Search

The importance of accurately predicting student performance in education, especially in the challenging curricular unit of Introductory Programming, cannot be overstated. As institutions struggle with high failure rates and look for solutions to improve the learning experience, the need for effective prediction methods becomes critical. This study aims to conduct a systematic review of the literature on methods for predicting student performance in higher education, specifically in Introductory Programming, focusing on machine learning algorithms. Through this study, we not only present different applicable algorithms but also evaluate their performance, using identified metrics and considering the applicability in the educational context, specifically in higher education and in Introductory Programming. The results obtained through this study allowed us to identify trends in the literature, such as which machine learning algorithms were most applied in the context of predicting students’ performance in Introductory Programming in higher education, as well as which evaluation metrics and datasets are usually used.

Autonomous vehicles (AVs) rely heavily on sensors to perceive their surrounding environment and then make decisions and act on them. However, these sensors have weaknesses, and are prone to failure, resulting in decision errors by vehicle controllers that pose significant challenges to their safe operation. To mitigate sensor failures, it is necessary to understand how they occur and how they affect the vehicle’s behavior so that fault-tolerant and fault-masking strategies can be applied. This survey covers 108 publications and presents an overview of the sensors used in AVs today, categorizes the sensor’s failures that can occur, such as radar interferences, ambiguities detection, or camera image failures, and provides an overview of mitigation strategies such as sensor fusion, redundancy, and sensor calibration. It also provides insights into research areas critical to improving safety in the autonomous vehicle industry, so that new or more in-depth research may emerge.

Summary: Humans are exposed to micro-and-nano plastics (MNPs) through various routes, but the adverse health effects of MNPs on different organ systems are not yet fully understood. This review aims to provide an overview of the potential impacts of MNPs on various organ systems and identify knowledge gaps in current research. The summarized results suggest that exposure to MNPs can lead to health effects through oxidative stress, inflammation, immune dysfunction, altered biochemical and energy metabolism, impaired cell proliferation, disrupted microbial metabolic pathways, abnormal organ development, and carcinogenicity. There is limited human data on the health effects of MNPs, despite evidence from animal and cellular studies. Most of the published research has focused on specific types of MNPs to assess their toxicity, while other types of plastic particles commonly found in the environment remain unstudied. Future studies should investigate MNPs exposure by considering realistic concentrations, dose-dependent effects, individual susceptibility, and confounding factors.

Large language models (LLMs) have had a significant impact on several domains, including software engineering. However, a comprehensive understanding of LLMs’ use, impact, and potential limitations in software engineering is still emerging and remains in its early stages. This paper analyzes the role of large language models (LLMs), such as ChatGPT-3.5, in software requirements engineering, a critical area in software engineering experiencing rapid advances due to artificial intelligence (AI). By analyzing several studies, we systematically evaluate the integration of ChatGPT into software requirements engineering, focusing on its benefits, challenges, and ethical considerations. This evaluation is based on a comparative analysis that highlights ChatGPT’s efficiency in eliciting requirements, accuracy in capturing user needs, potential to improve communication among stakeholders, and impact on the responsibilities of requirements engineers. The selected studies were analyzed for their insights into the effectiveness of ChatGPT, the importance of human feedback, prompt engineering techniques, technological limitations, and future research directions in using LLMs in software requirements engineering. This comprehensive analysis aims to provide a differentiated perspective on how ChatGPT can reshape software requirements engineering practices and provides strategic recommendations for leveraging ChatGPT to effectively improve the software requirements engineering process.

The exponential growth in data generation has become a ubiquitous phenomenon in today’s rapidly growing digital technology. Technological advances and the number of connected devices are the main drivers of this expansion. However, the exponential growth of data presents challenges across different architectures, particularly in terms of inefficient energy consumption, suboptimal bandwidth utilization, and the rapid increase in data stored in cloud environments. Therefore, data reduction techniques are crucial to reduce the amount of data transferred and stored. This paper provides a comprehensive review of various data reduction techniques and introduces a taxonomy to classify these methods based on the type of data loss. The experiments conducted in this study include distinct data types, assessing the performance and applicability of these techniques across different datasets.

Oncology has emerged as a crucial field of study in the domain of medicine. Computed tomography has gained widespread adoption as a radiological modality for the identification and characterisation of pathologies, particularly in oncology, enabling precise identification of affected organs and tissues. However, achieving accurate liver segmentation in computed tomography scans remains a challenge due to the presence of artefacts and the varying densities of soft tissues and adjacent organs. This paper compares artificial intelligence algorithms and traditional medical image processing techniques to assist radiologists in liver segmentation in computed tomography scans and evaluates their accuracy and efficiency. Despite notable progress in the field, the limited availability of public datasets remains a significant barrier to broad participation in research studies and replication of methodologies. Future directions should focus on increasing the accessibility of public datasets, establishing standardised evaluation metrics, and advancing the development of three-dimensional segmentation techniques. In addition, maintaining a collaborative relationship between technological advances and medical expertise is essential to ensure that these innovations not only achieve technical accuracy, but also remain aligned with clinical needs and realities. This synergy ensures their applicability and effectiveness in real-world healthcare environments.

The advent of technologies such as 5G makes it possible to improve the availability and quality of existing services and introduce new ones, such as enhanced mobile broadband, IoT applications, augmented reality, mission-critical services, cloud gaming, or smart infrastructure. It is now possible to get faster responses to a wide range of requests, but the addition of more users and services can still make normal operation difficult due to network congestion, bandwidth limitations, scalability issues, service differentiation, or security concerns. To solve this problem, and to help meet Service Level Agreements (SLAs), some services must be brought closer to the user. The Multi-access Edge Computing (MEC) initiative is a step in this direction, enabling cloud-like services to be moved closer to the end user, providing lower access latencies. However, the dynamic nature of edge computing environments, and the mobility of users, require the implementation of automated service delivery processes that can adapt to environmental conditions, such as the use of optimal policies tailored to the scenario. This article surveys the research focused on MECs, particularly those that use automated deployment mechanisms, namely Infrastructure as Code (IaC) tools. Today, these tools play a key role in automated deployment mechanisms, especially for maintaining optimal policies, which is still under investigation. The result of this assessment has been the identification of the relevance of IaC to these processes and the identification of future research directions.

Distributed databases and data transformation mechanisms are highly relevant to Business Intelligence and Data Analytics. Most enterprise systems today have multiple and distributed databases, and the growing dissemination of Edge-Oriented architectures is driving this trend across many industries and business domains. The Entity-Relationship (ER) model is fundamental to modeling complex enterprise systems, but it has shortcomings. In particular, the ER model cannot represent data transport between different locations (or databases) of a system, nor can it conceptually express data transformation operations, such as aggregate and line functions, that are standard in data analytics. Therefore, we propose ER+, an extension of the ER model, where data distribution, data transport, data transformation, and information generation for distributed operational and analytical systems can be visually identified. The new ER+ representation has another important benefit, it provides the basis for transforming conceptual models into physical implementations of distributed data-oriented systems. This work introduces new concepts in ER modeling and illustrates their application. The practicality of the ER+ model is also demonstrated through the TPC-H benchmark.

The healthcare industry has become increasingly challenging, requiring retrieval of knowledge from large amounts of complex data to find the best treatments. Several works have suggested the use of Data Mining tools to overcome the challenges; however, none of them has suggested the best tool to do so. To fill this gap, this paper presents a survey of popular open-source data mining tools in which data mining tool selection criteria based on healthcare application requirements is proposed and the best ones using the proposed selection criteria are identified. The following popular open-source data mining tools are assessed: KNIME, R, RapidMiner, Scikit-learn, and Spark. The study shows that KNIME and RapidMiner provide the largest coverage of healthcare data mining requirements.

Abstract Filaggrin (FLG) protein is indispensable for multiple aspects of the epidermal barrier function but its accumulation in a monomeric filaggrin form may initiate premature keratinocytes death; it is unclear how filaggrin levels are controlled before the formation of storing keratohyalin granules. Here we show that keratinocyte‐secreted small extracellular vesicles (sEVs) may contain filaggrin‐related cargo providing a route of eliminating excess filaggrin from keratinocytes; blocking of sEV release has cytotoxic effects on those cells. Filaggrin‐containing sEVs are found in plasma in both healthy individuals and atopic dermatitis patients. Staphylococcus aureus (S. aureus) enhances packaging and secretion of filaggrin‐relevant products within the sEVs for enhanced export via a TLR2‐mediated mechanism which is also linked to the ubiquitination process. This filaggrin removal system, preventing premature keratinocyte death and epidermal barrier dysfunction, is exploited by S. aureus which promotes filaggrin elimination from the skin that could help safeguard bacterial growth.

Conceptual modeling describes the physical or social aspects of the world abstractly, encompassing the interpretation of data production, gathering, visualization, and analysis. The quality of the data analysis system will limit the excellence of any decision-making process. Thus, accurately specifying the physical data model is essential. The primary goal of our work is to compare tools that can create this physical model. We recognize several types of data models, but we only include the relational data model. We evaluate free and commercial data modeling tools. But it is challenging to decide how to compare them and which elements are crucial. We propose a new approach for software tools’ evaluation based on the Business Readiness Rating (BRR) model and the OSSpal evaluation methodology. In this work, we show that this new methodology can be tailored to the needs of each individual developer or team, thus providing proper and meaningful results. Also, by applying this hybrid approach to the evaluation of data modelling tools, we show it can robustly handle the bias from lesser relevant evaluation categories.

The increase in the amounts of information used to analyze data is problematic since the memory necessary to store and process it is getting quite big. The interval inverted index representation was developed to reduce the required memory to store data, and Frag-Cubing is one of the most popular algorithms. In this paper, we propose two new data cubing algorithms: 3iCubing and M3iCubing. 3iCubing is a Frag-Cubing-based algorithm that uses the interval inverted index representation, while M3iCubing uses both a normal and interval inverted index data representation. The algorithms were compared using synthetic and real data sets in indexation and querying operations, both runtime and memory-wise. The experimental evaluation shows that 3iCubing can considerably reduce the memory needed to index a data set, reducing around 25% of the memory used by Frag-Cubing. Moreover, the results show that the interval inverted index representation is dependent on the data skewness to reduce the memory consumption, having positive results with highly skewed and real-world data sets.

Sphingolipids, ceramides and cholesterol are integral components of cellular membranes, and they also play important roles in signal transduction by regulating the dynamics of membrane receptors through their effects on membrane fluidity. Here, we combined biochemical and functional assays with single-particle tracking analysis of diffusion in the plasma membrane to demonstrate that the local lipid environment regulates CXCR4 organization and function and modulates chemokine-triggered directed cell migration. Prolonged treatment of T cells with bacterial sphingomyelinase promoted the complete and sustained breakdown of sphingomyelins and the accumulation of the corresponding ceramides, which altered both membrane fluidity and CXCR4 nanoclustering and dynamics. Under these conditions CXCR4 retained some CXCL12-mediated signaling activity but failed to promote efficient directed cell migration. Our data underscore a critical role for the local lipid composition at the cell membrane in regulating the lateral mobility of chemokine receptors, and their ability to dynamically increase receptor density at the leading edge to promote efficient cell migration.

NoSQL databases were created with the primary goal of addressing the shortcomings in the efficiency of relational databases, and can be of four types: document, column, key-value, and graph databases. Graph databases can store data and relationships efficiently, and have a flexible and easy-to-understand data schema. In this paper, we perform an experimental evaluation of the four most popular graph databases: JanusGraph, Nebula Graph, Neo4j, and TigerGraph. Database performance is evaluated using the Linked Data Benchmark Council’s Social Network Benchmark (LDBC SNB). In the experiments, we analyze the execution time of the queries, the loading time of the nodes and the RAM and CPU usage for each database. In our analysis, Neo4j was the graph database with the best performance across all metrics.

Loyalty platforms are designed to increase customer loyalty and thus increase consumers’ attraction to purchase. Although successful in increasing brand reach and sales, these platforms fail to meet their primary objective due to a lack of incentives and encouragement for customers to return. Along with the problem in originating sales, they bring excessive costs to brands due to the maintenance and infrastructure required to make the systems feasible. In that sense, recent blockchain technology can help to overcome some of these problems, providing capabilities such as smart contracts, which have the potential to reinvent the way loyalty systems work and solve current problems. Although blockchain is a relatively new technology, some brands are already investigating its usefulness and rebuilding their loyalty systems. However, these platforms are independent and linked directly to a brand. Thus, there is a need for a generic platform capable of creating and managing different loyalty programs, regardless of the size of the business. This paper explores the shortcomings of current loyalty programs identified through the literature review, and proposes a loyalty management system with blockchain integration that allows any retailer to create and manage their loyalty programs and have customers interact directly with multiple programs in a single application.

Transformers are models that implement a mechanism of self-attention, individually weighting the importance of each part of the input data. Their use in image classification tasks is still somewhat limited since researchers have so far chosen Convolutional Neural Networks for image classification and transformers were more targeted to Natural Language Processing (NLP) tasks. Therefore, this paper presents a literature review that shows the differences between Vision Transformers (ViT) and Convolutional Neural Networks. The state of the art that used the two architectures for image classification was reviewed and an attempt was made to understand what factors may influence the performance of the two deep learning architectures based on the datasets used, image size, number of target classes (for the classification problems), hardware, and evaluated architectures and top results. The objective of this work is to identify which of the architectures is the best for image classification and under what conditions. This paper also describes the importance of the Multi-Head Attention mechanism for improving the performance of ViT in image classification.

REST services are nowadays being used to support many businesses, with most major companies exposing their services via REST interfaces (e.g., Google, Amazon, Instagram, and Slack). In this type of scenarios, heterogeneity is prevalent and software is sometimes exposed to unexpected conditions that may activate residual bugs, leading service operations to fail. Such failures may lead to financial or reputation losses (e.g., information disclosure). Although techniques and tools for assessing robustness have been thoroughly studied and applied to a large diversity of domains, REST services still lack practical approaches that specialize in robustness evaluation. In this paper, we present a tool (named bBOXRT) for performing robustness tests over REST services, solely based on minimal information expressed in their interface descriptions. We used bBOXRT to evaluate an heterogeneous set of 52 REST services that comprise 1,351 operations and fit in distinct categories (e.g., public, private, in-house). We were able to disclose several different types of robustness problems, including issues in services with strong reliability requirements and also a few security vulnerabilities. The results show that REST services are being deployed preserving software defects that harm service integration, and also carrying security vulnerabilities that can be exploited by malicious users.

The increase of automated systems in space missions raises concerns about safety and reliability in operations carried out by satellites due to performance degradation. There have been several studies about the automatic planning process, but many approaches are generated with invalid states. The invalid state can be understood as a prohibited, degraded or risky scenario for the domain. This paper proposes an automated planning process with restrictions that enables automatic planners to not generate plans with invalid states. We implement a validator method for the planner software which proves that plan generation matches the restrictions imposed on the domain. In the experiments, we test an automatic planning process that is specific to the aerospace area, where a knowledge base with invalid states is available in the context of the operation of a satellite. Our proposal to carry out the verification of invalid states in automatic planning, can contribute to plans being generated with higher quality, ensuring that the goal of a plan is only achieved through valid intermediate states. It is also expected that the generated plans will be executed with better performance and will require less computational resources, since the search space is reduced.

Abstract The cytokine interferon‐gamma (IFN‐γ) is a master regulator of innate and adaptive immunity involved in a broad array of human diseases that range from atherosclerosis to cancer. IFN‐γ exerts it signaling action by binding to a specific cell surface receptor, the IFN‐γ receptor (IFN‐γR), whose activation critically depends on its partition into lipid nanodomains. However, little is known about the impact of specific lipids on IFN‐γR signal transduction activity. Here, a new conserved cholesterol (chol) binding motif localized within its single transmembrane domain is identified. Through direct binding, chol drives the partition of IFN‐γR2 chains into plasma membrane lipid nanodomains, orchestrating IFN‐γR oligomerization and transmembrane signaling. Bioinformatics studies show that the signature sequence stands for a conserved chol‐binding motif presented in many mammalian membrane proteins. The discovery of chol as the molecular switch governing IFN‐γR transmembrane signaling represents a significant advance for understanding the mechanism of lipid selectivity by membrane proteins, but also for figuring out the role of lipids in modulating cell surface receptor function. Finally, this study suggests that inhibition of the chol‐IFNγR2 interaction may represent a potential therapeutic strategy for various IFN‐γ‐dependent diseases.

NoSQL document databases emerged as an alternative to relational databases for managing large volumes of data. NoSQL document databases ensure big data storage and good query performance and are essential when the data scheme does not fit into the scheme of relational databases. They store their data in the form of documents and can handle unstructured, semi-structured, and structured data. This work evaluates the top three open-source NoSQL document databases: Couchbase, CouchDB, and MongoDB with Yahoo! Cloud Serving Benchmark (YCSB), which has become a standard for NoSQL database evaluation. The performance and scale-up of document databases are assessed using YCSB workloads with a different number of records and threads, where the runtime is measured for each database. In the experimental evaluation, we concluded that MongoDB is the database with the best runtime, except for the workload composed by scan operations. In addition, we identified CouchDB as the database with the best scale-up when varying the number of threads.

Background: Relational databases have been a prevalent technology for decades, using SQL (Structured Query Language) to manage data. However, the emergence of new technologies, such as the web and the cloud, has brought the requirement to handle more complex data. NewSQL is the latest technology that incorporates the ability to scale and ensures the availability of NoSQL (Not Only SQL) without losing the ACID properties (Atomicity, Consistency, Isolation, Durability) associated with relational databases. Methods: We evaluated CockroachDB, MariaDB Xpand, and VoltDB with OSSpal methodology and experimentally using the Star Schema Benchmark (SSB). The scalability and performance capabilities of each database were assessed. Results: Applying the OSSpal methodology, the results showed that MariaDB Xpand outperformed CockroachDB and VoltDB. On the other hand, we concluded that with Star Schema Benchmark, CockroachDB had better scalability, while VoltDB had a faster query execution time. Conclusions: CockroachDB and VoltDB are the best performing databases in terms of scalability and performance.

Abstract Objectives SARS-Cov-2 virus preferentially binds to the Angiotensin Converting Enzyme 2 (ACE2) on alveolar epithelial type II cells, initiating an inflammatory response and tissue damage which may impair surfactant synthesis contributing to alveolar collapse, worsening hypoxia and leading to respiratory failure. The objective of this study is to evaluate the feasibility, safety and efficacy of nebulised surfactant in COVID-19 adult patients requiring mechanical ventilation for respiratory failure. Trial design This study is a dose-escalating randomized open-label clinical trial of 20 COVID-19 patients. Participants This study is conducted in two centres: University Hospital Southampton and University College London Hospitals. Eligible participants are aged ≥18, hospitalised with COVID-19 (confirmed by PCR), who require endotracheal intubation and are enrolled within 24 hours of mechanical ventilation. For patients unable to consent, assent is obtained from a personal legal representative (PerLR) or professional legal representative (ProfLR) prior to enrolment. The following are exclusion criteria: imminent expected death within 24 hours; specific contraindications to surfactant administration (e.g. known allergy, pneumothorax, pulmonary hemorrhage); known or suspected pregnancy; stage 4 chronic kidney disease or requiring dialysis (i.e., eGFR < 30); liver failure (Child-Pugh Class C); anticipated transfer to another hospital, which is not a study site, within 72 hours; current or recent (within 1 month) participation in another study that, in the opinion of the investigator, would prevent enrollment for safety reasons; and declined consent or assent. Intervention and comparator Intervention: The study is based on an investigational drug/device combination product. The surfactant product is Bovactant (Alveofact®), a natural animal derived (bovine) lung surfactant formulated as a lyophilized powder in 108 mg vials and reconstituted to 45 mg/mL in buffer supplied in a prefilled syringe. It is isolated by lung lavage and, by weight, is a mixture of: phospholipid (75% phosphatidylcholine, 13% phosphatidylglycerol, 3% phosphatidylethanolamine, 1% phosphatidylinositol and 1% sphingomyelin), 5% cholesterol, 1% lipid-soluble surfactant-associated proteins (SP-B and SP-C), very low levels of free fatty acid, lyso-phosphatidylcholine, water and 0.3% calcium. The Drug Delivery Device is the AeroFact-COVID™ nebulizer, an investigational device based on the Aerogen® Solo vibrating mesh nebulizer. The timing and escalation dosing plans for the surfactant are as follows. Cohort 1: Three patients will receive 10 vials (1080 mg) each of surfactant at dosing times of 0 hours, 8 hours and 24 hours. 2 controls with no placebo intervention. Cohort 2: Three patients will receive 10 vials (1080 mg) of surfactant at dosing times of 0 hours and 8 hours, and 30 vials (3240 mg) at a dosing time of 24 hours. 2 controls with no placebo intervention. Cohort 3: Three patients will receive 10 vials (1080 mg) of surfactant at a dosing time of 0 hours, and 30 vials (3240 mg) at dosing times of 8 hours and 24 hours. 2 controls with no placebo intervention. Cohort 4: Three patients will receive 30 (3240 mg) vials each of surfactant at dosing times of 0 hours, 8 hours and 24 hours. 2 controls. 2 controls with no placebo intervention. The trial steering committee, advised by the data monitoring committee, will review trial progression and dose escalation/maintenance/reduction after each cohort is completed (48-hour primary outcome timepoint reached) based on available feasibility, adverse event, safety and efficacy data. The trial will not be discontinued on the basis of lack of efficacy. The trial may be stopped early on the basis of safety or feasibility concerns. Comparator: No placebo intervention. All participants will receive usual standard of care in accordance with the local policies for mechanically ventilated patients and all other treatments will be left to the discretion of the attending physician. Main outcomes The co-primary outcome is the improvement in oxygenation (PaO2/FiO2 ratio) and pulmonary ventilation (Ventilation Index (VI), where VI = [RR x (PIP − PEEP) × PaCO2]/1000) at 48 hours after study initiation. The secondary outcomes include frequency and severity of adverse events (AEs), Adverse Device Effects (ADEs), Serious Adverse Events (SAEs) and Serious Adverse Device Events (SADEs), change in pulmonary compliance, change in positive end-expiratory pressure (PEEP) requirement of ventilatory support at 24 and 48 hours after study initiation, clinical improvement defined by time to one improvement point on the ordinal scale described in the WHO master protocol (2020) recorded while hospitalised, days of mechanical ventilation, mechanical ventilator free days (VFD) at day 21, length of intensive care unit stay, number of days hospitalised and mortality at day 28. Exploratory end points will include quantification of SARS-CoV-2 viral load from tracheal aspirates using PCR, surfactant dynamics (synthesis and turnover) and function (surface tension reduction) from deep tracheal aspirate samples (DTAS), surfactant phospholipid concentrations in plasma and DTAS, inflammatory markers (cellular and cytokine) in plasma and DTAS, and blood oxidative stress markers. Randomisation After informed assent, patients fulfilling inclusion criteria will be randomised to 3:2 for the treatment and control arms using an internet-based block randomization service (ALEA tool for clinical trials, FormsVision BV) in combination with electronic data collection. Randomisation will be done by the recruiting centre with a unique subject identifier specific to that centre. Blinding (masking) This is an open-labelled unblinded study. Numbers to be randomised (sample size) The total sample size is 20 COVID-19 mechanically ventilated patients (12 intervention; 8 control). Trial Status Current protocol version is V2 dated 5th of June 2020. The recruitment is currently ongoing and started on the 14th of October 2020. The anticipated study completion date is November 2021. Trial registration ClinicalTrials.gov: NCT04362059 (Registered 24 April 2020), EUDAMED number: CIV-GB-20-06-033328, EudraCT number: 2020-001886-35 (Registered 11 May 2020) Full protocol The full protocol is attached as an additional file, accessible from the Trials website (Additional file 1). In the interest in expediting dissemination of this material, the familiar formatting has been eliminated; this Letter serves as a summary of the key elements of the full protocol. The study protocol has been reported in accordance with the Standard Protocol Items: Recommendations for Clinical Interventional Trials (SPIRIT) guidelines (Additional file 2).

The volume of data is growing exponentially and becoming more valuable to organizations that collect it, from e-commerce data, shipping, audio and video logs, text messages, internet search queries, stock market activity, financial transactions, the Internet of Things, and various other sources. The major challenges are related with the way to extract insights from such a rich data environment and whether Deep Learning can be successful with Big Data. To get some insight on these topics, social network data are employed as a case study on how sentiments can affect decisions in stock market environments. In this paper, we propose a generalized Deep Learning-based classification framework for Stock Market Sentiment Analysis. This work comprises the study, the development, and implementation of an automatic classification system based on Deep Learning and the validation of its adequacy and efficiency in any scenario, particularly Stock Market Sentiment Analysis. Distinct datasets and several Deep Learning approaches with different layers and embedded techniques are used, and their performances are evaluated. These developments show how Deep Learning reacts to distinct contexts. The results also give context on how different techniques with different parameter combinations react to certain types of data. Convolution obtained the best results when dealing with complex data inputs, and long short-term layers kept a memory of data, allowing inputs which are not as common to still be considered for decisions. The models that resulted from Stock Market Sentiment Analysis datasets were applied with some success to real-life problems. The best models reached accuracies of 73% in training and 69% in certain test datasets. In a simulation, a model was able to provide a Return on Investment of 4.4%. The results contribute to understanding how to process Big Data efficiently using Deep Learning and specialized hardware techniques.

Highlights • About 800 million people suffer from mental or behavioral illnesses, which represents more than one in ten people. • We present a web application to support mental health professionals to identify symptoms and diagnose mental disorders. • This application applies standardized diagnostic methods through decision trees, reducing the administration time and providing an instant classification. • Users do not need to have knowledge in programming languages or machine learning.AbstractMental and behavioral disorders are identified and diagnosed by clinical methods that include a detailed description of symptoms collected from the individual and others, including the family, a clinical examination to define mental status, specialized tests and examinations. Diagnostic methods based on structured interviews are considered one of the best approaches for gathering and analyzing symptoms. Therefore, we propose WAM-DT, a Web Application that implements a classification binary tree (CBT) analysis to diagnose mental disorders symptoms, which allows healthcare professionals to create their own questionnaires and generate their own decision trees based on the data collected from their patients.

In recent years, the use of personal data in marketing, scientific and medical investigation, and forecasting future trends has really increased. This information is used by the government, companies, and individuals, and should not contain any sensitive information that allows the identification of an individual. Therefore, data anonymization is essential nowadays. Data anonymization changes the original data to make it difficult to identify an individual. ARX Data Anonymization and Amnesia are two popular open-source tools that simplify this process. In this paper, we evaluate these tools in two ways: with the OSSpal methodology, and using a public dataset with the most recent tweets about the Pfizer and BioNTech vaccine. The assessment with the OSSpal methodology determines that ARX Data Anonymization has better results than Amnesia. In the experimental evaluation using the public dataset, it is possible to verify that Amnesia has some errors and limitations, but the anonymization process is simpler. Using ARX Data Anonymization, it is possible to upload big datasets and the tool does not show any error in the anonymization process. We concluded that ARX Data Anonymization is the one recommended to use in data anonymization.

Simulators allow for the simulation of real-world environments that would otherwise be financially costly and difficult to implement at a technical level. Thus, a simulation environment facilitates the implementation and development of use cases, rendering such development cost-effective and faster, and it can be used in several scenarios. There are some works about simulation environments in Edge Computing (EC), but there is a gap of studies that state the validity of these simulators. This paper compares the execution of the EdgeBench benchmark in a real-world environment and in a simulation environment using FogComputingSim, an EC simulator. Overall, the simulated environment was 0.2% faster than the real world, thus allowing for us to state that we can trust EC simulations, and to conclude that it is possible to implement and validate proofs of concept with FogComputingSim.

Decision support systems with machine learning can help organizations improve operations and lower costs with more precision and efficiency. This work presents a review of state-of-the-art machine learning algorithms for binary classification and makes a comparison of the related metrics between them with their application to a public diabetes and human resource datasets. The two mainly used categories that allow the learning process without requiring explicit programming are supervised and unsupervised learning. For that, we use Scikit-learn, the free software machine learning library for Python language. The best-performing algorithm was Random Forest for supervised learning, while in unsupervised clustering techniques, Balanced Iterative Reducing and Clustering Using Hierarchies and Spectral Clustering algorithms presented the best results. The experimental evaluation shows that the application of unsupervised clustering algorithms does not translate into better results than with supervised algorithms. However, the application of unsupervised clustering algorithms, as the preprocessing of the supervised techniques, can translate into a boost of performance.

Pulmonary surfactant is a complex mixture of lipids and proteins lining the interior of the alveoli, and constitutes the first barrier to both oxygen and pathogens as they progress toward blood circulation. Despite decades of study, the behavior of the pulmonary surfactant at the molecular scale is poorly understood, which hinders the development of effective surfactant replacement therapies, useful in the treatment of several lung-related diseases. In this work, we combined all-atom molecular dynamics simulations, Langmuir trough measurements, and AFM imaging to study synthetic four-component lipid monolayers designed to model protein-free pulmonary surfactant. We characterized the structural and dynamic properties of the monolayers with a special focus on lateral heterogeneity. Remarkably, simulations reproduce almost quantitatively the experimental data on pressure-area isotherms and the presence of lateral heterogeneities highlighted by AFM. Quite surprisingly, the pressure-area isotherms do not show a plateau region, despite the presence of liquid-condensed nanometer-sized domains at surface pressures larger than 20 mN/m. In the simulations, the liquid-condensed domains were small and transient, but they did not coalesce to yield a separate phase. They were only slightly enriched in DPPC and cholesterol, and their chemical composition remained very similar to the overall composition of the monolayer membrane. Instead, they differed from liquid-expanded regions in terms of membrane thickness (in agreement with AFM data), diffusion rates, as well as acyl chain packing and orientation. We hypothesize that such lateral heterogeneities are crucial for lung surfactant function, as they allow both efficient packing, to achieve low surface tension, and sufficient fluidity, critical for rapid adsorption to the air–liquid interface during the breathing cycle.

Differentiation of keratinocytes is critical for epidermal stratification and formation of a protective stratum corneum. It involves a series of complex processes leading through gradual changes in characteristics and functions of keratinocytes up to their programmed cell death via cornification. The stratum corneum is a relatively impermeable barrier, comprised of dead cell remnants (corneocytes) embedded in lipid matrix. Corneocyte membranes are comprised of specialized lipids linked to late differentiation proteins, contributing to the formation of a stiff and mechanically strengthened layer. To date, the assessment of the progression of keratinocyte differentiation is only possible through determination of specific differentiation markers, e.g., by using proteomics-based approaches. Unfortunately, this requires fixation or cell lysis, and currently there is no robust methodology available to study keratinocyte differentiation in living cells in real-time. Here, we explore new live-cell based approaches for screening differentiation advancement in keratinocytes, in a “calcium switch” model. We employ a polarity-sensitive dye, Laurdan, and Laurdan general polarization function (GP) as a reporter of the degree of membrane lateral packing order or condensation, as an adequate marker of differentiation. We show that the assay is straightforward and can be conducted either on a single cell level using confocal spectral imaging or on the ensemble level using a fluorescence plate reader. Such systematic quantification may become useful for understanding mechanisms of keratinocyte differentiation, such as the role of membrane in homogeneities in stiffness, and for future therapeutic development.

T-cell receptor stimulation induces the convergence of multivesicular bodies towards the microtubule-organizing centre (MTOC) and the polarization of the MTOC to the immune synapse (IS). These events lead to exosome secretion at the IS. We describe here that upon IS formation centrosomal area F-actin decreased concomitantly with MTOC polarization to the IS. PKCδ-interfered T cell clones showed a sustained level of centrosomal area F-actin associated with defective MTOC polarization. We analysed the contribution of two actin cytoskeleton-regulatory proteins, FMNL1 and paxillin, to the regulation of cortical and centrosomal F-actin networks. FMNL1$${\rm \beta}$$ phosphorylation and F-actin reorganization at the IS were inhibited in PKCδ-interfered clones. F-actin depletion at the central region of the IS, a requirement for MTOC polarization, was associated with FMNL1$${\rm \beta}$$ phosphorylation at its C-terminal, autoregulatory region. Interfering all FMNL1 isoforms prevented MTOC polarization; nonetheless, FMNL1$${\rm \beta}$$ re-expression restored MTOC polarization in a centrosomal area F-actin reorganization-independent manner. Moreover, PKCδ-interfered clones exhibited decreased paxillin phosphorylation at the MTOC, which suggests an alternative actin cytoskeleton regulatory pathway. Our results infer that PKCδ regulates MTOC polarization and secretory traffic leading to exosome secretion in a coordinated manner by means of two distinct pathways, one involving FMNL1$${\rm \beta}$$ regulation and controlling F-actin reorganization at the IS, and the other, comprising paxillin phosphorylation potentially controlling centrosomal area F-actin reorganization. Abbreviations Ab, antibody; AICD, activation-induced cell death; AIP, average intensity projection; APC, antigen-presenting cell; BCR, B-cell receptor for antigen; C, centre of mass; cent2, centrin 2; cIS, central region of the immune synapse; CMAC, CellTracker™ Blue (7-amino-4-chloromethylcoumarin); cSMAC, central supramolecular activation cluster; CTL, cytotoxic T lymphocytes; DAG, diacylglycerol; DGKα, diacylglycerol kinase α; Dia1, Diaphanous-1; dSMAC, distal supramolecular activation cluster; ECL, enhanced chemiluminescence; ESCRT, endosomal sorting complex required for traffic; F-actin, filamentous actin; Fact-low cIS, F-actin-low region at the centre of the immune synapse; FasL, Fas ligand; FMNL1, formin-like 1; fps, frames per second; GFP, green fluorescent protein; HBSS, Hank’s balanced salt solution; HRP, horseradish peroxidase; ILV, intraluminal vesicles; IS, immune synapse; MFI, mean fluorescence intensity; MHC, major histocompatibility complex; MIP, maximal intensity projection; MVB, multivesicular bodies; MTOC, microtubule-organizing centre; NS, not significant; PBL, peripheral blood lymphocytes; PKC, protein kinase C; PKCδ, protein kinase C δ isoform; PLC, phospholipase C; PMA, phorbol myristate acetate; Pol. Index, polarization index; pSMAC, peripheral supramolecular activation cluster; PSF, point spread function; ROI, region of interest; SD, standard deviation; shRNA, short hairpin RNA; SEE, Staphylococcus enterotoxin E; SMAC, supramolecular activation cluster; TCR, T-cell receptor for antigen; T-helper (Th); TRANS, transmittance; WB, Western blot.

COVID-19 has provoked enormous negative impacts on human lives and the world economy. In order to help in the fight against this pandemic, this study evaluates different databases’ systems and selects the most suitable for storing, handling, and mining COVID-19 data. We evaluate different SQL and NoSQL database systems using the following metrics: query runtime, memory used, CPU used, and storage size. The databases systems assessed were Microsoft SQL Server, MongoDB, and Cassandra. We also evaluate Data Mining algorithms, including Decision Trees, Random Forest, Naive Bayes, and Logistic Regression using Orange Data Mining software data classification tests. Classification tests were performed using cross-validation in a table with about 3 M records, including COVID-19 exams with patients’ symptoms. The Random Forest algorithm has obtained the best average accuracy, recall, precision, and F1 Score in the COVID-19 predictive model performed in the mining stage. In performance evaluation, MongoDB has presented the best results for almost all tests with a large data volume.

Cancer therapy using systemically administrated 4-1BB-targeting antibodies is often associated with severe toxicity due to the nonspecific activation of autoreactive T cells. Here, the authors have developed a trimeric antibody targeting both 4-1BB and EGFR, which activates T cells effectively and shows negligible cytotoxicity.

Abstract Businesses are nowadays deploying their services online, reaching out to clients all around the world. Many times deployed as web applications or web services, these business-critical systems typically perform large amounts of database operations; thus, they are dependent on the quality of the data to provide correct service to clients. Research and practice have shown that the quality of the data in an enterprise system gradually decreases overtime, bringing in diverse reliability issues to the applications that are using the data to provide services. These issues range from simple incorrect operations to aborted operations or severe system failures. In this paper, we present an approach to test data-centric services in presence of poor quality data. The approach has been designed to consider relational and NoSQL database nodes used by the system under test and is based on the injection of poor quality data on the database–application interface. The results indicate the effectiveness of the approach in discovering issues, not only at the application-level, but also in the middleware being used, contributing to the development of more reliable services.

Abstract Background Spectral imaging with polarity-sensitive fluorescent probes enables the quantification of cell and model membrane physical properties, including local hydration, fluidity, and lateral lipid packing, usually characterized by the generalized polarization (GP) parameter. With the development of commercial microscopes equipped with spectral detectors, spectral imaging has become a convenient and powerful technique for measuring GP and other membrane properties. The existing tools for spectral image processing, however, are insufficient for processing the large data sets afforded by this technological advancement, and are unsuitable for processing images acquired with rapidly internalized fluorescent probes. Results Here we present a MATLAB spectral imaging toolbox with the aim of overcoming these limitations. In addition to common operations, such as the calculation of distributions of GP values, generation of pseudo-colored GP maps, and spectral analysis, a key highlight of this tool is reliable membrane segmentation for probes that are rapidly internalized. Furthermore, handling for hyperstacks, 3D reconstruction and batch processing facilitates analysis of data sets generated by time series, z-stack, and area scan microscope operations. Finally, the object size distribution is determined, which can provide insight into the mechanisms underlying changes in membrane properties and is desirable for e.g. studies involving model membranes and surfactant coated particles. Analysis is demonstrated for cell membranes, cell-derived vesicles, model membranes, and microbubbles with environmentally-sensitive probes Laurdan, carboxyl-modified Laurdan (C-Laurdan), Di-4-ANEPPDHQ, and Di-4-AN(F)EPPTEA (FE), for quantification of the local lateral density of lipids or lipid packing. Conclusions The Spectral Imaging Toolbox is a powerful tool for the segmentation and processing of large spectral imaging datasets with a reliable method for membrane segmentation and no ability in programming required. The Spectral Imaging Toolbox can be downloaded from https://uk.mathworks.com/matlabcentral/fileexchange/62617-spectral-imaging-toolbox .

Wine is the second most popular alcoholic drink in the world behind beer. With the rise of e-commerce, recommendation systems have become a very important factor in the success of business. Recommendation systems analyze metadata to predict if, for example, a user will recommend a product. The metadata consist mostly of former reviews or web traffic from the same user. For this reason, we investigate what would happen if the information analyzed by a recommendation system was insufficient. In this paper, we explore the effects of a new wine ontology in a recommendation system. We created our own wine ontology and then made two sets of tests for each dataset. In both sets of tests, we applied four machine learning clustering algorithms that had the objective of predicting if a user recommends a wine product. The only difference between each set of tests is the attributes contained in the dataset. In the first set of tests, the datasets were influenced by the ontology, and in the second set, the only information about a wine product is its name. We compared the two test sets’ results and observed that there was a significant increase in classification accuracy when using a dataset with the proposed ontology. We demonstrate the general applicability of the methodology to other cases, applying our proposal to an Amazon product review dataset.

Receptor clustering is known to trigger signalling events that contribute to critical changes in cellular functions. Faithful imaging of such clusters by means of fluorescence microscopy relies on the application of adequate cell fixation methods prior to immunolabelling in order to avoid artefactual redistribution by the antibodies themselves. Previous work has highlighted the inadequacy of fixation with paraformaldehyde (PFA) alone for efficient immobilisation of membrane-associated molecules, and the advantages of fixation with PFA in combination with glutaraldehyde (GA). Using fluorescence microscopy, we here highlight how inadequate fixation can lead to the formation of artefactual clustering of receptors in lymphatic endothelial cells, focussing on the transmembrane hyaluronan receptors LYVE-1 and CD44, and the homotypic adhesion molecule CD31, each of which displays their native diffuse surface distribution pattern only when visualised with the right fixation techniques, i.e. PFA/GA in combination. Fluorescence recovery after photobleaching (FRAP) confirms that the artefactual receptor clusters are indeed introduced by residual mobility. In contrast, we observed full immobilisation of membrane proteins in cells that were fixed and then subsequently permeabilised, irrespective of whether the fixative was PFA or PFA/GA in combination. Our study underlines the importance of choosing appropriate sample preparation protocols for preserving authentic receptor organisation in advanced fluorescence microscopy.

Every year tens of thousands of Portuguese youths applied for a position in higher education, one of the most important choices that make throughout their lives. In this paper we propose a portal that contains all the criteria and indicators that are available and relevant during the process of choosing a course, providing the resources and the necessary options, so that it can be the candidates and not the State or any other entity to choose their priorities. According to the proposed algorithm, assigning a greater or lesser importance to criteria and indicators that one can consider relevant according to his/her objectives, the candidate can generate several simulations creating a personal ranking of the pair course/institution. Keywords: Higher Education Access; Education; Decision Support Systems.

The ability of keeping a record of geospatial information, knowing how it changed over time, is crucial for landscape analysis and territorial government. Land management is still a problem. Many governmental databases are incomplete, and there is a lack of reliable information. Good land management implies having a tool that can keep track of all the information available about a certain property and its changes over time. In this paper, we propose a land management tool where managers access all the information on a certain parcel of land—its boundaries, the land registration, a map which verifies the landcover, and the historic of updates of territorial limits. With the proposed tool, it is possible to edit the information of any property, whether it is active or not—that is, to also edit properties that no longer exist today, but that the user wants to add information to, for legal or other reasons. Keeping track of data properties’ revision history is groundbreaking due to the fact it is not well developed in existing tools. We will look at Brazil as a use case, where land management is a critical problem.

Cyber-Physical Systems (CPS) are a prominent component of the modern digital transformation, which combines the dynamics of the physical processes with those of software and networks. Critical infrastructures have built-in CPS, and assessing its risk is crucial to avoid significant losses, both economic and social. As CPS are increasingly attached to the world’s main industries, these systems’ criticality depends not only on software efficiency and availability but also on cyber-security awareness. Given this, and because Failure Mode and Effect Analysis (FMEA) is one of the most effective methods to assess critical infrastructures’ risk, in this paper, we show how this method performs in the analysis of CPS threats, also exposing the main drawbacks concerning CPS risk assessment. We first propose a risk prevention analysis to the Communications-Based Train Control (CBTC) system, which involves exploiting cyber vulnerabilities, and we introduce a novel approach to the failure modes’ Risk Priority Number (RPN) estimation. We also propose how to adapt the FMEA method to the requirement of CPS risk evaluation. We applied the proposed procedure to the CBTC system use case since it is a CPS with a substantial cyber component and network data transfer.

The growth of the Internet has increased the amount of data and information available to any person at any time. Recommendation Systems help users find the items that meet their preferences, among the large number of items available. Techniques such as collaborative filtering and content-based recommenders have played an important role in the implementation of recommendation systems. In the last few years, other techniques, such as, ontology-based recommenders, have gained significance when reffering better active user recommendations; however, building an ontology-based recommender is an expensive process, which requires considerable skills in Knowledge Engineering. This paper presents a new hybrid approach that combines the simplicity of collaborative filtering with the efficiency of the ontology-based recommenders. The experimental evaluation demonstrates that the proposed approach presents higher quality recommendations when compared to collaborative filtering. The main improvement is verified on the results regarding the products, which, in spite of belonging to unknown categories to the users, still match their preferences and become recommended.

Internet has become so widespread that most popular websites are accessed by hundreds of millions of people on a daily basis. Monolithic architectures, which were frequently used in the past, were mostly composed of traditional relational database management systems, but quickly have become incapable of sustaining high data traffic very common these days. Meanwhile, NoSQL databases have emerged to provide some missing properties in relational databases like the schema-less design, horizontal scaling, and eventual consistency. This paper analyzes and compares the consistency model implementation on five popular NoSQL databases: Redis, Cassandra, MongoDB, Neo4j, and OrientDB. All of which offer at least eventual consistency, and some have the option of supporting strong consistency. However, imposing strong consistency will result in less availability when subject to network partition events.

Ever since technologies enabled the characterisation of eukaryotic plasma membranes, heterogeneities in the distributions of its constituents were observed. Over the years this led to the proposal of various models describing the plasma membrane organisation such as lipid shells, picket-and-fences, lipid rafts, or protein islands, as addressed in numerous publications and reviews. Instead of emphasising on one model we in this review give a brief overview over current models and highlight how current experimental work in one or the other way do not support the existence of a single overarching model. Instead, we highlight the vast variety of membrane properties and components, their influences and impacts. We believe that highlighting such controversial discoveries will stimulate unbiased research on plasma membrane organisation and functionality, leading to a better understanding of this essential cellular structure.

The digital era has brought a number of significant changes in the world of communications. Although technological evolution has allowed the creation of new social event platforms to disclose events, it is still difficult to know what is happening around a location. Currently, a large number of social events are created and promoted on social networks. With the massive quantity of information created in these systems, finding an event is challenging because sometimes the data is ambiguous or incomplete. One of the main challenges in social event classification is related to the incompleteness and ambiguity of metadata created by users. This paper presents a new ontology, named LODSE (Linking Open Descriptions of Social Events) based on the LODE (Linking Open Descriptions of Events) ontology to describe the domain model of social events. The aim of this ontology is to create a data model that allows definition of the most important properties to describe a social event and to improve the classification of events. The proposed data model is used in an experimental evaluation to compare both ontologies in social event classification. The experimental evaluation, using a dataset based on real data from a popular social network, demonstrated that the data model based on the LODSE ontology brings several benefits in the classification of events. Using the LODSE ontology, the results show an increment of correctly classified events as well as a gain in execution time, when comparing with the data model based on the LODE ontology.