Software Fault Research Articles

Enhancing software defect prediction: a framework with improved feature selection and ensemble machine learning

Effective software defect prediction is a crucial aspect of software quality assurance, enabling the identification of defective modules before the testing phase. This study aims to propose a comprehensive five-stage framework for software defect prediction, addressing the current challenges in the field. The first stage involves selecting a cleaned version of NASA’s defect datasets, including CM1, JM1, MC2, MW1, PC1, PC3, and PC4, ensuring the data’s integrity. In the second stage, a feature selection technique based on the genetic algorithm is applied to identify the optimal subset of features. In the third stage, three heterogeneous binary classifiers, namely random forest, support vector machine, and naïve Bayes, are implemented as base classifiers. Through iterative tuning, the classifiers are optimized to achieve the highest level of accuracy individually. In the fourth stage, an ensemble machine-learning technique known as voting is applied as a master classifier, leveraging the collective decision-making power of the base classifiers. The final stage evaluates the performance of the proposed framework using five widely recognized performance evaluation measures: precision, recall, accuracy, F-measure, and area under the curve. Experimental results demonstrate that the proposed framework outperforms state-of-the-art ensemble and base classifiers employed in software defect prediction and achieves a maximum accuracy of 95.1%, showing its effectiveness in accurately identifying software defects. The framework also evaluates its efficiency by calculating execution times. Notably, it exhibits enhanced efficiency, significantly reducing the execution times during the training and testing phases by an average of 51.52% and 52.31%, respectively. This reduction contributes to a more computationally economical solution for accurate software defect prediction.

A type system for formal verification of cyber-physical systems C/C++ software

The subject: This study focuses on improving the quality of Cyber-Physical System (CPS) software by eliminating incorrect usage of units of measurement and orientation in C/C++ programs. Incorrect usage often leads to critical errors that conventional systems cannot effectively prevent. Manual examination of code using dimensional and orientation analysis can detect these errors in physical equations, but these methods become impractical when dealing with complex physical computations. Objectives: As suggested by Siano, the proposed approach uses physical quantities and prefixes defined by the International System of Units and orientation operations on physical objects. The elaborated system incorporates dimensional and orientation analysis and metaprogramming techniques. The methods used are dimensional & orientational analysis and metaprogramming. The following results were obtained: ensuring consistency of the units, incorporating orientation operations into the programming model for accurately handling physical object rotations and alignments, and using Siano’s work to precisely manipulate object orientation, thereby reducing the likelihood of orientation-related errors. Checking physical dimensions and orientations during the compilation stage identifies potential software defects before code execution, thereby reducing debugging time and lowering the cost of addressing issues later in development. The elaborated system represents a crucial step towards safer and more dependable Cyber-Physical System applications. This approach allows us to identify approximately 90% of incorrect usage of program variables; additionally, it detects over 50% of erroneous operations during compilation and execution of large-scale programs in real-world conditions. Conclusions. Scientific novelty: it proposed and developed a specialized C++-type library for formal compile-time software verification of Cyber-Physical Systems software. The proposed C++-type library leverages dimensional and orientational analysis to enhance software quality, reliability, and real-time formal verification. Although the proposed method for formal verification is not tailor-made for cyber-physical objects and systems, given its primary focus on software-level concerns, it does exhibit adaptability for verifying general-purpose software that incorporates various physical parameters. This versatility extends to diverse domains such as educational, gaming, and simulation software.

Structural topic modeling-based systematic review for service robots’ fault diagnosis: content, popularity and regional differences

Abstract The market for service robots is expanding as labor costs continue to rise. Faced with intricate working environments, fault detection and diagnosis are crucial to ensure the proper functioning of service robots. The objective of this review is to systematically investigate the realm of service robots’ fault diagnosis through the application of Structural Topic Modeling. A total of 289 papers were included, culminating in ten topics, including advanced algorithm application, data learning-based evaluation, automated equipment maintenance, actuator diagnosis for manipulator, non-parametric method, distributed diagnosis in multi-agent systems, signal-based anomaly analysis, integrating complex control framework, event knowledge assistance, mobile robot particle filtering method. These topics spanned service robot hardware and software failures, diverse service robot systems, and a range of advanced algorithms for fault detection in service robots. Asia-Pacific, Europe, and the Americas, recognized as three pivotal regions propelling the advancement of service robots, were employed as covariates in this review to investigate regional disparities. The review found that current research tends to favor the use of artificial intelligence (AI) algorithms to address service robots’ complex system faults and vast volumes of data. The topics of algorithms, data learning, automated maintenance, and signal analysis are advancing with the support of AI, gaining increasing popularity as a burgeoning trend. Additionally, variations in research focus across different regions were found. The Asia-Pacific region tends to prioritize algorithm-related studies, while Europe and the Americas show a greater emphasis on robot safety issues. The integration of diverse technologies holds the potential to bring forth new opportunities for future service robot fault diagnosis.Simultaneously, regional standards about data, communication, and other aspects can streamline the development of methods for service robots’ fault diagnosis.

Software bug localization based on optimized and ensembled deep learning models

AbstractAn automated task for finding the essential buggy files among software projects with the help of a given bug report is termed bug localization. The conventional approaches suffer from the challenges of performing lexical matching. Particularly, the terms utilized for describing the bugs in the bug reports are observed to be irrelevant to the terms used in the source code files. To resolve these problems, we propose an optimized and ensemble deep learning model for software bug localization. These features are reduced by the principle component analysis (PCA). Then, they are selected by the weighted convolutional neural network (CNN) model with the support of the Modified Scatter Probability‐based Coyote Optimization Algorithm (MSP‐COA). Finally, the optimal features are subjected to the ensemble deep neural network and long short‐term memory (DNN‐LSTM), with parameter tuning by the MSP‐COA. Experimental results show that the proposed approach can achieve higher bug localization accuracy than individual models.

An updated software reliability model using the shanker model and failure data

AbstractSoftware developers' goal is to develop reliable and superior software. Due to the fact that software errors frequently generate large societal or financial losses, software reliability is essential. Software reliability growth models are a widely used technique for software reliability assessment. This study examines various nonhomogeneous Poisson process models with the newly developed software reliability distribution and evaluates the unknown model parameters based on frequentist and Bayesian methods of estimation. Finally, we conduct evaluations on real datasets using a variety of evaluation criteria to compare the results of previous software reliability growth models and show how the proposed model may be applied under both approaches in a practical setting. According to this study, the innovative model's mean square error, R2, , bias, predicted relative variation, Theil statistic, and mean error of prediction values show the lowest values under the Bayesian approach for data sets II to IV, and both approaches perform well for data set I. These implementation findings demonstrate the effectiveness of our specific approach based on our examination of failure data.

Modeling the Process of Evaluation of Information Security Risks and its Methodology by the Application of Computer Systems

The main aim of the article is to study modeling the process of evaluation of information security risks by the application of computer systems. Thus, information security risk assessment and analysis of existing methods of information security risk assessment had been also studied in the article. The term "risk of information security" (IS) applies to the damage that can attack the information technology systems. IS risk is a wide range of potential threats, especially includes data violations, control measures, financial costs, reputation damage, etc. cover issues such as. IS risks include the failure of hardware and software, human errors, spams, viruses and harmful attacks, as well as natural disasters such as fires, cyclones or floods. Security risk assessment determines, assesses and implements key security controls in applications. It also focuses on prevention of software security defects and vulnerabilities. Information security risk is the potential probability of using vulnerabilities of an asset or group of assets as a specific threat to damage the organization.

Intelligent Detection and Analysis of Software Vulnerabilities based on Encryption Algorithms and Feature Extraction

Implement status detection of ship software, identify the source of faults in problematic software, and release new software versions. Based on the above requirements, the author regards the detection and control of ship software status as the core research content. Based on the actual operating environment of ship software, the functional requirements of software status detection were studied and analyzed, and a set of ship software status detection was designed and implemented, a software inspection and maintenance platform that integrates ship software operation and maintenance, as well as ship software version release and update. The author conducted practical verification of the SM3 and SM2 hybrid encryption algorithm and selected software on the ship for detection. After analyzing the experimental results, it has been proven that using a hybrid algorithm for encryption and decryption, the server can accurately obtain software information on the ship's platform, detect the software status on the ship, and locate specific problem files. For software that does not meet the standard status, the server can accurately transmit software information to the ``component integration framework'' and put the component in a ``prohibited'' scheduling state. After the server repairs the problematic software, the detection results of the software change and display as legal, while the software is in the ``allowed'' scheduling state in the ``component integration framework''.

SCFS-securing flying ad hoc network using cluster-based trusted fuzzy scheme

AbstractFlying Ad hoc Networks have emerged as a promising technology for number of real-time applications. However, the flexible and unstructured characteristics of these networks make them vulnerable to security threats posed by malicious nodes, such as denial of service attacks, node impersonation, and information breaches. Another major issue is the consideration of those nodes being unable to prove their trustworthiness due to factors like hardware or software failure, or by link interruptions, during the processing of detection of false nodes in the network. The existing mechanisms encompassing encryption, authentication, and intrusion detection highlight limitations to secure real-time applications and services due to the high speed of flying nodes and the absence of fixed network structures. To overcome these constraints, this research paper incorporates a novel framework for evaluating and improving the security of network by introducing an innovative cluster-based approach. Moreover, it presents a fuzzy model that dynamically estimates the trust levels of both individual nodes and clusters, by assigning weight to the parameters to address vulnerabilities. Additionally, a trust reconfiguration mechanism is further proposed to address the issue of nodes unable to substantiate their trust by providing them with additional chances based on the collective trust from previous evaluations. Further, the paper incorporates a dynamic reputation system to proficiently identify and separate malicious and selfish nodes from the network. Simulation results indicate a significant improvement in performance metrics, with a considerable reduction in delay and drop ratio by 41.46% and 36.37%, respectively, while the sufficient rise of 54.71% and 46.05% in throughput and coverage, respectively, comparing with the considered state-of-art.

Programming memristor arrays with arbitrarily high precision for analog computing.

In-memory computing represents an effective method for modeling complex physical systems that are typically challenging for conventional computing architectures but has been hindered by issues such as reading noise and writing variability that restrict scalability, accuracy, and precision in high-performance computations. We propose and demonstrate a circuit architecture and programming protocol that converts the analog computing result to digital at the last step and enables low-precision analog devices to perform high-precision computing. We use a weighted sum of multiple devices to represent one number, in which subsequently programmed devices are used to compensate for preceding programming errors. With a memristor system-on-chip, we experimentally demonstrate high-precision solutions for multiple scientific computing tasks while maintaining a substantial power efficiency advantage over conventional digital approaches.

Neighborhood contrastive learning-based graph neural network for bug triaging

Researchers have been developing automatic bug triaging techniques by leveraging bug information sourced from bug tracking systems. Recent studies have modeled the bug-developer relationship as a graph, introducing graph neural networks for bug triaging. Despite achieving positive outcomes, these methods overlook issues related to data sparsity and fail to fully exploit implicit relationships within the graph. In addressing these challenges, we present the Neighborhood Contrastive Learning-based Graph Neural Network Bug Triaging framework, abbreviated as NCGBT. Our approach models the relationship between bugs and developers as a bipartite graph. We utilize a pre-trained language model to acquire the initial representation of bug nodes. Employing a basic graph neural network framework, we learn the representation of all nodes and leverage these representations to predict developers for a given bug. Our proposed strategy involves neighborhood contrastive learning applied to the basic graph neural network approach. We take into account the neighbors of nodes from both structural and semantic perspectives, utilizing them as contrastive objects. Extensive experiments conducted on three public datasets demonstrate the effectiveness of the NCGBT framework.

Evaluation of the effect of sagging correction calibration errors in radiotherapy software on image matching

To investigate the impact of sagging correction calibration errors in radiotherapy software on image matching. Three software applications were used, with and without a polymethyl methacrylate rod supporting the ball bearings (BB). The calibration error for sagging correction across nine flex maps (FMs) was determined by shifting the BB positions along the Left–Right (LR), Gun–Target (GT), and Up–Down (UD) directions from the reference point. Lucy and pelvic phantom cone-beam computed tomography (CBCT) images underwent auto-matching after modifying each FM. Image deformation was assessed in orthogonal CBCT planes, and the correlations among BB shift magnitude, deformation vector value, and differences in auto-matching were analyzed. The average difference in analysis results among the three softwares for the Winston–Lutz test was within 0.1 mm. The determination coefficients (R2) between the BB shift amount and Lucy phantom matching error in each FM were 0.99, 0.99, and 1.00 in the LR-, GT-, and UD-directions, respectively. The pelvis phantom demonstrated no cross-correlation in the GT direction during auto-matching error evaluation using each FM. The correlation coefficient (r) between the BB shift and the deformation vector value was 0.95 on average for all image planes. Slight differences were observed among software in the evaluation of the Winston–Lutz test. The sagging correction calibration error in the radiotherapy imaging system was caused by an auto-matching error of the phantom and deformation of CBCT images.

Alleviating Class Imbalance Issue in Software Fault Prediction Using DBSCAN-Based Induced Graph Under-Sampling Method

Alleviating Class Imbalance Issue in Software Fault Prediction Using DBSCAN-Based Induced Graph Under-Sampling Method

Hybrid deep architecture for software defect prediction with improved feature set

Hybrid deep architecture for software defect prediction with improved feature set

Interpretable Software Defect Prediction from Project Effort and Static Code Metrics

Software defect prediction models enable test managers to predict defect-prone modules and assist with delivering quality products. A test manager would be willing to identify the attributes that can influence defect prediction and should be able to trust the model outcomes. The objective of this research is to create software defect prediction models with a focus on interpretability. Additionally, it aims to investigate the impact of size, complexity, and other source code metrics on the prediction of software defects. This research also assesses the reliability of cross-project defect prediction. Well-known machine learning techniques, such as support vector machines, k-nearest neighbors, random forest classifiers, and artificial neural networks, were applied to publicly available PROMISE datasets. The interpretability of this approach was demonstrated by SHapley Additive exPlanations (SHAP) and local interpretable model-agnostic explanations (LIME) techniques. The developed interpretable software defect prediction models showed reliability on independent and cross-project data. Finally, the results demonstrate that static code metrics can contribute to the defect prediction models, and the inclusion of explainability assists in establishing trust in the developed models.

Profiling with trust: system monitoring from trusted execution environments

AbstractLarge-scale attacks on IoT and edge computing devices pose a significant threat. As a prominent example, Mirai is an IoT botnet with 600,000 infected devices around the globe, capable of conducting effective and targeted DDoS attacks on (critical) infrastructure. Driven by the substantial impacts of attacks, manufacturers and system integrators propose Trusted Execution Environments (TEEs) that have gained significant importance recently. TEEs offer an execution environment to run small portions of code isolated from the rest of the system, even if the operating system is compromised. In this publication, we examine TEEs in the context of system monitoring and introduce the Trusted Monitor (TM), a novel anomaly detection system that runs within a TEE. The TM continuously profiles the system using hardware performance counters and utilizes an application-specific machine-learning model for anomaly detection. In our evaluation, we demonstrate that the TM accurately classifies 86% of 183 tested workloads, with an overhead of less than 2%. Notably, we show that a real-world kernel-level rootkit has observable effects on performance counters, allowing the TM to detect it. Major parts of the TM are implemented in the Rust programming language, eliminating common security-critical programming errors.

Identification of Risk Factors in the Software Design Stage Using the C4.5 Algorithm

A strong design phase is necessary for good software. However, design errors in software can cause serious issues with its creation and use. Therefore, the goal of this study is to find risk variables that could have an early impact on software development. In this study, a machine learning technique called technique C4.5 is employed to create decision tree models. 100 respondents with software design experience participated in the online surveys and questionnaires that collected the data for this study in 2022. The C4.5 Algorithm was used in this study to analyze the data and determine the risk variables that affect the success of software design. The study's findings show that the C4.5 Algorithm-based model has a high level of accuracy (93.33%), which means that the data can offer crucial insights into understanding potential risks that may arise during the software design stage, enabling software developers to take the necessary precautions to lessen or eliminate these risks. In order to enhance the caliber and effectiveness of software design, this research is anticipated to provide a significant contribution to practitioners and academics in the field of software development.

Assessing the Impact of Compiler Optimizations on GPUs Reliability

Graphics Processing Units (GPUs) compilers have evolved in order to support general-purpose programming languages for multiple architectures. NVIDIA CUDA Compiler (NVCC) has many compilation levels before generating the machine code and applies complex optimizations to improve performance. These optimizations modify how the software is mapped in the underlying hardware; thus, as we show in this article, they can also affect GPU reliability. We evaluate the effects on the GPU error rate of the optimization flags applied at the NVCC Parallel Thread Execution (PTX) compiling phase by analyzing two NVIDIA GPU architectures (Kepler and Volta) and two compiler versions (NVCC 10.2 and 11.3). We compare and combine fault propagation analysis based on software fault injection, hardware utilization distribution obtained with application-level profiling, and machine instructions radiation-induced error rate measured with beam experiments. We consider eight different workloads and 144 combinations of compilation flags, and we show that optimizations can impact the GPUs’ error rate of up to an order of magnitude. Additionally, through accelerated neutron beam experiments on a NVIDIA Kepler GPU, we show that the error rate of the unoptimized GEMM (-O0 flag) is lower than the optimized GEMM’s (-O3 flag) error rate. When the performance is evaluated together with the error rate, we show that the most optimized versions (-O1 and -O3) always produce a higher amount of correct data than the unoptimized code (-O0).

Automatic program bug fixing by focusing on finding the shortest sequence of changes

Automatic bug repair as the last step in program repair has attracted a lot of research attention. Various ideas and techniques have been presented in this field. Recent bug fixing techniques use machine learning and graphs to generate fixes. Despite the promising results of recent approaches, maintaining high speed and accuracy as well as recording a wide range of errors may still be a problem. In this paper, a new approach is presented in the field of automatic bug fixing based on graphs and model checking. For this purpose, we have used the graph transformation and model checking system to create a sequence of edits and produce fixes. Then, using meta-heuristic algorithms, we have selected the best solution and fix from the generated solutions. We use the extracted graphs from the buggy JavaScript code and their corresponding bug-free ones. In evaluating the effectiveness of the proposed method, we implement it in GROOVE, which is a toolbox used to design and check graph transformation systems. Experimental results on identical dataset demonstrate that the proposed method outperforms other related methods in generating fixes. Also, this method covers a wider range of bugs compared to previous methods.

Feature importance feedback with Deep Q process in ensemble-based metaheuristic feature selection algorithms

Feature selection is an indispensable aspect of modern machine learning, especially for high-dimensional datasets where overfitting and computational inefficiencies are common concerns. Traditional methods often employ either filter, wrapper, or embedded approaches, which have limitations in terms of robustness, computational load, or capability to capture complex interactions among features. Despite the utility of metaheuristic algorithms like Particle Swarm Optimization (PSO), Firefly Algorithm (FA), and Whale Optimization (WOA) in feature selection, there still exists a gap in efficiently incorporating feature importance feedback into these processes. This paper presents a novel approach that integrates the strengths of PSO, FA, and WOA algorithms into an ensemble model and further enhances its performance by incorporating a Deep Q-Learning framework for relevance feedbacks. The Deep Q-Learning module intelligently updates feature importance based on model performance, thereby fine-tuning the selection process iteratively. Our ensemble model demonstrates substantial gains in effectiveness over traditional and individual metaheuristic approaches. Specifically, the proposed model achieved a 9.5% higher precision, an 8.5% higher accuracy, an 8.3% higher recall, a 4.9% higher AUC, and a 5.9% higher specificity across multiple software bug prediction datasets and samples. By resolving some of the key issues in existing feature selection methods and achieving superior performance metrics, this work paves the way for more robust and efficient machine learning models in various applications, from healthcare to natural language processing scenarios. This research provides an innovative framework for feature selection that promises not only superior performance but also offers a flexible architecture that can be adapted for a variety of machine learning challenges.

Eğitim Yönetiminde Etkili Metodik Değerlendirme Süreçlerinin İncelenmesi

Methodically making a negative assessment of the effectiveness and efficiency of a policy or program immediately, rather than over a broader period of time, may lead to the termination of the program. When evaluating an institution in educational management, the content of the evaluation, it and the values involved in the event will affect the final evaluation as a whole. In this context, one of the decisions that should be taken before making any policy evaluation is to determine who should make the evaluation. The specific circumstances surrounding the necessity for any evaluation will have a significant impact on the alternative that will be selected. Within this framework, the suitability of time, financing situation, subject matter expertise, understanding of evaluation findings and management support, etc. the concepts of internal evaluation and external evaluation stand out in terms of the functioning of processes. Internal assessments have other advantages, such as low cost. In addition, internal evaluators are more predisposed to the organization itself, the evaluated program and related policies, corporate stakeholders and the target audience. The biggest disadvantage of internal evaluation is its real and perceived biases. There is real prejudice. Because it is human nature for corporate members to perceive program errors through the lens of their common corporate experience. External evaluations also have a number of advantages in terms of minimizing biases, at least internal biases. The most important advantage of external evaluation is that evaluators are perceived as impartial because they have no interest in the evaluation results. The plus advantage of using external evaluators is that they are not part of the corporate culture, program and evaluated policy, and their professionalism, previous experience, training, values and perceptions offer them a completely different perspective on problem identification and potential improvements. An important disadvantage of using external evaluators is that it is more expensive than internal evaluation and more time is needed. External evaluations take more time. Because evaluators need to internalize the programs or institutions as well as the policies, programs and institutions they evaluate. Key Words: Educational Management, Educational Effectiveness, Methodical Evaluation