Nikola Luburić, Simona Prokić, Katarina-Glorija Grujić et al. · Zenodo (CERN European Organization for Nuclear Research) · 2022
This dataset includes open-source projects written in C# programing language, annotated for the presence of Long Method and God Class code smells. Each instance was manually annotated by at least two annotators. We explain our motivation and methodology for creating this dataset in our preprint: Luburić, N., Prokić, S., Grujić, K.G., Slivka, J., Kovačević, A., Sladić, G. and Vidaković, D., 2021. Towards a systematic approach to manual annotation of code smells. The dataset contains two excel datasheets: <em>DataSet_Large Class.xlsx</em> – C# classes annotated for the Large Class code smell severity. <em>DataSet_Long Method.xlsx</em> – C# methods annotated for the Long method code smell severity. The columns in the datasheet represent: <em>Code Snippet ID</em> – the full name of the code snippet. For classes, this is the package/namespace name followed by the class name. The full name of inner classes also contains the names of any outer classes (e.g., <em>namespace.subnamespace.outerclass.innerclass</em>). For methods, this is the full name of the class and the methods’s signature (e.g., <em>namespace.class.method(param1Type, param2Type)</em> ). <em>Link </em>– The GitHub link to the code snippet, including the commit and the start and end LOC. <em>Code Smell </em>– code smell for which the code snippet is examined (Large Class or Long Method). <em>Project Link </em>– the link to the version of the code repository that was annotated. <em>Metrics </em>– a list of metrics for the code snippet, calculated by our platform. Our dataset provides 25 class-level metrics for Large Class detection and 18 method-level metrics for Long Method detection The list of metrics and their definitions is available here. <em>Final annotation </em>– a single severity score calculated by a majority vote. <em>Annotators </em>– each annotator's (1, 2, or 3) assigned severity score. To help guide their reasoning for evaluating the presence and the severity of a code smell, three annotators independently annotated whether the considered heuristics apply to an evaluated code snippet. We provide these results in two separate excel datasheets: <em>LargeClass_Heuristics.xlsx </em>- C# classes annotated for the presence of heuristics relevant for the Large Class code smell. <em>LongMethod_Heuristics.xlsx </em>- C# classes annotated for the presence of heuristics relevant for the Large Class code smell. The columns of these two datasheets are: <em>Code Snippet ID </em>- the full name of the code snippet (matching the IDs from <em>DataSet_Large Class.xlsx </em>and <em>DataSet_Long Method.xlsx</em>) <em>Annotators</em> – heuristics labelled by each of the annotators (1, 2, or 3). <em>Heuristics </em>– whether the heuristic is applicable to the examined code snippet or not (Section 1.2.4 lists heuristics relevant for the Large Class detection, and Section 1.2.5 lists the heuristics relevant for the Long Method detection).
Cited by 1Open access
O.Ye. Demenko · Juridical scientific and electronic journal · 2023
Деменко О.Є., к
Cited by 0Open access
Jarmo Koponen, Keijo Haataja, Pekka Toivanen · F1000Research · 2022
<b>Background:</b> Manufacturing markings printed on products play an important role in the handling and use of pharmaceuticals and perishable foods. Currently, optical character recognition, neural networks, deep learning-based methods, and combinations of these methods are used to recognize these codes. <b>Methods:</b> This systematic review was performed to find papers that can answer the following research questions: How have machine vision methods that can recognize product texts evolved over the past eight years? What are the most common difficulties in recognizing product texts? Articles published between 2012 and 2020 were systematically searched from Science Direct/SCOPUS, and Google Scholar in November-December 2020. Ten studies were eligible, with inclusion criteria: details about the recognition method used, performance analysis result, imaging method, product and the text printed on it. <b>Results:</b> Product text recognition methods have evolved significantly over the last two years to tolerate the most common difficulties in the field. This is due to the ability of the deep learning neural network (DNN) architectures such as convolutional neural networks (CNN) to extract and learn salient character features straight from packaging surface images. Four of the most recent methods use two consecutive deep learning networks, one detecting the text area based on an image captured from the product package's surface and the other recognizing the characters within. Furthermore, this paper presents solutions to the most common product text recognition difficulties. <b>Conclusions:</b> There were a limited number of studies that met the eligibility criteria for this systematic review. The study's aim was to evaluate the development of machine vision methods for recognizing manufacturing marking texts printed on the surface of products. The study results demonstrated how methods have evolved over time, beginning with optical character recognition, and advancing to methods which can recognize texts despite the field's most common problems.
Cited by 4Open access
Saima Afrin, MD Zahidul Haque, Antonio Mastropaolo · ACM Transactions on Software Engineering and Methodology · 2026
The rise of Artificial Intelligence (AI)-and particularly Large Language Models (LLMs) for code–has reshaped Software Engineering (SE) by enabling the automation of tasks such as code generation, bug detection, and repair. However, these models require significant computational resources for training and fine-tuning, posing challenges for real-world adoption in resource-constrained environments. To address this, the research community has increasingly turned to Parameter-Efficient Fine-Tuning (PEFT)–a class of techniques that enables the adaptation of large models by updating only a small subset of parameters, rather than the entire model. In this Systematic Literature Review (SLR), we examine the growing application of PEFT techniques–across a wide range of software engineering tasks. We analyze how these methods are used to optimize various deep learning (DL) architectures, focusing on their impact on both performance and efficiency. Our study synthesizes findings from 28 peer-reviewed papers, identifying patterns in configuration strategies and adaptation trade-offs. The outcome of this review is a comprehensive taxonomy that categorizes PEFT usage by task type, distinguishing between generative (e.g., Code Summarization) and non-generative (e.g., Code Clone Detection) scenarios. Our findings aim to inform future research and guide the practical deployment of PEFT in sustainable, AI-powered software development. Our artifacts are publicly available at https://github.com/alvi75/SLR-PEFT
Cited by 0
Satwinder Singh, Sharanpreet Kaur · Ain Shams Engineering Journal · 2017
Cited by 90Open access
Atik Ul Mussanadah, Muhammad Adlin Saputra, Hasan Mukhibad et al. · Zenodo (CERN European Organization for Nuclear Research) · 2026
Dataset supporting the systematic review on accounting information system governance (bibliographic metadata and author-generated analysis) entitled "Code, Control, and Performance: A Systematic Review on AIS-Governance Nexus"
Cited by 0Open access
Lucy Chesson, Sarah Whitehead, Kirsten Flanagan et al. · Journal of science and medicine in sport · 2020
Cited by 8
Princis, Henrijs, Sharma, Arindam, David, Cristina · ArXiv.org · 2025
Large language models (LLMs) have shown remarkable ability to generate code, yet their outputs often violate syntactic or semantic constraints when guided only through natural language prompts. We introduce TreeCoder, the most general and flexible framework to date for exploring decoding strategies, constraints, and hyperparameters in LLMs, and use it in code generation to enforce correctness and structure during decoding rather than relying on prompt engineering. TreeCoder represents decoding as a tree search over candidate programs, where both decoding strategies and constraint functions - such as style, syntax, execution - are treated as first-class, optimisable components. This design enables systematic exploration and automatic tuning of decoding configurations using standard optimisation techniques. Experiments on the MBPP (Python) and SQL-Spider benchmarks show that TreeCoder consistently improves accuracy across open-source models such as CodeLlama, Mistral and DeepSeek, often outperforming their unconstrained baselines by considerable margins.
Cited by 0Open access
Chih-Hua Chien · SHILAP Revista de lepidopterología · 2016
The lookup table methods for decoding binary systematic Quadratic Residue (QR) code are presented in this paper. The key ideas behind this decoding technique are based on one to one corresponding mapping between the syndromes and the correctable error patterns. Such algorithms determine the error locations directly by lookup tables without the operations of addition and multiplication over a finite field. Moreover, the methods to dramatically reduce the memory requirement by shift-search decoding are utilized. Two new algorithm have been verified through a software simulation in C language. The new approach is modular, regular and naturally suitable for System on Chip (SOC) software implementation.
Cited by 1Open access
Lukas Rumbutis, Asta Slotkienė, Birutė Pliuskuvienė · New Trends in Computer Sciences · 2024
Identifying and understanding reasons for deriving software development defects is crucial for ensuring software product quality attributes such as maintainability. This paper presents a systematic literature review and the objective is to analyze the suggestions of other authors regarding software code defect prediction using machine learning, deep learning, or other artificial intelligence methods for the identification of software quality. The systemic literature review reveals that many analyzed papers considered multiple software code defects, but they were analyzed individually. However, more is needed to identify software quality attributes. The more profound analysis of code smells indicates the significance when considering multiple detected code smells and their interconnectedness; it helps to identify the software quality sub-attributes of maintainability.
Cited by 0Open access
Cristina David, Arindam Sharma, Henrijs Princis · Zenodo (CERN European Organization for Nuclear Research) · 2026
Cited by 0Open access
Cristina David, Arindam Sharma, Henrijs Princis · Zenodo (CERN European Organization for Nuclear Research) · 2026
Cited by 0Open access
Muhammad Atif Javed, Uwe Zdun · 2014
The links between the software architecture and the source code of a software system should be based on solid traceability mechanisms in order to effectively perform quality control and maintenance of the software system. There are several primary studies on traceability between software architecture and source code but so far no systematic literature review (SLR) has been undertaken. This study presents an SLR which has been carried out to discover the existing traceability approaches and tools between software architecture and source code, as well as the empirical evidence for these approaches, their benefits and liabilities, their relations to software architecture understanding, and issues, barriers, and challenges of the approaches. In our SLR the ACM Guide to Computing Literature has been electronically searched to accumulate the biggest share of relevant scientific bibliographic citations from the major publishers in computing. The search strategy identified 742 citations, out of which 11 have been included in our study, dated from 1999 to July, 2013, after applying our inclusion and exclusion criteria. Our SLR resulted in the identification of the current state-of-the-art of traceability approaches and tools between software architecture and source code, as well as gaps and pointers for further research. Moreover, the classification scheme developed in this paper can serve as a guide for researchers and practitioners to find a specific approach or set of approaches that is of interest to them.
Cited by 44
Princis, Henrijs, Sharma, Arindam, David, Cristina · arXiv (Cornell University) · 2025
Large language models (LLMs) have shown remarkable ability to generate code, yet their outputs often violate syntactic or semantic constraints when guided only through natural language prompts. We introduce TreeCoder, the most general and flexible framework to date for exploring decoding strategies, constraints, and hyperparameters in LLMs, and use it in code generation to enforce correctness and structure during decoding rather than relying on prompt engineering. TreeCoder represents decoding as a tree search over candidate programs, where both decoding strategies and constraint functions - such as style, syntax, execution - are treated as first-class, optimisable components. This design enables systematic exploration and automatic tuning of decoding configurations using standard optimisation techniques. Experiments on the MBPP (Python) and SQL-Spider benchmarks show that TreeCoder consistently improves accuracy across open-source models such as CodeLlama, Mistral and DeepSeek, often outperforming their unconstrained baselines by considerable margins.
Cited by 0Open access
Maha Tufail Agro, Atharva Kulkarni, Karima Kadaoui et al. · ArXiv.org · 2025
Motivated by a growing research interest into automatic speech recognition (ASR), and the growing body of work for languages in which code-switching (CS) often occurs, we present a systematic literature review of code-switching in end-to-end ASR models. We collect and manually annotate papers published in peer reviewed venues. We document the languages considered, datasets, metrics, model choices, and performance, and present a discussion of challenges in end-to-end ASR for code-switching. Our analysis thus provides insights on current research efforts and available resources as well as opportunities and gaps to guide future research.
Cited by 0Open access
Wolf, Christian K. · ICPSR Data Holdings · 2023
Data and code to accompany the article "Fiscal Stimulus and the Systematic Response of Monetary Policy".
Cited by 0Open access
Claudia de las Mercedes Rodríguez Ponce, Rene Santaolaya Salgado, Blanca Dina Valenzuela Robles et al. · International Journal of Combinatorial Optimization Problems and Informatics. · 2026
The accumulation of technical debt in legacy software is primarily manifested through the recurring need to modify code in order to correct defects, adapt it to new requirements, or alter existing functionalities. This situation reduces the useful lifespan of software systems by increasing their fragility, as a consequence of the continuous interventions required for maintenance and correction. Code smells constitute one of the principal contributors to this problem in software development. Consequently, it is essential to address this issue through appropriate design and coding practices in order to enhance code quality, maintainability, and readability. The objective of the present research is to identify the relative frequency of practices employed by developers during the design and coding phases that lead to the generation of malformed code, thereby contributing to entropy and increased technical debt in legacy systems. The study focuses on evaluating factors related to structural design and programming practices, while also conducting statistical analyses to identify recurring patterns associated with code smells. The results provide a detailed understanding of the most common practices and their impact on software quality. Furthermore, they highlight relevant gaps in current software engineering research and development, suggesting potential directions for future investigations aimed at addressing these shortcomings. Smart citations: https://scite.ai/reports/10.61467/2007.1558.2025.v16i1.947 Dimensions.Open Alex.
Cited by 0Open access
Fengji Zhang, Zexian Zhang, Jacky Keung et al. · SSRN Electronic Journal · 2024
Cited by 1Open access
United States, National Risk Management, D. A. Kanzer · 1997
Effective use of ground-water simulation codes as management decision tools requires the establishment of their functionality, performance characteristics, and applicability to the problems at hand. This is accomplished through systematic code-testing protocol and code selection strategy. The protocol contains two main elements: functionality analysis and performance evaluation. Functionality analysis is the description and measurement of the capabilities of a simulation code; performance evaluation concerns the appraisal of a code’s operational characteristics (e.g., computational accuracy and efficiency, sensitivity for problem design and parameter selection, and reproducibility). Testing of qround-water simulation codes may take the form of (1) benchmarking with known, independently derived analytical solutions; (2) intracomparison using different code functions inciting the same system responses; (3) intercomparison with comparable simulation codes; or (4) comparison with field or laboratory experiments. The results of the various tests are analyzed using standardized statistical and graphical techniques to identify performance strengths and weaknesses of code and testing procedures. The protocol is demonstrated and evaluated using a three-dimensional finite difference flow and solute transport simulation code, FTWORK.
Cited by 0
Atik Ul Mussanadah, Muhammad Adlin Saputra, Hasan Mukhibad et al. · Zenodo (CERN European Organization for Nuclear Research) · 2026
Dataset supporting the systematic review on accounting information system governance (bibliographic metadata and author-generated analysis) entitled "Code, Control, and Performance: A Systematic Review on AIS-Governance Nexus"
Cited by 0Open access