Tuning research tools for scalability and performance: The NiCad experience

Cordy, James R.; Roy, Chanchal K.

doi:10.1016/j.scico.2011.11.002

Cited by 10 publications

(9 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The last two techniques allow to discard or to abstract in the comparison, respectively, a subset of the nonterminals of the language-specific grammar employed by the parser. NiCad has been chosen as reference because it is a mature and flexible tool [23], which allows for customization of both the language grammar and of the detection process, making it possible to detect Type-2 Java clones. NiCad can be configured to recognize the clones as defined in Section 3.1 if one allows for:…”

Section: Resultsmentioning

confidence: 99%

CD-Form: A clone detector based on formal methods

Cuomo

Santone

Villano

2014

Science of Computer Programming

View full text Add to dashboard Cite

h i g h l i g h t s• We propose the use of formal methods (Calculus of Communicating Systems) for detection of software clones.• We introduce a technique for translating Java bytecode to process algebra representation.• We present an optimized implementation able to deal with real-world software projects.• We provide a comparison with a state-of-the-art tool, obtaining comparable results. a r t i c l e i n f o a b s t r a c tThis paper presents the design and implementation of CD-Form (Clone Detector based on FORmal Methods), a tool targeted at the detection of Type-2 clones in Java code. CD-Form is based on a novel approach for detecting code clones. The methodology adopted performs the analysis on Java bytecode and not on the original Java source. The bytecode is transformed into CCS (Calculus of Communicating Systems) processes, which are successively checked for equivalence. After a thorough description of the methodology used for detecting clones, the design of the tool is presented. The results obtained by evaluating sample Java codes are validated by comparing them to those obtained by a state-of-the-art tool for clone detection.

show abstract

Section: Resultsmentioning

confidence: 99%

CD-Form: A clone detector based on formal methods

Cuomo

Santone

Villano

2014

Science of Computer Programming

View full text Add to dashboard Cite

show abstract

“…Different setups in corresponding clone types might result in different stability scenarios. However, the NiCad setups that we have used for detecting three types clones are considered standard [23,24,26,4,29] and thus we contend that the clone detection results that we have investigated are reliable.…”

Section: Threats To Validitymentioning

confidence: 89%

An empirical study on clone stability

Mondal

Roy

Schneider

2012

SIGAPP Appl. Comput. Rev.

Self Cite

View full text Add to dashboard Cite

Code cloning is a controversial software engineering practice due to contradictory claims regarding its effect on software maintenance. Code stability is a recently introduced measurement technique that has been used to determine the impact of code cloning by quantifying the changeability of a code region. Although most existing stability analysis studies agree that cloned code is more stable than non-cloned code, the studies have two major flaws: (i) each study only considered a single stability measurement (e.g., lines of code changed, frequency of change, age of change); and, (ii) only a small number of subject systems were analyzed and these were of limited variety. In this paper, we present a comprehensive empirical study on code stability using four different stability measuring methods. We use a recently introduced hybrid clone detection tool, NiCAD, to detect the clones and analyze their stability in different dimensions: by clone type, by measuring method, by programming language, and by system size and age. Our in-depth investigation on 12 diverse subject systems written in three programming languages considering three types of clones reveals that: (i) cloned code is generally less stable than non-cloned code, and more specifically both Type-1 and Type-2 clones show higher instability than Type-3 clones; (ii) clones in both Java and C systems exhibit higher instability compared to the clones in C# systems; (iii) a system's development strategy might play a key role in defining its comparative code stability scenario; and, (iv) cloned and non-cloned regions of a subject system do not follow any consistent change pattern. 1

show abstract

“…The raw samples from GitHub, however, contain clones which may affect the training performance. To reduce the impact of duplicated data on classification results, we use NiCad [13] to detect clones among the data, and remove all of the Type 1, Type 2 and Type 3 clones with a dissimilarity lower than 10%. After the clone removal, we obtained 4,932 algorithm files in Java and 4,732 for C++.…”

Section: A Datasetsmentioning

confidence: 99%

Bilateral Dependency Neural Networks for Cross-Language Algorithm Classification

Nghi

Jiang

2019

2019 IEEE 26th International Conference on Software Analysis, Evolution and Reengineering (SANER)

View full text Add to dashboard Cite

Algorithm classification is to automatically identify the classes of a program based on the algorithm(s) and/or data structure(s) implemented in the program. It can be useful for various tasks, such as code reuse, code theft detection, and malware detection. Code similarity metrics, on the basis of features extracted from syntax and semantics, have been used to classify programs. Such features, however, often need manual selection effort and are specific to individual programming languages, limiting the classifiers to programs in the same language. To recognize the similarities and differences among algorithms implemented in different languages, this paper describes a framework of Bilateral Neural Networks (Bi-NN) that builds a neural network on top of two underlying sub-networks, each of which encodes syntax and semantics of code in one language. A whole Bi-NN can be trained with bilateral programs that implement the same algorithms and/or data structures in different languages and then be applied to recognize algorithm classes across languages. We have instantiated the framework with several kinds of token-, tree-and graph-based neural networks that encode and learn various kinds of information in code. We have applied the instances of the framework to a code corpus collected from GitHub containing thousands of Java and C++ programs implementing 50 different algorithms and data structures. Our evaluation results show that the use of Bi-NN indeed produces promising algorithm classification results both within one language and across languages, and the encoding of dependencies from code into the underlying neural networks helps improve algorithm classification accuracy further. In particular, our custom-built dependency trees with tree-based convolutional neural networks achieve the highest classification accuracy among the different instances of the framework that we have evaluated. Our study points to a possible future research direction to tailor bilateral and multilateral neural networks that encode more relevant semantics for code learning, mining and analysis tasks.

show abstract

Tuning research tools for scalability and performance: The NiCad experience

Cited by 10 publications

References 24 publications

CD-Form: A clone detector based on formal methods

CD-Form: A clone detector based on formal methods

An empirical study on clone stability

Bilateral Dependency Neural Networks for Cross-Language Algorithm Classification

Contact Info

Product

Resources

About