Software Engineering Research Articles

The role of firm AI capabilities in generative AI-pair coding

ABSTRACT Generative Artificial Intelligence (genAI) is the latest evidence of the transformative value of AI in organisations. One promising avenue is in software engineering, where genAI can contribute to coding by pairing with developers. Based on a sample of global companies, two key findings emerge from an analysis of the productivity impact of genAI pair coding. Coding quality is negatively correlated with productivity throughput gains, while quality-adjusted productivity gains depend on the extent to which firms have deployed AI capabilities in the form of data, skills upgrading and AI governance. As observed with other digital technologies, the success of using genAI is closely linked to complementary technical skills and organisational resources.

The Role of Artificial Intelligence in Modern Software Engineering

Abstract. The rapid advancement of Artificial Intelligence (AI) has significantly influenced various industries, including software engineering. This paper explores the integration of AI into software engineering, focusing on its applications across different stages of the software development life cycle, including design, development, testing, project management, and maintenance. AI's ability to automate tasks, enhance efficiency, and improve code quality is revolutionizing how software is built and maintained. The paper also addresses the challenges and risks associated with AI-driven software engineering, such as dependency on AI tools, ethical concerns, and security vulnerabilities. Finally, the paper highlights future trends in AI-powered software engineering, including adaptive and self-healing systems, AI-enhanced collaboration, and full software automation. The role of AI in shaping the future of software engineering is both profound and transformative, making it a critical area of study.

Foundational Competencies and Responsibilities of a Research Software Engineer

The term Research Software Engineer, or RSE, emerged a little over 10 years ago as a way to represent individuals working in the research community but focusing on software development. The term has been widely adopted and there are a number of high-level definitions of what an RSE is. However, the roles of RSEs vary depending on the institutional context they work in. At one end of the spectrum, RSE roles may look similar to a traditional research role. At the other extreme, they resemble that of a software engineer in industry. Most RSE roles inhabit the space between these two extremes. Therefore, providing a straightforward, comprehensive definition of what an RSE does and what experience, skills and competencies are required to become one is challenging. In this community paper we define the broad notion of what an RSE is, explore the different types of work they undertake, and define a list of foundational competencies as well as values that outline the general profile of an RSE. Further research and training can build upon this foundation of skills and focus on various aspects in greater detail. We expect that graduates and practitioners will have a larger and more diverse set of skills than outlined here. On this basis, we elaborate on the progression of these skills along different dimensions. We look at specific types of RSE roles, propose recommendations for organisations, give examples of future specialisations, and detail how existing curricula fit into this framework.

Exploring job competency related to intelligent construction in China using a text mining method

PurposeThe construction industry is experiencing digital transformation, which is also defined as intelligent construction. With the rise of intelligent construction, job characteristics are changing rapidly. Current knowledge about job competencies required by intelligent construction is lacking. Therefore, the aim of this paper is to explore job competencies related to intelligent construction by text mining recruitment information. It is expected to reveal the trend of talent development for the intelligent construction industry.Design/methodology/approachA total of 375 job advertisements regarding the demanding professionals and industrial workers related to intelligent construction were collected and analyzed to reveal the demands of the current labor market. Different job posts related to intelligent construction were classified into 11 categories. Job competencies were extracted and analyzed using the latent Dirichlet allocation (LDA) model, frequency–inverse document frequency (TF-IDF) algorithm and k-means cluster analysis method. The text mining results identified 10 job competencies.FindingsCurrently, there was a high demand for high-tech talents in the labor market related to intelligent construction. Those high-tech job posts, such as software engineers and R&D staff, required digital technology, R&D skills, electrical automation knowledge and programming capability. Current employees demanding for intelligent construction are expected to be capable of both using information technology and having a general knowledge of the construction industry.Originality/valueThrough text mining of current job advertisements, the overall demand for compound talents in the labor market of intelligent construction were explored. The results provide empirical reference for personnel training and talent cultivation in the development of intelligent construction. Higher educational institutions, human resources professionals, as well as experts that are already employed or aspire to be employed in intelligent construction companies, would benefit from the results of our analysis.

Challenges and Opportunities in Cloud Computing Education

The complexity of cloud computing development often presents significant challenges for junior software engineers. Our mixed-method study examines the challenges, motivations, and importance of acquiring knowledge and practical experience in cloud computing for junior software engineers before they enter the job market in Saudi Arabia. Furthermore, our work highlights the necessity of revising traditional educational approaches by integrating practical projects into the curriculum, as well as addressing challenges related to cost and accessibility.

Classifying software security requirements into confidentiality, integrity, and availability using machine learning approaches

Security requirements are considered one of the most important non-functional requirements of software. The CIA (confidentiality, integrity, and availability) triad forms the basis for the development of security systems. Each dimension is expressed as having many security requirements that should be designed, implemented, and tested. However, requirements are written in a natural language and may suffer from ambiguity and inconsistency, which makes it harder to distinguish between different security dimensions. Recognizing the security dimensions in a requirements document should facilitate tracing the requirements and ensuring that a dimension has been implemented in a software system. This process should be automated to reduce time and effort for software engineers. In this paper, we propose to classify the security requirements into CIA triads using Term frequency-inverse document frequency and sentence-transformer embedding as two different technologies for feature extraction. For both techniques, we developed five models by using five well-known machine learning algorithms: (1) support vector machine (SVM), (2) K-nearest neighbors (KNN), (3) Random Forest (RF), (4) gradient boosting (GB), and (5) Bernoulli Naive Bayes (BNB). Also, we developed a web interface that facilitates real-time analysis and classifies security requirements into CIA triads. Our results revealed that SVM with the sentence-transformer technique outperformed all classifiers by 87% accuracy in predicting a type of security dimension.

Determining the Reliability of Software Electronic Engineering Tools in the Control of Vestibular Disorders in Inclusive Physical Education of Students

Background. Due to the increase in the proportion of mine-blast injuries sustained during the Russian-Ukrainian war, the number of students in higher education institutions after blast TBI has risen. Objectives. The study aimed to determine the reliability of the tool for test control of gait parameters, created based on software electronic engineering for students after blast TBI in inclusive physical education. Materials and methods. The theoretical stage of the research was realized by analyzing the literature on the studied problem, evaluating and synthesizing information from scientific sources. For the empirical part of the experiment, the Dynamic Gait Index test was used to assess gait parameters, dynamic balance, and fall risk. The test results were processed using the methods of statistical mathematical analysis. The experiment involved 30 students (men) after sustaining an explosive traumatic brain injury. Results. The study presents the developed tool for testing gait parameters using the Dynamic Gait Index, which is based on the integration of various types of software electronic engineering services. The development uses motion, speed, pressure, and touch sensors. The information about the test exercise, recorded by the sensors, is sequentially transmitted by wireless communication means to the controller (Arduino Leonardo based on the ATmega32u4 microcontroller platform) and then to a PC, where the results are displayed on the monitor in digital and graphical form. The development uses a neural network and the Microsoft OneDrive cloud service to process information, which provides processing and storage of large amounts of test control data. The results of the empirical establishment of the Dynamic Gait Index quality measure, the characteristics of reliability and validity that were recorded by the teacher, are significantly inferior in numerical values to the correlation coefficient obtained by the results of the Dynamic Gait Index when the outcomes were recorded using the developed tool. Conclusions. The introduction of modern software engineering services into inclusive physical education of students represents a significant step forward in implementing an impartial control mechanisms for restoring damaged gait and balance functions, which have been recognized as a leading factor in disability caused by blast TBI, and contributes to ensuring the effectiveness of this process.

A Day in the Life of a Food Packaging Engineer

A Day in the Life of a Food Packaging Engineer

Creating and Validating a Ground Truth Dataset of Unified Modeling Language Diagrams Using Deep Learning Techniques

UML (Unified Modeling Language) diagrams are graphical representations used in software engineering which play a vital role in the design and development of software systems and various engineering processes. Large, good-quality datasets containing UML diagrams are essential for different areas in the industry, research, and teaching purposes; however, few exist in the literature and it is common to find duplicate elements in the existing datasets. This might affect the evaluation of the models obtained when using these datasets. This paper addresses the challenge of creating a ground truth dataset of UML diagrams, including semi-automated inspection to remove duplicates and ensuring the correct labeling of all UML diagrams contained in the dataset. In particular, a dataset of six UML diagram classes was assembled, comprising a total of 2626 images (426 activity diagrams, 636 class diagrams, 352 component diagrams, 357 deployment diagrams, 435 sequence diagrams, and 420 use case diagrams). Importantly, unlike other existing datasets, ours contains no duplicate elements and all diagrams are correctly labeled. Our curated dataset is a valuable and unique resource for the research community, serving as a foundation for training and evaluating diverse artificial intelligence models. In this paper, we demonstrate this by training and testing several deep learning models using our dataset, achieving highly satisfactory results compared to those presented in other works in the literature. Additionally, our experimental results highlight the potential of visual transformers for UML diagram classification, setting our approach apart from others that predominantly used convolutional neural networks for similar tasks.

Applying bayesian data analysis for causal inference about requirements quality: a controlled experiment

It is commonly accepted that the quality of requirements specifications impacts subsequent software engineering activities. However, we still lack empirical evidence to support organizations in deciding whether their requirements are good enough or impede subsequent activities. We aim to contribute empirical evidence to the effect that requirements quality defects have on a software engineering activity that depends on this requirement. We conduct a controlled experiment in which 25 participants from industry and university generate domain models from four natural language requirements containing different quality defects. We evaluate the resulting models using both frequentist and Bayesian data analysis. Contrary to our expectations, our results show that the use of passive voice only has a minor impact on the resulting domain models. The use of ambiguous pronouns, however, shows a strong effect on various properties of the resulting domain models. Most notably, ambiguous pronouns lead to incorrect associations in domain models. Despite being equally advised against by literature and frequentist methods, the Bayesian data analysis shows that the two investigated quality defects have vastly different impacts on software engineering activities and, hence, deserve different levels of attention. Our employed method can be further utilized by researchers to improve reliable, detailed empirical evidence on requirements quality.

Evaluating an Ethical Hacking Module: Case of a University in South Africa

The University of Johannesburg has Ethical Hacking content covered within the Information Security in the WWW module. This study was conducted at the University of Johannesburg, with honours students registered at the Academy of Computer Science and Software Engineering. This study aims to evaluate the Information Security in the WWW module. Specifically, it seeks to evaluate the content, teaching and learning, and extra activities such as participating in the first round of the National Cybersecurity Hackathon, through student perspectives. It also seeks to understand the views of the students on what they like the most about the learning module as a basis for continuous improvement. The instrument utilised was an amalgamation of module evaluation questionnaires in the body of knowledge. It was modified to suit the context of the study. The online questionnaire had both open-ended and closed questions that seek to get a better understanding of the concepts presented. The frequency distribution, percentage distribution, and weighted mean were calculated to offer the level of agreement and satisfaction. The findings suggest that students were satisfied with the content, teaching and learning, in the Information Security in the WWW module. The results further highlight some essential insights from the respondents that lecturers may consider when improving the instructional material. Further, considerations have been put forward for future improvement of the learning material.

Motivations, Challenges, Best Practices, and Benefits for Bots and Conversational Agents in Software Engineering: A Multivocal Literature Review

Bots are software systems designed to support users by automating specific processes, tasks, or activities. When these systems implement a conversational component to interact with users, they are also known as conversational agents or chatbots . Bots—particularly in their conversation-oriented version and AI-powered—have seen increased adoption over time for software development and engineering purposes. Despite their exciting potential, which has been further enhanced by the advent of Generative AI and Large Language Models, bots still face challenges in terms of development and integration into the development cycle, as practitioners report that bots can add difficulties rather than provide improvements. In this work, we aim to provide a taxonomy for characterizing bots, as well as a series of challenges for their adoption in software engineering, accompanied by potential mitigation strategies. To achieve our objectives, we conducted a multivocal literature review , examining both research and practitioner literature. Through such an approach, we hope to contribute to both researchers and practitioners by providing (i) a series of future research directions to pursue, (ii) a list of strategies to adopt for improving the use of bots for software engineering purposes, and (iii) fostering technology and knowledge transfer from the research field to practice—one of the primary goals of multivocal literature reviews.

AI‐Augmented Software Engineering: Revolutionizing or Challenging Software Quality and Testing?

ABSTRACTWith organizations seeking faster, cheaper, and smarter ways of delivering higher quality software, many are looking towards generative artificial intelligence (AI) to drive efficiencies and innovation throughout the software development lifecycle. However, generative AI can suffer from several fundamental issues, including a lack of traceability in concept generation and decision‐making, the potential for making incorrect inferences (hallucinations), shortcomings in response quality, and bias. Quality engineering (QE) has long been utilized to enable more efficient and effective delivery of higher quality software. A core aspect of QE is adopting quality models to support various lifecycle practices, including requirements definition, quality risk assessments, and testing. In this position paper, we introduce the application of QE to AI systems, consider shortcomings in existing AI quality models from the International Organization for Standardization (ISO), and propose extensions to ISO models based on the results of a survey. We also reflect on skills that IT graduates may need in the future, to support delivery of better‐quality AI.

Analysis of Datasets and Large Language Models for Vulnerability Detection in Imperative Programming Language Code

Large language models are machine learning models that enable the classification and generation of both natural language text and code in various programming languages. These models have billions of parameters and are trained on vast datasets. In recent years, such models have been successfully applied to a wide range of tasks in software engineering. The paper presents data on publication activity on the topic under study, which is obtained on the basis of statistical analysis of search results for relevant key queries. In addition, a review of recent publications in the field of using large language models to detect vulnerabilities in program code is carried out, and the results of the analysis of data sets that are used in training neural network models to search for vulnerabilities in program code are presented.

Control Software Engineering Approaches for Cyber-Physical Systems: A Systematic Mapping Study

Cyber-physical systems (CPS), robotics, the Internet of Things (IoT), and automotive systems are integral to modern technology. They are characterized by their safety criticality, accuracy, and real-time control requirements. Control software plays a crucial role in achieving these objectives by managing and coordinating the operations of various sub-systems. This paper presents a novel systematic mapping study (SMS) for control software engineering, analyzing 115 peer-reviewed papers. The study identifies, classifies, and maps existing solutions, providing a comprehensive and structured overview for practitioners and researchers. Our contributions include (i) a unique classification of literature into six research themes—engineering phases, engineering approaches, engineering paradigms, engineering artefacts, target application domains, and engineering concerns; (ii) insights into the specificity of approaches to target technologies and phases; (iii) the prominence of model-driven approaches for design and testing; (iv) the lack of end-to-end engineering support in existing approaches; and (v) the emerging role of agile-based methods versus the dominance of waterfall-based methods. This paper’s significance lies in its thorough analysis and the high-level mapping of the solution space, offering new perspectives and a detailed roadmap for future research and innovation in control software engineering. The findings will guide advancements and best practices in the field, underscoring the paper’s impact.

Incremental model transformations with triple graph grammars for multi-version models and multi-version pattern matching

AbstractLike conventional software projects, projects in model-driven software engineering require adequate management of multiple versions of development artifacts, importantly allowing living with temporary inconsistencies. In previous work, we have introduced multi-version models for model-driven software engineering, which allow checking well-formedness and finding merge conflicts for multiple versions of the same model at once. However, situations where different models are linked via automatic model transformations also have to be handled for multi-version models. In this paper, we propose a technique for jointly handling the transformation of multiple versions of a source model into corresponding versions of a target model. This enables the use of a more compact representation that may afford improved execution time of both the transformation and further analysis. Our approach is based on the well-known formalism of triple graph grammars and the aforementioned encoding of model version histories called multi-version models. In addition to batch transformation of an entire history, the technique covers incremental synchronization of changes in the framework of multi-version models. Our solution is complemented by a dedicated pattern matching technique for multi-version models. We show the correctness of our approach with respect to the standard semantics of triple graph grammars and conduct an empirical evaluation to investigate the performance of our technique regarding execution time and memory consumption. Our results indicate that the proposed solution affords lower memory consumption and may improve execution time for batch transformation of large version histories, but can also come with computational overhead in unfavorable cases.

The Impact of Product Packaging Characteristics on Order Picking Performance in Grocery Retailing

ABSTRACTIncreasing labor cost levels and workforce shortages have caused retailers to pay increased attention to their order‐fulfillment operations, which continue to largely depend on manual order picking systems. The operations and logistics management literature suggests that optimizing tertiary packaging, which groups products into full unit loads for storage and shipping, is an important way to improve order picking performance. While most retailers handle products at the level of secondary packaging when fulfilling orders, this packaging level remains largely unexplored. To address this gap, we analyze 3,380,596 picks performed by 185 order pickers of 4957 products in a grocery retail warehouse in Germany. Our findings indicate that secondary packaging characteristics directly affect order picking performance and that this effect is moderated by traditional product characteristics (e.g., product weight and volume), as well as elements of warehouse design (e.g., pick and stack levels). From a managerial perspective, our findings may help to bridge the gap between logistics managers and packaging engineers and provoke further research on the trade‐off between operational and environmental performance.

Retraction Note: Studying efficacy of traditional software quality parameters in quantum software engineering

Retraction Note: Studying efficacy of traditional software quality parameters in quantum software engineering

Research on the design and image perception of cultural landscapes based on digital roaming technology

The preservation and advancement of cultural heritage continue to present significant challenges, arising from conflicting objectives, concepts, methodologies, and limited resources. Image perception not only serves as an observational tool for cultural heritage but also as a research method to uncover deeper meanings and implications. To explore the complex dynamics of cultural heritage and address the risks of its deterioration, this study employed virtual engine software (Unity 2019) to create digital landscapes for immersive exploration. An image perception evaluation framework was established through field surveys and behavioral synthesis experiments across five key areas, including ancient tree ecosystems and architectural sites. Our findings revealed that participants showed the greatest interest in the ancient tree ecological area, indicating a strong attachment to historical continuity and natural progression. Heart rate variability (HRV) analysis further demonstrated that participants experienced positive emotional responses, suggesting that well-designed cultural landscapes can enhance mental well-being and uplift mood. Additionally, participants expressed a strong willingness to discuss the ancient tree eco-region, with image perception playing a crucial role in motivating information sharing and driving conservation efforts. This study introduces an innovative model—Design Practice, Immersive Experience, and Image Perception Evaluation—which offers a new perspective for the preservation and enhancement of cultural heritage sites.

A New Multi-Axial Functional Stress Analysis Assessing the Longevity of a Ti-6Al-4V Dental Implant Abutment Screw.

This study investigates the impact of tightening torque (preload) and the friction coefficient on stress generation and fatigue resistance of a Ti-6Al-4V abutment screw with an internal hexagonal connection under dynamic multi-axial masticatory loads in high-cycle fatigue (HCF) conditions. A three-dimensional model of the implant-abutment assembly was simulated using ANSYS Workbench 16.2 computer aided engineering software with chewing forces ranging from 300 N to 1000 N, evaluated over 1.35 × 107 cycles, simulating 15 years of service. Results indicate that the healthy range of normal to maximal mastication forces (300-550 N) preserved the screw's structural integrity, while higher loads (≥800 N) exceeded the Ti-6Al-4V alloy's yield strength, indicating a risk of plastic deformation under extreme conditions. Stress peaked near the end of the occluding phase (206.5 ms), marking a critical temporal point for fatigue accumulation. Optimizing the friction coefficient (0.5 µ) and preload management improved stress distribution, minimized fatigue damage, and ensured joint stability. Masticatory forces up to 550 N were well within the abutment screw's capacity to sustain extended service life and maintain its elastic behavior.