Regression testing is an essential part of software development, but it can be costly and require significant computational resources. Regression Test Selection (RTS) improves regression testing efficiency by only re-executing the tests that have been affected by code changes. Recently, dynamic and static RTS techniques for Java projects showed that selecting tests at a coarser granularity, class-level, is more effective than selecting tests at a finer granularity, method- or statement-level. However, prior techniques are mainly considering Java object-oriented projects but not modularized Java projects. Given the explicit support of architectural constructs introduced by the Java Platform Module System (JPMS) in the ninth edition of Java, these research efforts are not customized for component-based Java projects. To that end, we propose two static component-based RTS approaches called CORTS and its variant C2RTS tailored for component-based Java software systems. CORTS leverages the architectural information such as components and ports, specified in the module descriptor files, to construct module-level dependency graph and identify relevant tests. The variant, C2RTS, is a hybrid approach in which it integrates analysis at both the module and class levels, employing module descriptor files and compile-time information to construct the dependency graph and identify relevant tests.We evaluated CORTS and C2RTS on 1200 revisions of 12 real-world open source software systems, and compared the results with those of class-level dynamic (Ekstazi) and static (STARTS) RTS approaches. The results showed that CORTS and C2RTS outperformed the static class-level RTS in terms of safety violation that measures to what extent an RTS technique misses test cases that should be selected. Using Ekstazi as the baseline, the average safety violation with respect to Ekstazi was 1.14% for CORTS, 2.21% for C2RTS, and 3.19% for STARTS. On the other hand, the results showed that CORTS and C2RTS selected more test cases than Ekstazi and STARTS. The average reduction in test suite size was 22.78% for CORTS and 43.47% for C2RTS comparing to the 68.48% for STARTS and 84.21% for Ekstazi. For all the studied subjects, CORTS and C2RTS reduced the size of the static dependency graphs compared to those generated by static class-level RTS, leading to faster graph construction and analysis for test case selection. Additionally, CORTS and C2RTS achieved reductions in overall end-to-end regression testing time compared to the retest-all strategy.

Artificial Intelligence (AI) is increasingly applied across various domains, including education, where it enhances numerous aspects of the learning process, from course design to assessment. Despite its benefits in efficiency, scalability, and consistency, AI in education is applied in different learning and educational stages. This paper focuses on the use of AI in the assessment stage. To that end, this paper proposes a taxonomy of AI-based learner assessment educational technologies (EduTech) from both research and industrial perspectives. The taxonomy provides a comprehensive understanding and identifies gaps in the field. Using the PRISMA framework, we systematically review related research papers and tools.

Software architecture is critical for defining a software system’s structure and behavior, influencing quality attributes such as performance, scalability, and security. Architectural drift, the gap between planned and implemented architectures, presents significant challenges, particularly in Java, which lacks explicit architectural constructs. The Java Platform Module System (JPMS) introduced in Java 9 addresses these issues by supporting modularity and dependency management. However, transitioning existing Java applications to JPMS is complex and error-prone, often leading to modules with excessive privileges. Automated component recovery tools, such as the Algorithm for Comprehension-Driven Clustering (ACDC), are essential for facilitating this transition.This paper evaluates ACDC in the context of JPMS. We examined seven real-world Java software systems to determine if ACDC can recover component-based architectures that conform to JPMS’s modular constraints and the Java packaging structures of both the core application and its test suites. Our study found that over 50% of modules had cyclic dependencies, more than 50% of program packages were split across multiple modules with some subjects exceeding 90%, and over 50% of test packages were split across modules in most of the seven subjects. These violations necessitate extensive refactoring. Additionally, ACDC often produced overly fragmented module structures and imbalanced class distributions, complicating maintenance and reducing modular efficiency.

In today’s rapidly evolving digital landscape, the surge in smartphone usage is paralleled by an increasing wave of cyberthreats, highlighting the limitations of existing signature-based malware detection methods. To address this problem, our research introduces a Time-Aware Machine Learning (TAML) framework specifically designed for Android malware detection. Our framework extracts the best time-correlated features and then it builds time-aware and time-agnostic machine learning (ML) models. The ML models are trained on the KronoDroid dataset, which contains more than 41,000 benign Android apps and more than 36,000 malicious apps developed between 2008 to 2020. Our experimental evaluation revealed that the Last Modification Date ‘LastModDate’ feature is a critical variable for time-aware classification. Moreover, our empirical analysis reveals that real-device detection outperforms emulator-based detection. Impressively, the time-correlated features boosts the detection performance and achieving an outstanding 99.98

The global spread of Coronavirus (COVID-19) has prompted imperative research into scalable and effective detection methods to curb its outbreak. The early diagnosis of COVID-19 patients has emerged as a pivotal strategy in mitigating the spread of the disease. Automated COVID-19 detection using Chest X-ray (CXR) imaging has significant potential for facilitating large-scale screening and epidemic control efforts. This paper introduces a novel approach that employs state-of-the-art Convolutional Neural Network models (CNNs) for accurate COVID-19 detection. The employed datasets each comprised 15,000 X-ray images. We addressed both binary (Normal vs. Abnormal) and multi-class (Normal, COVID-19, Pneumonia) classification tasks. Comprehensive evaluations were performed by utilizing six distinct CNN-based models (Xception, Inception-V3, ResNet50, VGG19, DenseNet201, and InceptionResNet-V2) for both tasks. As a result, the Xception model demonstrated exceptional performance, achieving 98.13% accuracy, 98.14% precision, 97.65% recall, and a 97.89% F1-score in binary classification, while in multi-classification it yielded 87.73% accuracy, 90.20% precision, 87.73% recall, and an 87.49% F1-score. Moreover, the other utilized models, such as ResNet50, demonstrated competitive performance compared with many recent works.

A year-round performance of a new design of a solar water heater with natural circulation was studied experimentally at Amman, Jordan (latitude of 32 degrees N and the altitude of 985 m above sea level). The measured parameters were temperatures of ambient, water inlet, and outlet of the collector; the temperature of the heated water stored in the storage tank, and the solar incident radiation on the tilted angle. The obtained results showed that the performance of the solar water heater depended highly on both of the flow rate through the collector and the solar tilted incident radiation. A typical day analysis of the system showed a high solar collector efficiency especially around mid-day (a time when it receives the highest solar energy). Also, data revealed that the maximum total useful heat gain without water consumption was 200.55 MJ recorded during September, while the maximum useful heat gain with water consumption was 304.81MJ recorded during October. Meanwhile, the maximum daily efficiency without water consumption was 54% recorded during May, while the maximum daily efficiency with consumption was 55% recorded during September. Based on the results of this study, the new design of a solar water heater can be recommended for the optimal use of solar energy to heat water.

The increasing interactive content in the Internet motivated researchers and data scientists to conduct Aspect-Based Sentiment Analysis (ABSA) research to understand the various sentiments and the different aspects of a product in a single user’s comment. Determining the various aspects along with their polarities (positive, negative, or neutral) from a single comment is a challenging problem. To this end, we have designed and developed a deep learning model based on Gated Recurrent Units (GRU) and features extracted using the Multilingual Universal Sentence Encoder (MUSE). The proposed Pooled-GRU model trained on a Hotels’ Arabic reviews to address two ABSA tasks: (1) aspect extraction, and (2) aspect polarity classification. The proposed model achieved high results with 93.0% F1 score in the former task and 90.86% F1 score in the latter task. Our experimental results show that our proposed model outperforms the baseline model and the related research methods evaluated on the same dataset. More precisely, our proposed model showed 62.1% improvement in the F1 score over the baseline model for the aspect extraction task and 15% improvement in the accuracy over the baseline model for the aspect polarity classification task.

Cloud computing technology significantly emerged in various systems to enhance the platform and infrastructure of services. Fog and Edge computing technologies transform computing power and processing close to the end-user and resolve many data processing challenges. Healthcare IoT systems refer to using medical sensors and mobile computing to manage patients’ conditions and use cloud-based technologies to transfer collected patients’ data to the cloud computing storage. Many previous research efforts focused on the use of edge, fog, and cloud computing technologies in the field of healthcare IoT systems. However, no previous research study summarized the main research topics and trends in these research efforts. To that end, we conducted a systematic and comprehensive empirical research study to understand the main research topics and the employed computing technologies and trends in these research topics. This study investigated 100 research studies focusing on cloud-based technologies in the filed of healthcare IoT systems published in highly reputable venues over the past decade. We created a taxonomy of the main research topics of the investigated research manuscripts and reported the main focus of these research efforts and the trends in this field. Our investigation concluded that the researchers mainly focused on the following five research topics: big data, security, network latency, energy efficiency, and QoS. However, some research topics are well studied while others are poorly investigated and need more attention. We found that the highest number of research papers focused on big data, i.e., 63% of the studied papers, while the energy efficiency topic had the lowest number of papers, i.e., 17% of the papers. Researchers and practitioners interested in the cloud-based healthcare IoT systems can use our findings to gear their work on the areas that are poorly investigated to find better solutions.

Text classification has many applications in various fields; such as news categorization, sentiment analysis, E-mail spam filtering, and others. However, handling textual data is a challenging task owing to the potentially massive number of features (words). The presence of redundant irrelevant features deteriorates the performance of a learning algorithm and makes the process of text classification more complex. This research conducts a comparison study of several filtering-based feature se-lection methods in the context of Arabic text classification. Arabic is a highly complex language syntactically and morphologically which leads to more complicated learning tasks. Proposing a ro-bust classification model is demanding. Remarkably, integrating filtering approaches results in significant improvements in the performance of classification algorithms.

An experimental study of the performance of a refrigerator using two refrigerants R-134a and R-410A has been conducted. The R-134a was the original design refrigerant, while the R-410A was the drop-in refrigerant. The study was performed on a small refrigerator charged with each refrigerant alone at nearly the same ambient conditions. Temperatures at various locations in the refrigeration system were measured using thermocouples during the running of the experiments, and the data collected were processed into performance refrigeration parameters. The results have indicated that both the refrigeration capacity and coefficient of performance were remarkably higher for the refrigerant R-410a by about 23 and 24%, respectively.

The ability to automatically understand and analyze human language attracted researchers and practitioners in the Natural Language Processing (NLP) field. Detecting humor is an NLP task needed in many areas, including marketing, politics, and news. However, such a task is challenging due to the context, emotion, culture, and rhythm. To address this problem, we have proposed a robust model called BFHumor, a BERT-Flair-based Humor detection model that detects humor through news headlines. It is an ensemble model of different state-of-the-art pre-trained models utilizing various NLP techniques. We used public humor datasets from the SemEval-2020 workshop to evaluate the proposed model. As a result, the model achieved outstanding performance with 0.51966 as Root Mean Squared Error (RMSE) and 0.62291 as accuracy. In addition, we extensively investigated the underlying reasons behind the high accuracy of the BFHumor model in humor detection tasks. To that end, we conducted two experiments on the BERT model: vocabulary level and linguistic capturing level. Our investigation shows that BERT can capture surface knowledge in the lower layers, syntactic in the middle, and semantic in the higher layers.

Deoxyribonucleic acid (DNA) sequencing is the process of locating the sequence of the main chemical bases in the DNA. Next-generation sequencing (NGS) is the state-of-the-art DNA sequencing technique. The NGS technique advanced the biological science in analyzing human DNA due to its scalability, high throughput, and speed. Analyzing human DNA is crucial to determine the ability of a person to develop certain diseases and his ability to respond to certain medications. ChIP-sequencing is a method that combines chromatin immunoprecipitation (ChIP) with NGS sequencing to analyze protein interactions with DNA to identify binding sites. Many online web tools have been developed to conduct ChIP-Seq data analysis to either discover or find motifs, i.e., patterns of binding sites. Since these ChIP-Seq web tools need to be used by clinical practitioners, they must comply to the web-related usability tasks including effectiveness, efficiency and satisfaction to enhance the user experience (UX). To that end, we have conducted an empirical study to understand their UX design. Specifically, we have evaluated the usability of 8 widely used ChIP-Seq web tools against 6 known usability quality metrics. Our study shows that the design of the studied ChIP-Seq web tools does not follow the UX design principles.

Many popular object-oriented (OO) programming languages, such as Java, do not provide explicit support for architecture-based development, i.e., do not provide programming-language constructs that are at the granularity of architectural constructs, such as components and ports. The gap between how engineers design their systems and how they implement them has been one of the leading causes of architectural drift—a situation in which the prescriptive architecture (the designed architecture) does not match the descriptive architecture (the implemented architecture). To mitigate this challenge, in its ninth iteration, Java introduced the concept of Java Platform Module System (JPMS), which for the first time provides explicit implementation-level support for well-known architectural constructs, such as components (called modules) and ports (called module directives). Despite this, the majority of existing Java applications (apps) are still purely OO programs that do not make use of the new constructs, because converting them to well-structured component-based (CB) programs is a tedious and error-prone task. In fact, prior research has shown that when engineers convert OO apps to CB apps, they tend to be highly over-privileged, i.e., components are granted more access privileges than they actually need. To mitigate these challenges, we have developed OO2CB, an approach for conversion of an OO Java app to a least-privilege CB Java app. OO2CB employs component recovery techniques to assist the developer in determining a given OO app's components. It then statically analyzes the source code of the app to determine the dependencies among its recovered components and the required port types for facilitating their interaction. Finally, OO2CB generates a functionally equivalent CB app that satisfies the least-privilege security principle. Our experiments on several large real-world OO Java apps corroborate the effectiveness of OO2CB.

Aim: To evaluate the effects of the COVID-19 pandemic and lockdown on orthodontic patients' apprehension and inflow and to investigate the treatment-related problems encountered during the pandemic. Materials and Methods: A cross-sectional study was carried out in Saudi Arabia, and patients with active orthodontic treatment were reached via an online questionnaire. A total of 260 orthodontic patients agreed to participate in the study. The sample size calculation was performed using the Raosoft sample size calculator based on the estimation of 75% of the population need orthodontic treatment. The questionnaire included three sections: demographic data, patient fear, and troubles encountered during the lockdown. A link was sent to the participants via different social media platforms and applications. The chi-square goodness-of-fit test was used to assess differences between the variables. Pearson correlation, binomial logistic regression, and multiple logistic regression tests were used to assess the extent of the relationship between patient apprehension and safety measures of COVID-19, as well as between patient orthodontic-related problems. Results: About half of the participants (52.3%) were not afraid of COVID-19 or panicked; however, 54% of them felt depressed during the lockdown. More than 80% were not afraid of visiting the orthodontists or thinking of changing their orthodontists for safety measures and did not want to postpone their treatment. The participants felt that excellent disinfection was the most important measure of infection control followed by wearing masks, face shields, and protective clothing. Conclusion: COVID-19 and the lockdown have a noticeable impact on the patients' apprehension and dejection. A very high percentage of patients in this study showed no hesitation to visit their orthodontists, and they did not want to stop their treatment during the pandemic. The most common treatment-related problem reported was cheek injury, followed by bracket breakage.

Currently, the whole world is fighting a very dangerous and infectious disease caused by the novel coronavirus, called COVID-19. The COVID-19 is rapidly spreading around the world due to its high infection rate. Therefore, early discovery of COVID-19 is crucial to better treat the infected person as well as to slow down the spread of this virus. However, the current solution for detecting COVID-19 cases including the PCR test, CT images, epidemiologically history, and clinical symptoms suffer from high false positive. To overcome this problem, we have developed a novel transfer deep learning approach for detecting COVID-19 based on x-ray images. Our approach helps medical staff in determining if a patient is normal, has COVID-19, or other pneumonia. Our approach relies on pre-trained models including Inception-V3, Xception, and MobileNet to perform two tasks: i) binary classification to determine if a person infected with COVID-19 or not and ii) a multi-task classification problem to distinguish normal, COVID-19, and pneumonia cases. Our experimental results on a large dataset show that the F1-score is 100% in the first task and 97.66 in the second task.

The impact of graphene reinforcement on the mechanical properties of metals has been a subject of intense investigation over the last decade in surface applications to mitigate the impact of tribological loadings or for strengthening purposes when dispersed into a bulk material. Here, the effect on the plastic indentation response of a single graphene layer grown on copper is analyzed for two configurations: one with graphene at the surface, the other with graphene sandwiched under a 100 nm thick copper cap layer. Nanoindentation under both displacement and load control conditions show both earlier and shorter pop-in excursions compared to systems without graphene. Atomic force microscopy reveals much smoother pile-ups with no slip traces in the presence of a surface graphene layer. The configuration with the intercalated graphene layer appears as an ideal elementary system to address bulk hardening mechanisms by indentation testing. Transmission electron microscopy (TEM) cross-sections below indents show more diffuse and homogeneous dislocation activity in the presence of graphene. 3D dislocation dynamics simulations allow unraveling of the origin of these 3D complex phenomena and prove that the collective dislocation mechanisms are dominantly controlled by the strong back stress caused by the graphene barrier. These results provide a quantitative understanding of the impact of graphene on dislocation mechanisms for both surface and bulk applications, but with an impact that is not as large as anticipated from other studies or general literature claims.

Diabetic retinopathy is an eye disease caused by high blood sugar and pressure which damages the blood vessels in the eye. Diabetic retinopathy is the root cause of more than 1% of the blindness worldwide. Early detection of this disease is crucial as it prevents it from progressing to a more severe level. However, the current machine learning-based approaches for detecting the severity level of diabetic retinopathy are either, i) rely on manually extracting features which makes an approach unpractical, or ii) trained on small dataset thus cannot be generalized. In this study, we propose a transfer learning-based approach for detecting the severity level of the diabetic retinopathy with high accuracy. Our model is a deep learning model based on global average pooling (GAP) technique with various pre-trained convolutional neural net- work (CNN) models. The experimental results of our approach, in which our best model achieved 82.4% quadratic weighted kappa (QWK), corroborate the ability of our model to detect the severity level of diabetic retinopathy efficiently.

Optical satellites generally provide high-resolution panchromatic but low-resolution multispectral images which provide structural details of features and spectral information respectively. Nowadays, fusion of the two types of resolutions, to have complementary information, becomes increasingly essential for many applications such as microscopic, astronomical and satellite imagery. In this paper, a novel hybrid pixel-level image fusion method is proposed for benefiting from both panchromatic (PAN) and multispectral (MUL) images. The proposed method integrates Gram Schmidt (GS) and curvelet transforms (CVT), by the aid of local energy and maximum fusion rules, for reducing individual method limitations and achieving both better spectral consistency and spatial details preservation. After a pre-processing stage, orthonormal bases are obtained for low spatial resolution images by using GS transform. Then, high-resolution and low-resolution images are fused using CVT by the aid of histogram matching. Finally, the fused image is obtained by applying both curvelet and GS inverse transforms. The performance of the proposed method is evaluated using publicly available Pleiades benchmark-datasets. Consequently, the spectral and spatial qualities of the fused images are assessed subjectively as well as objectively using different quality metrics. Moreover, the proposed method is compared with state-of-the-art fusion techniques and results show the robustness of the proposed method that has the best result in spatial and spectral evaluation metrics such as, Quality with No Reference (QNR), Peak Signal to Noise Ratio (PSNR), Standard Deviation (SD), Entropy (ENT) and Spectral Correlation Coefficient (SCC) metrics.

Question-answering platforms serve millions of users seeking knowledge and solutions for their daily life problems. However, many knowledge seekers are facing the challenge to find the right answer among similar answered questions and writer’s responding to asked questions feel like they need to repeat answers many times for similar questions. This research aims at tackling the problem of learning the semantic text similarity among different asked questions by using deep learning. Three models are implemented to address the aforementioned problem: i) a supervised-machine learning model using XGBoost trained with pre-defined features, ii) an adapted Siamese-based deep learning recurrent architecture trained with pre-defined features, and iii) a Pre-trained deep bidirectional transformer based on BERT model. Proposed models were evaluated using a reference Arabic dataset from the mawdoo3.com company. Evaluation results show that the BERT-based model outperforms the other two models with an F1=92.99%, whereas the Siamese-based model comes in the second place with F1=89.048%, and finally, the XGBoost as a baseline model achieved the lowest result of F1=86.086%.

Humans have become the weakest point in the information security chain, and social engineers take advantage of that fact. Social engineers manipulate people psychologically to convince them to divulge sensitive information or to perform malicious acts. Social engineering security attacks can be severe and difficult to detect. Therefore, to prevent these attacks, employees and their organizations should be aware of relevant defense mechanisms. This research develops a taxonomy of social engineering defense mechanisms that can be used to develop educational materials for use in various kinds of organizations. To develop the taxonomy, the authors conducted a systematic literature review of related research efforts and extracted the main target points of social engineers and the defense mechanisms regarding each target point.