Anxiety and depression are psychological disorders characterized by persistent and impairing symptoms. They affect millions of people worldwide and have a significant impact on individuals' well-being and daily functioning. Although highly effective treatments exist, delayed diagnoses and limited access to mental health care contribute to a significant number of undiagnosed individuals. Therefore, it is important to explore predictive modeling to anticipate and address potential issues before the symptoms increase. In that context, this study proposes a machine learning approach to predict anxiety and depression scores based on the Generalized Anxiety Disorder (GAD-7) and Patient Health Questionnaire (PHQ-9). In a regression scenario the proposed multi-layer perceptron (MLP) achieved the lowest MAE values of 5.3924 for anxiety and 5.06 for depression, as well as the lowest MAPE values of 0.1101 for anxiety and 0.1043 for depression. For a classification scenario the best-performing models were the random forest (RF) and LightGBM with an F1score of 0.8997 and 0.8918 for anxiety, respectively, and 0.7593 and 0.7480 for depression. These results highlight the potential of neural network-based models to outperform traditional ensemble and kernel-based approaches to predict mental disorder scores. Additionally, the classification results also suggest that tree and kernel-based models can effectively maintain balanced predictive performance.

The High-Pressure Die Casting (HPDC) process is a critical manufacturing technique for producing complex and high-precision parts, but it faces significant challenges related to equipment longevity, predictive maintenance, and decision support. This paper addresses three major challenges in HPDC environments: predicting the remaining useful life (RUL) of dies, developing an integrated pipeline for anomaly detection and maintenance forecasting, and implementing a real-time dashboard for operational decision-making. A predictive maintenance framework was proposed, combining traditional machine learning models as Light Gradient Boosting Machine (LightGBM) with advanced Large Language Models (LLMs) enhanced by Retrieval-Augmented Generation (RAG). This framework enables proactive fault detection, anomaly classification, failure impact foresight, and operational interruption forecasting. To ensure that these predictive insights are accessible to industry practitioners, a customized dashboard was developed to integrate real-time monitoring, predictive alerts, and maintenance recommendations. The proposed approach enhances operational efficiency, optimizes die usage, and promotes a data-driven maintenance culture in HPDC industries.

A mental disorder is a clinically significant disturbance in an individual's cognition, emotional, or behavioral functioning.Mental disorders such as anxiety and depression can be accessed by psychiatrists using auxiliary tools such as the depression anxiety stress scale (DASS), patient reported outcome (PRO), patient reported outcome measures (PROMs) and patient reported outcomes measurement information system (PROMIS ® ).However, many individuals affected by the symptoms of mental disorders do not receive a proper diagnosis.In that context, this work proposes a machine learning approach to predict the score of anxiety and depression using PROMIS ® questionnaires by performing a comparative study between supervised learning models to estimate the scores of anxiety and depression from individuals.Through the proposed model an average MAPE of 6.31%, R² of 0.76, and Spearman coefficient of 88.86 were achieved, outperforming widely used linear models such as support vector machines (SVM), random forest (RF), and gradient boosting (GB).In conclusion, the utilization of machine learning algorithms with PROMIS ® questionnaires has shown promise as a methodology for assessing anxiety and depression scores from the participants' perspective, aligning with their perceptions of well-being.

The growth of assistive technologies brings new perspectives to Speech Sound Disorders (SSD) treatment. For example, automatic Speech Recognition (ASR) tools recognize and convert sound signals into text in multiple languages. Commonly, these tools rely on models trained with samples from typically developed speakers, but most of them can deal well with sonorous variations such as accents. Hence, there is an expectation that they may also transcribe phonological disorders, such as those produced by people with SSD. However, this potential remains poorly known. Here, we analyze the potential of one of the ASR tools, Google’s speech-to-text API©, as a multilevel indicator of speech intelligibility. We used pronunciations of volunteer actors, which simulated people with a broad spectrum of speech impairments. The tool indicated speech intelligibility at a general level and was marginally capable of determining the SSD type, but it could not map the syllable exchanges accurately. In short, our results suggest that ASRs have great potential as components of assistive tools in many contexts. Our contribution goes beyond the tests themselves, as we propose a simple, robust, systematic, and automated method to quantify speech intelligibility using ASRs. The method, which still needs clinical validation, can be replicated in other versions and tools and the pronunciations of people who are genuinely SSD carriers or in other languages, as long as they use the appropriate protocols. The goal is to enhance ASR tools’ capabilities to promote even higher digital inclusion for people with phonological disorders.

Symmetries express the invariance of a system towards sets of mathematical transformations. In more practical terms, symmetries greatly reduce or simplify the computational efforts required to evaluate relevant properties of a system. In this paper, two methods are proposed to implement spin symmetries which simplify the analysis of the spreading of diseases in an agent-based epidemic model. We perform a set of simulations to measure the efficiency gains compared to traditional methods. Our findings show symmetry-based algorithms improve the performance of the Monte Carlo simulation and the exact Markov process.

For people with speech disorders, language therapies usually employ speech exercises performed at home, without close specialized supervision. There are fundamental movements to train speech ability called NonVerbal Praxis (NVP). These movements include blows and tongue snaps, which develop both the neuromotor capacities of people with apraxia and the pronunciation of phonemes. However, current research on NVP recognition has employed standard speech recognition techniques without much success. This paper describes a method to identify NVP under field conditions. The method covers the input audio signal segmentation until the audio segments are classified and scored. Our method was experimented with, and NVP sounds were efficiently classified, which helped guide patients and therapists through home therapy. It also provides valuable information for speech-language professionals to evaluate the patient's therapeutic performance.

The aim of this work is twofold. First, we identify web accessibility requirements for deaf people who communicate using Sign Language. Second, we define automation approaches for accessibility testing according to the requirements identified. The requirements were identified through a literature review that considered laws, standards, guides, and scientific articles. This review showed the lack of tools that automate the accessibility test for the Deaf. Thus, we propose two approaches to automating accessibility testing: one that requires the programmer to encode metadata for accessibility testing, and another that relies only on descriptive analysis of data from other web pages accessible to the deaf. Our work was based on the analysis of 150 sites, 100 of them with metadata analysis, and 50 with descriptive analysis. We discuss the tradeoff for each approach.

The low performance of software projects generally arises from erroneous, omitted, misinterpreted, or conflicting requirements. To produce better quality specifications, the practise of requirements reuse through requirement patterns has been widely debated in the secondary literature. However, a tertiary study that provides an overview of secondary studies on the state of the art and the practise of requirement patterns does not exist. This study describes a study of secondary literature on requirement patterns under a perspective on research and practise. The identification and selection methods of secondary studies include automatic search on five sources, inclusion, and exclusion criteria, the snowballing technique, and the quality assessment of those studies. Four secondary studies are considered relevant according to the purpose of this research from a 26‐distinct‐study group. The authors’ contribution is two‐fold: the tertiary study itself and a preliminary research agenda dealing with state of the art and practise on requirement patterns.

The literature describes examples of software frameworks providing developers with generic and reusable functionality for building healthcare applications.Using concepts and technologies from Information Retrieval, Machine Learning, and Semantic Web, we present a novel software framework called HSSF (Health Surveillance Software Framework) which aims to facilitate the development of applications to support health professionals in the prevention of chronic diseases.The main contribution of this paper includes lessons learned distilled from (i) the reuse and evolution of the HSSF components on the development of three new health surveillance applications, and (ii) a quantitative evaluation of the HSSF reusability in terms of time spent and artifacts reused on such development task.Lessons learned are summarized as advantages and drawbacks regarding HSSF reusability.The HSSF allows healthcare applications not only to relate scientific research evidences, exams and treatments, but also to incorporate them together into the clinical practice.

This paper demonstrates our ongoing experience on a documenter system for medical grand rounds. The system captures and synchronizes the set of material presented and corresponding physicians' interactions, automatically relates clinical cases of patients, and then generates web-accessible documents with all information captured. The resulting documentation can be used for several purposes such as teaching, research and presurgical decision taking.

Both content driven web authors and application designers may have their attention deviated from their main task when they have to be concerned with the generation of elaborated linking structures. This work aims to demonstrate how a metadata-enhanced web-based open linking service can be exploited towards supporting content driven authors in their tasks. The following results are presented in this paper: (a) the Web Linking Service (WLS), a novel open hypermedia system that stores and exchanges metadata in RDF standard syntax for hypertext structures across the wire and (b) two case studies in which applications offer to their users the ability to create linking structures upon existing contents by making use the WLS service.