MDE has seen significant levels of adoption in many different domains, including aviation, automotive, aerospace, and other highreliability embedded systems applications.While MDE use is growing, and engineers and customers are starting to experience the benefits of using model-based techniques [1], there are still many obstacles to overcome. One example is model quality evaluation, that is, the ability to assess quality of the models and artifacts of interest across the MDE life cycle [2], [3]. When it comes to more traditional software engineering processes, such as those that center around third generation programming languages, quality assurance is a well-researched and established area. In contrast, much less is understood about quality assurance for models in the MDE context. Improving our knowledge and making it easier to reason about model quality is a necessary step in continuing the growth of MDE.One approach to assessing software quality involves detecting the existence of design patterns [4] and antipatterns [5] in

show abstract

Section: B Design Patterns and Antipatternsmentioning

confidence: 99%

Identifying Instances of Model Design Patterns and Antipatterns Using Model Clone Detection

Stephan

Cordy

2015

2015 IEEE/ACM 7th International Workshop on Modeling in Software Engineering

View full text Add to dashboard Cite

show abstract

“…Antipatterns are general representations of bad practices and properties that software engineers should avoid. Patterns have been utilized in a variety of domains for different purposes, including Java enterprise system design patterns [22], predicting performance [23], and finding design patterns in multi-agent systems [24]. In all cases, the patterns have a description of their applicable contexts and an abstraction of the solution, allowing it to be used in different situations.…”

Section: Design Patterns and Antipatternsmentioning

confidence: 99%

Identification of Simulink model antipattern instances using model clone detection

Stephan

Cordy

2015

2015 ACM/IEEE 18th International Conference on Model Driven Engineering Languages and Systems (MODELS)

View full text Add to dashboard Cite

Abstract-One challenge facing the Model-Driven Engineering community is the need for model quality assurance. Specifically, there should be better facilities for analyzing models automatically. One measure of quality is the presence or absence of good and bad properties, such as patterns and antipatterns, respectively. We elaborate on and validate our earlier idea of detecting patterns in model-based systems using model clone detection by devising a Simulink antipattern instance detector. We chose Simulink because it is prevalent in industry, has mature model clone detection techniques, and interests our industrial partners. We demonstrate our technique using near-miss crossclone detection to find instances of Simulink antipatterns derived from the literature in four sets of public Simulink projects. We present our detection results, highlight interesting examples, and discuss potential improvements to our approach. We hope this work provides a first step in helping practitioners improve Simulink model quality and further research in the area.I. INTRODUCTION While Model-Driven Engineering (MDE) is becoming increasingly prevalent in the Software Engineering community, especially in both business and embedded domains [1], there are still numerous areas that need to be addressed as both model-based development and systems mature. In MDE, higher-level abstractions, or models, are the primary artifacts in all phases of the Software Engineering life cycle, including requirements, design, implementation, testing, and maintenance. Given the importance and longevity of these models, ascertaining and improving quality of the models and other artifacts of interest in MDE projects, known as model quality assurance, becomes imperative [2], [3]. Compared to quality assurance (QA) for more traditional software development paradigms such as code-based development, model quality assurance is much less refined and researched [4]. If MDE is to continue to flourish and allow engineers to fully reap all of its rewards, then evaluation and improvement of model quality is essential.One method of assessing the quality of Software Engineering systems is to identify and report well-established "good" and "bad" ways of solving specific design questions and constructing the systems' artifacts. These are known as design patterns [5] and antipatterns [6], respectively, and will be referred to collectively simply as "patterns" herein when we are referring to both. Detection of instances of these patterns provides analysts the facility to identify functional and nonfunctional properties that can be used to reason about the

show abstract

“…There are various approaches that aim to detect design patterns based on a static analysis of source code such as: FUJABA RE [11], [12], SPQR [18], [19], PINOT [17], PTIDEJ [9] or MAISA [21]. The design pattern detection mechanisms search for information defining an entire pattern or for sub-elements of patterns which can be combined to build patterns [1] or evaluate the similarity of the code structure with higher-level models as UML diagrams [3] or graph representations [20].…”

Section: Introductionmentioning

confidence: 99%

JADEPT: Dynamic Analysis for Behavioral Design Pattern Detection

Arcelli

Perin

Raibulet

et al. 2009

Proceedings of the 4th International Conference on Evaluation of Novel Approaches to Software Engineering

View full text Add to dashboard Cite

In the context of reverse engineering, the recognition of design patterns provides additional information related to the rationale behind the design. This paper presents our approach to the recognition of design patterns based on dynamic analysis of Java software. The idea behind our approach is to identify a set of rules capturing information necessary to identify a design pattern instance. Rules are characterized by weights indicating their importance in the detection of a specific design pattern. The core behavior of each design pattern may be described through a subset of these rules forming a macrorule. Macrorules define the main traits of a pattern. JADEPT (JAva DEsign Pattern deTector) is our software for design pattern identification based on this idea. It captures static and dynamic aspects through a dynamic analysis of the software by exploiting the JPDA (Java Platform Debugger Architecture). The extracted information is stored in a database. Queries to the database implement the rules defined to recognize design patterns. The tool has been validated with positive results on different academic implementations of design patterns and on systems as JADEPT itself.

show abstract

Design patterns in performance prediction

Abstract: Categories and Subject Descriptors

Cited by 9 publications

References 1 publication

Identifying Instances of Model Design Patterns and Antipatterns Using Model Clone Detection

Identifying Instances of Model Design Patterns and Antipatterns Using Model Clone Detection

Identification of Simulink model antipattern instances using model clone detection

JADEPT: Dynamic Analysis for Behavioral Design Pattern Detection

Contact Info

Product

Resources

About