Software Engineering Knowledge Research Articles

Representation Learning for Stack Overflow Posts: How Far Are We?

The tremendous success of Stack Overflow has accumulated an extensive corpus of software engineering knowledge, thus motivating researchers to propose various solutions for analyzing its content. The performance of such solutions hinges significantly on the selection of representation models for Stack Overflow posts. As the volume of literature on Stack Overflow continues to burgeon, it highlights the need for a powerful Stack Overflow post representation model and drives researchers’ interest in developing specialized representation models that can adeptly capture the intricacies of Stack Overflow posts. The state-of-the-art (SOTA) Stack Overflow post representation models are Post2Vec and BERTOverflow, which are built upon neural networks such as convolutional neural network and transformer architecture (e.g., BERT). Despite their promising results, these representation methods have not been evaluated in the same experimental setting. To fill the research gap, we first empirically compare the performance of the representation models designed specifically for Stack Overflow posts (Post2Vec and BERTOverflow) in a wide range of related tasks (i.e., tag recommendation, relatedness prediction, and API recommendation). The results show that Post2Vec cannot further improve the SOTA techniques of the considered downstream tasks, and BERTOverflow shows surprisingly poor performance. To find more suitable representation models for the posts, we further explore a diverse set of transformer-based models, including (1) general domain language models (RoBERTa, Longformer, and GPT2) and (2) language models built with software engineering related textual artifacts (CodeBERT, GraphCodeBERT, seBERT, CodeT5, PLBart, and CodeGen). This exploration shows that models like CodeBERT and RoBERTa are suitable for representing Stack Overflow posts. However, it also illustrates the “No Silver Bullet” concept, as none of the models consistently wins against all the others. Inspired by the findings, we propose SOBERT, which employs a simple yet effective strategy to improve the representation models of Stack Overflow posts by continuing the pre-training phase with the textual artifact from Stack Overflow. The overall experimental results demonstrate that SOBERT can consistently outperform the considered models and increase the SOTA performance significantly for all the downstream tasks.

MR2-KG: A multi-relation multi-rationale knowledge graph for modeling software engineering knowledge on Stack Overflow

MR²-KG: A multi-relation multi-rationale knowledge graph for modeling software engineering knowledge on Stack Overflow

Modeling software processes from different domains using SPEM and BPMN notations: An experience report of teaching software processes

In a current application development scenario in different environments, technologies and contexts, such as IoT, Blockchain, Machine Learning and Cloud Computing, there is a need for particular solutions for domain-specific software development processes. The proper definition of software processes requires the understanding for the involved teams and organization’s particularities and specialized technical knowledge in Software Engineering. Although it is an essential part of Software Engineering, many university curricula do not dedicate as much effort to teaching software processes, focusing more on the basic principles of Software Engineering, such as requirements, architecture and programming languages. Another important aspect of software processes is modeling. The modeling of a software process provides a basis for managing, automating and supporting the software process improvement. In this context, teaching software process modeling becomes challenging, mainly due to the great emphasis on theory and few practices. This work presents an experience report teaching the definition and modeling of software processes in different domains. In the discipline of software processes, we applied a practice for defining and modeling processes in various application domains, such as: IoT, cloud, mobile, critical systems, self-adaptive systems, machine learning, blockchain and games. The processes were modeled in the Software Systems Process Engineering Metamodel (SPEM) and Business Process Model and Notation (BPMN) notations based on references from the literature for each domain. We evaluated the process modeling practice with the SPEM and BPMN in two classes of the software processes discipline, and we had discussions about the use of the two notations applied to the different domains. In general, students reported good experiences in defining processes, highlighting the importance of practical modeling applications for professional life. As the main results of the study in teaching process modeling, we have that: (i) students accepted HEFLO tool better than EPF Composer tool; (ii) most students are not aware of specific domains and that anticipating the study of these domains in the discipline is a good strategy; and, (iii) the students also highlighted the need for more support for the two notation tools.

Towards Automation of the FORM/BCS Method

The software industry is facing more complex computer systems, with short development and sustainability issues. To deliver good software with these constraints, software reuse has become a central concept for minimizing design and realization costs. This study improves upon Feature-Oriented Reuse Method with Business Component Semantics (FORM/BCS), a software development method that produces adaptable architectures from reusable domain components. This is a promising method for reusable software assets and model creation. The objective of the FORM/BCS is to bring the industrial production chain to the software. This study proposes a model to automatically transform the FORM/BCS business subsystem component into a process business component. Two metamodels for business subsystems and process business components were developed. In addition, this study establishes correspondences between the source metamodel and target metamodel classes, transformation rules, and the instance of the source metamodel and generates the target metamodel instance. Detailed findings can help practitioners reduce software design costs and development time, and contribute to the advancement of knowledge in software engineering.

Enhancing learning and exploratory search with concept semantics in online healthcare knowledge management systems: An interactive knowledge visualization approach

Healthcare is a knowledge intensive process, which requires large amount of information from various healthcare knowledge providers. However, this growing information makes challengeable for users in finding useful healthcare information to support their healthcare decision making. Knowledge management and knowledge management systems are drawing increasing attention in healthcare community. This paper proposes an online healthcare knowledge management framework and system prototype, featured with integrating knowledge learning and exploratory search for solving problems efficiently. Inspired by the knowledge to action framework in healthcare and the iterative process in software engineering, an iterative and incremental knowledge management model is designed to organize the activities during the knowledge learning and searching procedure. Supported by the knowledge-enhanced infrastructure of online healthcare information designed, domain knowledge learning and task knowledge learning environment can work collaboratively and interactively to improve the result of learning and exploratory search. Knowledge base also provides helpful information in an interactive way for solving problems. Given exploiting probabilistic information retrieval model and concept semantics described in Description Logics, a search intent discovery method and search intent guided learning and searching mechanism are designed within this environment for achieving the learning and search goal. The results of the experiments have proved the efficacy of the methods and process model used in the leaning-focused healthcare knowledge management system.

A reference implementation for knowledge assisted robot development for planetary and orbital robotics

The use of modular, yet reusable and interchangeable components offers flexibility for system designers and thus leads to a higher degree of reconfigurability with cost effective, adaptable and scalable solutions. The development of a robotic system remains, however, a complex task with increasing demands depending on the targeted capabilities, functionalities and operational constraints of the robot. This requires expert knowledge in mechanical, electrical and software engineering, as well as tools that can collect, process and use relevant data. As a result, the amount of available and managed information is increasing, while its relevance, usability, and knowledge that can be extracted from it are consequently decreasing: a situation described as information overload paradox.The design of a modular system can be based on the use of standardized modules with universal interconnects, but it should be aligned with a strong formalism, e.g. the semantic representation of components, to allow for a better management of the overall complexity. An ontology-based knowledge representation, for instance, which provides domain and application specific knowledge for the design of robotic systems, can be used as a method of storing and sharing data in a uniform, machine-readable and standardized way.In this paper, we introduce the Knowledge-based Open Robot voCabulary as Utility Toolkit (korcut) as a core component to support the ontology-driven development of robotic systems as part of a reference implementation of the Q-Rock development cycle. Q-Rock gathers methods to assist users in designing robot configuration by: (a) automatically exploring robot capabilities based on robot hardware, (b) proposing robot designs to meet the user’s needs, (c) refining the proposed robot designs. We use korcut to improve the robot design process in orbital and planetary robotics from the requirements definition to the development phases of complex modular robotic structures (e.g. composition of an unmanned ground vehicle with robotic arms). This is implemented by embedding semantic component descriptions in a state-of-the-art open-source 3D modeling software. Furthermore, the korcut ontology family includes various sub-ontologies to address specific space-related tasks, such as the modeling of a Standard Interconnect (SI) for On Orbit Services (OOS) and Orbital Factory. The ontology design follows criteria that have been derived from a survey conducted as part of this work. This paper also covers an evaluation of its current applications from a methodological, knowledge representation, and software tool perspectives. In the application part, we introduce the use of the ontology to model a multi-functional SI that enables mechanical connections between various (modular) robotic components, as well as transfer of power and data. Finally, we provide a critical analysis of our work and outline its future work.

Undergraduate research in software engineering. An experience and evaluation report  

The purpose of this paper is to present an undergraduate research experience process model and the evaluation of seven years of its application in an undergraduate research program in software engineering. Undergraduate students who participated in research projects between 2015 and 2022 were surveyed to find out a) their motivations for participating in research projects in software engineering, b) the skills they consider they have acquired or improved by participating in those projects, and c) their perception of benefits and utility for their future work and professional activities. Results reveal that participation in real research projects in software engineering is highly valued by undergraduate students, who perceive benefits in the development of research and soft skills, and for their future professional activity. In addition, these undergraduate research projects and the process followed show that it is feasible to make original contributions to the body of knowledge of software engineering.

Exploring the Intersection between Software Maintenance and Machine Learning—A Systematic Mapping Study

While some areas of software engineering knowledge present great advances with respect to the automation of processes, tools, and practices, areas such as software maintenance have scarcely been addressed by either industry or academia, thus delegating the solution of technical tasks or human capital to manual or semiautomatic forms. In this context, machine learning (ML) techniques play an important role when it comes to improving maintenance processes and automation practices that can accelerate delegated but highly critical stages when the software launches. The aim of this article is to gain a global understanding of the state of ML-based software maintenance by using the compilation, classification, and analysis of a set of studies related to the topic. The study was conducted by applying a systematic mapping study protocol, which was characterized by the use of a set of stages that strengthen its replicability. The review identified a total of 3776 research articles that were subjected to four filtering stages, ultimately selecting 81 articles that were analyzed thematically. The results reveal an abundance of proposals that use neural networks applied to preventive maintenance and case studies that incorporate ML in subjects of maintenance management and management of the people who carry out these tasks. In the same way, a significant number of studies lack the minimum characteristics of replicability.

OSS in Software Engineering Education

Software Engineering is a crucial topic in undergraduate computing-related courses and provides the basic knowledge and skills necessary for professional practice in the software industry. Teaching Software Engineering principles, concepts, and practices and relating them to real-world scenarios are challenging tasks, and the adoption of Open Source Software (OSS) projects can help to face these challenges. On the other hand, adopting OSS projects as a didactic resource may introduce additional challenges to instructors who are not familiar with the OSS ecosystem. Objective: In this paper, we identified and mapped the profiles of instructors of Software Engineering courses concerning their classroom practices and use of OSS projects in Software Engineering Education. Method: We surveyed 90 higher education instructors in Brazil to collect data regarding their familiarity with the Software Engineering knowledge areas, pedagogical methods and resources used, and familiarity with and use of OSS projects in the classroom. Then, we resorted to data mining techniques, for instance, K-modes and Decision Tree algorithms, to identify instructors’ characteristics according to their classroom practices and use of OSS projects in the course activities. Results: Our findings include the characterization of instructors who use and instructors that do not use OSS projects in Software Engineering Education and the grouping of instructors after the application of the K-modes algorithm, and after the application of the Decision Tree algorithm, with similar characteristics of the pedagogical practices. The main result of this work is that the familiarity with OSS projects and the use of active learning methods were characteristics present in the application of the K-modes and Decision Tree algorithms, that distinguished instructors who used OSS projects from those that did not use them in Software Engineering Education. Finally, we confirmed that familiarity with OSS projects could have a positive influence on the instructors’ interest and potential for adopting this approach in Software Engineering Education.

App development in a sports science setting: A systematic review and lessons learned from an exemplary setting to generate recommendations for the app development process.

The digital health sector is rapidly growing. With only 4% of publishers out of academic settings, it is under-represented in app development. The objective of this study is to assess the current state of app development with a systematic review and a survey within an exemplary academic setting along the following research questions: (Q1) Are software engineering principles sufficiently known in the sports science app development context? (Q2) Is the role of sports scientists in the context of app development sufficiently understood? The systematic review was conducted by two independent reviewers within databases Pubmed, Scopus, Web of Science, and IEEE Xplore. The PICO schema was used to identify the search term. We subtracted information about five main topics: development process, functional requirements and features, security, technology, and dissemination. The survey was developed by a multidisciplinary team and focused on five main topics. Out of 701 matches, 21 were included in the review. The development process was only described in seven studies. Functional requirements and features were considered in 11 studies, security in 3, technology in 13, and dissemination in 12 with varying details. Twelve respondents [mean age 33(7) years, 58% women] replied to the survey. The survey revealed limited knowledge in realization of security measures, underlying technology and source code management, and dissemination. Respondents were able to provide input on development processes as well as functional requirements and features. The involvement of domain experts is given in seven review studies and described in two more. In 50% of survey respondents, the role in app development is defined as a research assistant. We conclude that there is a varying degree of software engineering knowledge in the sports science app development context (Q1). Furthermore, we found that the role of sports scientists within app development is not sufficiently defined (Q2). We present recommendations for improving the success probability and sustainability of app development and give orientation on the potential roles of sports scientists as domain experts. Future research should focus on the generalizability of these findings and the reporting of the app development process.

Development of an app to monitor queues at an ophthalmological institute

Applications in the health area, known as these m-health, are important accessories for the management of productive systems, especially with the increase in demand for health services. Therefore, this study aims to develop an application for monitoring queues in an ophthalmology institute located in the São Francisco Valley region. The work performed as prototyping steps with Figma, analysis of the application development with Flutter/Dart and Firebase and evaluation of the management of the alpha version. The evaluation was based on the MARS scale, which obtained the dimension “functionality” with the highest evaluation (3.75) and “aesthetics” as the lowest (3.0). The overall quality was indicated as (3.70), above recommended (3.70). It is concluded that it was possible to develop an application using knowledge and techniques of Software Engineering and Queue Management.

Effects of Flipped Learning Approaches on Students’ Learning Performance in Software Engineering Education

Software engineering education plays an important role in keeping students educated with software technologies, processes, and practices that are needed by industries. Nevertheless, the nature of software engineering learning activities in traditional classrooms is limited in scope and time, making it more difficult to achieve a proper balance between theory and practice and address industrial demands. This makes scant provision for assisting students in keeping their software engineering knowledge current. To support software engineering education, flipped learning is a suitable strategy. Prior studies have shown that students’ perceptions in flipped learning environments are better than those in traditional learning environments. Nevertheless, in flipped learning, students may not have sufficient ability to conduct learning out of class. Therefore, the flipped learning strategy should aim to meet the needs of students to ensure that they get the appropriate support or feedback during the learning process before the class. The aim of this study was to propose a flipped learning diagnosis approach to promote students’ learning out of class in the flipped classroom. To explore students’ learning performance in software engineering courses, three classes of students were invited to learn with three different learning approaches (traditional learning approach, flipped learning approach, and flipped learning diagnosis approach). The results showed that the students who learned with the flipped learning diagnosis approach outperformed those students who learned with the flipped learning approach or the traditional learning approach.

A case study of industry–academia communication in a joint software engineering research project

AbstractEmpirical software engineering research relies on good communication with industrial partners. Conducting joint research both requires and contributes to bridging the communication gap between industry and academia (IA) in software engineering. This study aims to explore communication between the two parties in such a setting. To better understand what facilitates good IA communication and what project outcomes such communication promotes, we performed a case study, in the context of a long‐term IA joint project, followed by a validating survey among practitioners and researchers with experience of working in similar settings. We identified five facilitators of IA communication and nine project outcomes related to this communication. The facilitators concern the relevance of the research, practitioners' attitude and involvement in research, frequency of communication and longevity of the collaboration. The project outcomes promoted by this communication include, for researchers, changes in teaching and new scientific venues, and for practitioners, increased awareness, changes to practice, and new tools and source code. Besides, both parties gain new knowledge and develop social‐networks through IA communication. Our study presents empirically based insights that can provide advise on how to improve communication in IA research projects and thus the co‐creation of software engineering knowledge that is anchored in both practice and research.

Software Engineering Applications Enabled by Blockchain Technology: A Systematic Mapping Study

The novel, yet disruptive blockchain technology has witnessed growing attention, due to its intrinsic potential. Besides the conventional domains that benefit from such potential, such as finance, supply chain and healthcare, blockchain use cases in software engineering have emerged recently. In this study, we aim to contribute to the body of knowledge of blockchain-oriented software engineering by providing an adequate overview of the software engineering applications enabled by blockchain technology. To do so, we carried out a systematic mapping study and identified 22 primary studies. Then, we extracted data within the research type, research topic and contribution type facets. Findings suggest an increasing trend of studies since 2018. Additionally, findings reveal the potential of using blockchain technologies as an alternative to centralized systems, such as GitHub, Travis CI, and cloud-based package managers, and also to establish trust between parties in collaborative software development. We also found out that smart contracts can enable the automation of a variety of software engineering activities that usually require human reasoning, such as the acceptance phase, payments to software engineers, and compliance adherence. In spite of the fact that the field is not yet mature, we believe that this systematic mapping study provides a holistic overview that may benefit researchers interested in bringing blockchain to the software industry, and practitioners willing to understand how blockchain can transform the software development industry.

A systematic literature review on Technical Debt prioritization: Strategies, processes, factors, and tools

BackgroundSoftware companies need to manage and refactor Technical Debt issues. Therefore, it is necessary to understand if and when refactoring of Technical Debt should be prioritized with respect to developing features or fixing bugs. ObjectiveThe goal of this study is to investigate the existing body of knowledge in software engineering to understand what Technical Debt prioritization approaches have been proposed in research and industry. MethodWe conducted a Systematic Literature Review of 557 unique papers published until 2020, following a consolidated methodology applied in software engineering. We included 44 primary studies. ResultsDifferent approaches have been proposed for Technical Debt prioritization, all having different goals and proposing optimization regarding different criteria. The proposed measures capture only a small part of the plethora of factors used to prioritize Technical Debt qualitatively in practice. We present an impact map of such factors. However, there is a lack of empirical and validated set of tools. ConclusionWe observed that Technical Debt prioritization research is preliminary and there is no consensus on what the important factors are and how to measure them. Consequently, we cannot consider current research conclusive. In this paper, we therefore outline different directions for necessary future investigations.

Analysis of 13 implementations of the software engineering management and engineering basic profile guide of ISO/IEC 29110 in very small entities using different life cycles

AbstractWe are living in an age of growing demand for software products. This growing demand creates opportunities for very small entities (VSEs) to not only survive but also flourish. In this context, VSEs need to produce high‐quality products to meet market needs. However, in their quest to produce high‐quality software, VSEs need to overcome the challenge of implementing international standards, which they find difficult to do because of lack of knowledge and practical experience. This paper provides an analysis performed to 13 teams of VSEs, using different life cycles, which achieved the implementation of the ISO/IEC 29110, to analyze the effort each team invests to implement the best practices provided by the standard. Besides, the paper provides an analysis of the difficulties, and the benefits of using the six‐step method are included. This analysis is of interest because software engineering knowledge developed by researchers should be transferred to the industry to reduce the gap between them. The results highlight, on the one hand, the practices representing more effort for teams by life cycle. On the other hand, the results highlight the six‐step method that allowed the 13 teams to achieve a high level of coverage of the ISO/IEC 29110 international standard.

Ontologic Design of Software Engineering Knowledge Area Knowledge Components

Ontologic Design of Software Engineering Knowledge Area Knowledge Components

Sustainable Software Engineering:A Perspective of Individual Sustainability

Sustainable software engineering is a mean of developing sustainable software with sustainable software engineering process activities while balancing its various dimensions for instance economic, environmental, social, technical and individual. It is conveyed that the economic, technical, environmental and social dimensions are explored to satisfactory degree however the individual dimension of sustainable software engineering which is concerned with wellbeing of software engineers is not explored to satisfactory degree with respect to its understanding and challenges. Therefore, the aim of the study is to highlight and prioritize the challenges regarding individual sustainability dimension. The study also provides the mitigation strategies for the top five individual sustainability challenges. The systematic literature review has been performed to report the challenges and mitigation strategies. The study finding shows that lack of domain knowledge, lack of methodologies and tool support, lack of education, varying and unidentified situations and lack of sustainable software engineering practices are top most challenges regarding individual sustainability. These challenges need an urgent attention to achieve the goal of sustainable software engineering. The study also reports various mitigation strategies to overcome the risk of identified top most individual sustainability challenges such as to introduce sustainable software engineering education and knowledge in software engineering curricula, development of knowledge sharing frameworks and awareness regarding unclear and varying situations for each software engineering activity etc. The study will be beneficial for sustainable software engineering body of knowledge, sustainable software engineering practitioners and researchers by providing classified list of individual sustainability challenges and their mitigation strategies.

Software Engineering for Smart Cyber-Physical Systems (SEsCPS 2018) - Workshop Report

Smart Cyber-Physical Systems (sCPS) are a novel kind of Cyber- Physical System engineered to take advantage of large-scale cooperation between devices, users and environment to achieve added value in the face of uncertainty and changing environments. Examples of sCPS include modern traffic systems, Industry 4.0 systems, systems for smart buildings, and smart energy grids. The uniting aspect of all these systems is that to achieve their high level of intelligence, adaptivity and ability to optimize and learn, they rely heavily on software. This makes them software-intensive systems, where software becomes their most complex part. Engineering sCPS thus becomes a recognized software engineering discipline, which, due to specifics of sCPS, can only partially rely on the existing body of knowledge in software engineering. In fact, it turns out that many of the traditional approaches to architecture modeling and software development fall short in their ability to cope with the high dynamicity and uncertainty of sCPS. This calls for innovative approaches that jointly reflect and address the specifics of such systems. This paper maps the discussions and results of the Fourth International Workshop on Software Engineering for Smart Cyber-Physical Systems (SEsCPS 2018), which focuses on challenges and promising solutions in the area of software engineering for sCPS.

Ten simple rules for helping newcomers become contributors to open projects.

To survive and thrive, a must attract new members, retain them, and help them be productive [1]. As openness becomes the norm in research, software development, and education, knowing how to do this has become an essential skill for principal investigators and managers alike. A growing body of knowledge in sociology, anthropology, education, and software engineering can guide decisions about how to facilitate this. What exactly do we mean by community? In the case of open source and open science, the most usual meaning is a community of practice. As defined by Lave and Wenger [2, 3], groups as diverse as knitting circles, oncology researchers, and web designers share three key characteristics: Participants have a common product or purpose that they work on or toward. They are mutually engaged, i.e., they assist and mentor each another. They develop shared resources and domain knowledge. Brown [4] specializes this to define a community of as …a formed in pursuit of a common goal. The goal can be definite or indefinite in time, and may not be clearly defined, but it is something that (generally speaking) the is aligned on. People working to preserve coral reefs in the face of global climate change are an example of such a community. No central organization coordinates their work, but the scientists who study coral reefs, the environmentalists who work to protect them, and the citizens who support them financially and politically are aware of each other’s efforts, collaborate in ad hoc ways, and are conscious of contributing toward a shared purpose. Open-source software projects are also communities of effort. E.g., the Mozilla Firefox [5] includes a mix of paid professionals, highly involved volunteers, and occasional contributors who not only create software, documentation, and tutorials but also organize events, answer questions in online forums, mentor newcomers, and advocate for open standards. Every of effort has unique features, but they have enough in common to profit from one another’s experience. The 10 rules laid out below are based on studies of such communities and on the authors’ experience as members, leaders, and observers. Our focus is on small and medium-sized projects, i.e., ones that have a handful of to a few hundred participants and are a few months to a few years old but may not (yet) have any formal legal standing, such as incorporation as a nonprofit.