Information

Santana-Mancilla, Pedro C.;Castrejón-Mejía, Oscar E.;Fajardo-Flores, Silvia B.;Anido-Rifón, Luis E.

doi: 10.3390/info14120625pmid: N/A

Wearable Internet of Medical Things (IoMT) technology, designed for non-invasive respiratory monitoring, has demonstrated considerable promise in the early detection of severe diseases. This paper introduces the application of supervised machine learning techniques to predict respiratory abnormalities through frequency data analysis. The principal aim is to identify respiratory-related health risks in older adults using data collected from non-invasive wearable devices. This article presents the development, assessment, and comparison of three machine learning models, underscoring their potential for accurately predicting respiratory-related health issues in older adults. The convergence of wearable IoMT technology and machine learning holds immense potential for proactive and personalized healthcare among older adults, ultimately enhancing their quality of life.

Garay-Jiménez, Laura I.;Romero-Lujambio, Jose Fausto;Santiago-Horta, Amaury;Tovar-Corona, Blanca;Gómez-Miranda, Pilar;Mata-Rivera, Miguel Félix

doi: 10.3390/info14120626pmid: N/A

Environmental research calls for a multidisciplinary approach, where highly specialized research teams collaborate in data analysis. Nevertheless, managing the data lifecycle and research artifacts becomes challenging because the project teams require techniques and tools tailored to their study fields. Another pain point is the unavailability of essential analysis and data representation formats for querying and interpreting the shared results. In addition, managing progress reports across the teams is demanding because they manage different platforms and systems. These concerns discourage the knowledge-sharing process and lead to researchers’ low adherence to the system. A hybrid methodology based on Design Thinking and an Agile approach enables us to understand and attend to the research process needs. As a result, a microservices-based architecture of the system, which can be deployed in cloud, hybrid, or standalone environments and adapt the computing resources according to the actual requirements with an access control system based on users and roles, enables the security and confidentiality, allowing the team’s lead to share or revoke access. Additionally, intelligent assistance is available for document searches and dataset analyses. A multidisciplinary researchers’ team that uses this system as a knowledge-sharing workspace reported an 83% acceptance.

Põldvere, Nele;Uddin, Zia;Thomas, Aleena

doi: 10.3390/info14120627pmid: N/A

This study presents a new dataset for fake news analysis and detection, namely, the PolitiFact-Oslo Corpus. The corpus contains samples of both fake and real news in English, collected from the fact-checking website PolitiFact.com. It grew out of a need for a more controlled and effective dataset for fake news analysis and detection model development based on recent events. Three features make it uniquely placed for this: (i) the texts have been individually labelled for veracity by experts, (ii) they are complete texts that strictly correspond to the claims in question, and (iii) they are accompanied by important metadata such as text type (e.g., social media, news and blog). In relation to this, we present a pipeline for collecting quality data from major fact-checking websites, a procedure which can be replicated in future corpus building efforts. An exploratory analysis based on sentiment and part-of-speech information reveals interesting differences between fake and real news as well as between text types, thus highlighting the importance of adding contextual information to fake news corpora. Since the main application of the PolitiFact-Oslo Corpus is in automatic fake news detection, we critically examine the applicability of the corpus and another PolitiFact dataset built based on less strict criteria for various deep learning-based efficient approaches, such as Bidirectional Long Short-Term Memory (Bi-LSTM), LSTM fine-tuned transformers such as Bidirectional Encoder Representations from Transformers (BERT) and RoBERTa, and XLNet.

Oh, Seung-Taek;Lim, Jae-Hyun

doi: 10.3390/info14120628pmid: N/A

Natural light continuously changes its correlated color temperature (CCT) from sunrise to sunset, providing the best color reproducibility and healthy light. In the lighting field, efforts have been made to improve the Color Rendering Index (CRI) to provide light quality at the same level as natural light. A unique light source technology that mixes and controls multiple LED light sources with different spectral or CCT characteristics or provides a high color rendering index has been introduced. However, the characteristics of natural light, which provide high CRI light while changing color temperature every moment, could not be reproduced as they were. Therefore, in this paper, we propose a CRI-based smart lighting system that reproduces natural light characteristics, provides light with high color reproducibility, and maintains homeostasis even under the changing environment of natural light CCT. After extracting the CCT for each day from the characteristics of measured natural light, the light with the highest CRI under the CCT condition for each hour was provided through a CRI-based CCT matching algorithm. Performance evaluation was conducted for four-channel LED experimental lighting. For each clear and cloudy day, daily natural light was reproduced with a light quality higher than average CRI 98 within the MAE range of CCT 6.78 K.

Gupta, Khushi;Oladimeji, Damilola;Varol, Cihan;Rasheed, Amar;Shahshidhar, Narasimha

doi: 10.3390/info14120629pmid: N/A

Social media applications have been ubiquitous in modern society, and their usage has grown exponentially over the years. With the widespread adoption of these platforms, social media has evolved into a significant origin of digital evidence in the domain of digital forensics. The increasing utilization of social media has caused an increase in the number of studies focusing on artifact (digital remnants of data) recovery from these platforms. As a result, we aim to present a comprehensive survey of the existing literature from the past 15 years on artifact recovery from social media applications in digital forensics. We analyze various approaches and techniques employed for artifact recovery, structuring our review on well-defined analysis focus categories, which are memory, disk, and network. By scrutinizing the available literature, we determine the trends and commonalities in existing research and further identify gaps in existing literature and areas of opportunity for future research in this field. The survey is expected to provide a valuable resource for academicians, digital forensics professionals, and researchers by enhancing their comprehension of the current state of the art in artifact recovery from social media applications. Additionally, it highlights the need for continued research to keep up with social media’s constantly evolving nature and its consequent impact on digital forensics.

Shen, Guan-Ting;Huang, Yin-Fu

doi: 10.3390/info14120630pmid: N/A

Image segmentation is the process of partitioning an image into multiple segments where the goal is to simplify the representation of the image and make the image more meaningful and easier to analyze. In particular, semantic segmentation is an approach of detecting the classes of objects, based on each pixel. In the past, most semantic segmentation models were for only one single style, such as urban street views, medical images, or even manga. In this paper, we propose a semantic segmentation model called the Dual-Pyramid Wide Residual Network (DPWRN) to solve the segmentation on cross-style datasets, which is suitable for diverse segmentation applications. The DPWRN integrated the Pyramid of Kernel paralleled with Dilation (PKD) and Multi-Feature Fusion (MFF) to improve the accuracy of segmentation. To evaluate the generalization of the DPWRN and its superiority over most state-of-the-art models, three datasets with completely different styles are tested in the experiments. As a result, our model achieves 75.95% of mIoU on CamVid, 83.60% of F1-score on DRIVE, and 86.87% of F1-score on eBDtheque. This verifies that the DPWRN can be generalized and shows its superiority in semantic segmentation on cross-style datasets.

Florea, Marcu;Esteves, Beatriz

doi: 10.3390/info14120631pmid: N/A

Personal Information Management Systems (PIMS) are acquiring a prominent role in the data economy by promoting services that help individuals to have more control over the processing of their personal data, in line with the European data protection laws. One of the highlighted solutions in this area is Solid, a new protocol that is decentralizing the storage of data, through the usage of interoperable web standards and semantic vocabularies, to empower its users to have more control over the processing of data by agents and applications. However, to fulfill this vision and gather widespread adoption, Solid needs to be aligned with the law governing the processing of personal data in Europe, the main piece of legislation being the General Data Protection Regulation (GDPR). To assist with this process, we analyze the current efforts to introduce a policy layer in the Solid ecosystem, in particular, related to the challenge of obtaining consent for processing personal data, focusing on the GDPR. Furthermore, we investigate if, in the context of using personal data for biomedical research, consent can be expressed in advance, and discuss the conditions for valid consent and how it can be obtained in this decentralized setting, namely through the matching of privacy preferences, set by the user, with requests for data and whether this can signify informed consent. Finally, we discuss the technical challenges of an implementation that caters to the previously identified legal requirements.

Lishner, Itai;Shtub, Avraham

doi: 10.3390/info14120632pmid: N/A

The selection of an appropriate development methodology is a critical strategic decision when managing a New Product Development (NPD) project. However, accurately estimating project duration based on the chosen methodology remains a challenge. This paper addresses the limitations of existing models and proposes a novel NPD project model that allows for testing and evaluation of different product development strategies. The model considers Waterfall, Spiral, Agile, and Hybrid methodologies and provides system engineers and project managers with decision-making tools to determine the optimal strategy and understand associated tradeoffs. The model is validated using real projects from various organizations and methodologies. It incorporates stochastic variables, risk management, and dynamic resource allocation, while addressing both Waterfall and Agile methodologies. The study contributes to the body of knowledge by offering practical tools for system engineers and project managers for choosing development methodology, improving project duration estimation, and identifying critical processes and risks in NPD projects. The research results also provide a basis for further studies and can benefit researchers interested in systems engineering methodologies. The proposed model fills a gap in the literature by providing a validated NPD model to evaluate the impact of different product development methodologies on project duration.

Single, Sam;Iranmanesh, Saeid;Raad, Raad

doi: 10.3390/info14120633pmid: N/A

The accurate classification of landfill waste diversion plays a critical role in efficient waste management practices. Traditional approaches, such as visual inspection, weighing and volume measurement, and manual sorting, have been widely used but suffer from subjectivity, scalability, and labour requirements. In contrast, machine learning approaches, particularly Convolutional Neural Networks (CNN), have emerged as powerful deep learning models for waste detection and classification. This paper analyses VGG-16, InceptionResNetV2, DenseNet121, Inception V3, and MobileNetV2 models to classify real-life waste when trained on pristine and unadulterated materials, versus samples collected at a landfill site. When training on DiversionNet, the unadulterated material dataset with labels required for landfill modelling, classification accuracy was limited to 49.69% in the real environment. Using real-world samples in the newly formed RealWaste dataset showed that practical applications for deep learning in waste classification are possible, with Inception V3 reaching 89.19% classification accuracy on the full spectrum of labels required for accurate modelling.

Guesmi, Amira;Hanif, Muhammad Abdullah;Shafique, Muhammad

doi: 10.3390/info14120634pmid: N/A

Vision-based perception modules are increasingly deployed in many applications, especially autonomous vehicles and intelligent robots. These modules are being used to acquire information about the surroundings and identify obstacles. Hence, accurate detection and classification are essential to reach appropriate decisions and take appropriate and safe actions at all times. Current studies have demonstrated that “printed adversarial attacks”, known as physical adversarial attacks, can successfully mislead perception models such as object detectors and image classifiers. However, most of these physical attacks are based on noticeable and eye-catching patterns for generated perturbations making them identifiable/detectable by the human eye, in-field tests, or in test drives. In this paper, we propose a camera-based inconspicuous adversarial attack (AdvRain) capable of fooling camera-based perception systems over all objects of the same class. Unlike mask-based FakeWeather attacks that require access to the underlying computing hardware or image memory, our attack is based on emulating the effects of a natural weather condition (i.e., Raindrops) that can be printed on a translucent sticker, which is externally placed over the lens of a camera whenever an adversary plans to trigger an attack. Note, such perturbations are still inconspicuous in real-world deployments and their presence goes unnoticed due to their association with a natural phenomenon. To accomplish this, we develop an iterative process based on performing a random search aiming to identify critical positions to make sure that the performed transformation is adversarial for a target classifier. Our transformation is based on blurring predefined parts of the captured image corresponding to the areas covered by the raindrop. We achieve a drop in average model accuracy of more than 45% and 40% on VGG19 for ImageNet dataset and Resnet34 for Caltech-101 dataset, respectively, using only 20 raindrops.