IoT has gradually exposed society to intelligent environments. Software developed for these environments requires efficient data processing, low response time, and proper functioning of sensors, devices, and systems. To meet these software requirements, we can leverage edge, fog, and cloud computing. However, the use of these computational resources presents challenges for software engineering, such as determining which architectures to employ for developing software in intelligent environments. Considering these challenges, this work addresses the research question: How can a self-adaptive architecture support automated computational resource allocation in e-health environments? To answer this research question, we propose a self-adaptive IoT architecture that uses artificial intelligence to manage computational resource usage in intelligent environments, enabling the management of physical spaces and ensuring the correct functioning of applications. A case study was conducted in an e-health environment to support our arguments. The Design Science Research methodology was used to develop the research, and its execution cycles in a real e-health corporate environment, through a case study, enabled the incremental construction of the architecture. The results demonstrate that the proposed architecture enhances the efficiency of allocating computational resources - encompassing edge, fog, and cloud computing — while ensuring the functioning of applications and supporting the management of the physical environment using artificial intelligence. As contributions, the study shows: (i) the self-adaptive architecture construction phases; (ii) how architecture adapts to the demands of the IoT intelligent environment; (iii) how artificial intelligence can support the allocation of computational resources.

One of the challenges of predictive maintenance is making data-driven decisions in an agile and assertive way. Connected sensors and operational data favor intelligent processing techniques to enrich information and support decision-making. Digital Twins (DTs) provide a real-time representation of physical machines, enabling data-driven predictive maintenance. This enables assertive, quick decision-making, aligning with the Industry 5.0 pillars: human-centric, sustainable, and resilient industry transformation. The main contribution of this work is the specification of a suite of services for DT creation, called DT-CREATE, focused on informed decision support for predictive maintenance. DT-CREATE suite utilizes intelligent techniques, processes semantic data, and incorporates self-adaptation features. The evaluation combines two real industrial case studies (textile and heat-treatment furnace) with a comparison against single-algorithm and framework-level baselines, an ablation, a retrospective analysis of operational KPIs, and an evaluation on the public NASA C-MAPSS benchmark. The results demonstrate the importance of using DT-CREATE in the specification of DTs, considering the following aspects: (i) collection, storage, and intelligent processing of data generated by sensors, (ii) enrichment of information through machine learning and ontologies, (iii) use of intelligent techniques to select predictive models that adhere to the available dataset, and (iv) decision support and self-adaptation. The proposal addresses human centricity through an informed decision-making process, sustainability through preventive equipment maintenance, extended equipment usage time, and resilience to prevent equipment breakdowns.

Intelligent systems, particularly in e-health, must meet fundamental ethical and privacy requirements, especially in contexts involving extensive data collection. We propose an architecture based on federated learning and edge computing to detect potential stress moments. The proposed solution ensures scalability, security, and model customization while protecting sensitive data. Federated training enables machine learning models to be trained locally on users’ devices, thereby improving efficiency and security. Personal data is not shared; only the parameters of the locally trained models are used to build a global model, thereby preserving information privacy. An intelligent mobile system was developed and used by five volunteer users. Data from these users was collected, and the trained models detected potential moments of stress.

One of the challenges of predictive maintenance is making decisions based on data in an agile and assertive way. Connected sensors and operational data favor intelligent processing techniques to enrich information and enable decision-making. IoT devices connected to equipment can generate data that predictive maintenance can use to make assertive and quick decisions. Intelligent techniques can be used to process this information and support decision-making. The main contribution of this work is a mapping study considering works related to the use of intelligent data analysis techniques in the industrial predictive maintenance context. Through systematic mapping, 34 studies were analyzed, revealing four main categories of approaches: learning from historical data, context-sensitive data analysis, data classification in quality control, and maintenance planning through simulation. The results show that 75% of studies use supervised learning models, with artificial neural networks being the most frequently applied technique, followed by decision trees and logistic regression. The research identified significant gaps regarding self-adaptation capabilities of systems and the absence of semantic models combined with artificial intelligence algorithms to promote system adaptability.

As Tecnologias Digitais da Informação e Comunicação (TDICs) têm transformado o cenário educacional, exigindo novas abordagens pedagógicas. Este estudo, de natureza qualitativa e quantitativa, é o relato de experiência com a plataforma Khan Academy como ferramenta complementar ao ensino de Matemática para alunos do 9º ano do Ensino Fundamental. O estudo avaliou a potencialidade, usabilidade e interatividade da plataforma, bem como sua contribuição para a compreensão de conteúdos matemáticos e a extração de dados para acompanhamento do desempenho discente. A Khan Academy, com seus recursos de videoaulas, artigos e exercícios, demonstrou ser um ambiente promissor para promover a aprendizagem personalizada e auxiliar o trabalho docente, conforme corroborado por estudos anteriores sobre o uso de TDICs na educação.

Background/Objectives: Electronic medical record systems play a crucial role in the operation of modern healthcare institutions, enabling the foundational data necessary for advancements in personalized medicine. Despite their importance, the software supporting these systems frequently experiences data availability and integrity issues, particularly concerning patients’ personal information. This study aims to present a decentralized architecture that integrates both clinical and personal patient data, with a provenance mechanism to enable data tracing and auditing, ultimately supporting more precise and personalized healthcare decisions. Methods: A system implementation based on the solution was developed, and a feasibility study was conducted with synthetic medical records data. Results: The system was able to correctly receive data of 190 instances of the entities designed, which included different types of medical records, and generate 573 provenance entries that captured in detail the context of the associated medical information. Conclusions: For the first cycle of the research, the system developed served to validate the main features of the solution, and through that, it was possible to infer the feasibility of a decentralized EHR and PHR health system with formal provenance data tracking. Such a system lays a robust foundation for secure and reliable data management, which is essential for the effective implementation and future development of personalized medicine initiatives.

One of the challenges of predictive maintenance is making decisions based on data in an agile and assertive way. Connected sensors and operational data favor intelligent processing techniques to enrich information and enable decision-making. Digital Twins (DTs) can be used to process information and support decision-making. DTs are a real-time representation of physical machines and generate data that predictive maintenance can use to make assertive and quick decisions. The main contribution of this work is the specification of a suite of services for specifying DTs, called DT-Create, focused on decision support in predictive maintenance. DT-Create suite is based on intelligent techniques, semantic data processing, and self-adaptation. This suite was developed using the Design Science Research (DSR) methodology through two development cycles and evaluated through case studies. The results demonstrate the feasibility of using DT-Create in specifying DTs considering the following aspects: (i) collection, storage, and intelligent processing of data generated by sensors, (ii) enrichment of information through machine learning and ontologies, (iii) use of intelligent techniques to select predictive models that adhere to the available data set, and (iv) decision support and self-adaptation.

Background. The rapid growth of online data has made retrieving relevant information a challenging task, prompting the rise of Knowledge Base Question Answering (KBQA) systems that handle complex, multi-hop queries. Purpose. This extended work refines our previous pipeline by introducing structured dummy templates, a Hereditary Tree-LSTM (HTL) for classification, and more comprehensive analyses of entity recognition, property extraction, and SPARQL assembly. Methods. We enhanced the LC-QUAD 2.1 dataset with standardized templates and evaluated a flexible pipeline that integrates DeepPavlov, Falcon, SpaCy, qualifiers constraints, and reverse lookups. Results. Our experiments reveal that multi-tool entity recognition outperforms single-tool methods, while property extraction benefits from extended property sets and refined ranking strategies. Overall SPARQL correctness reaches up to 70–80% in mid-complex queries but remains lower in domain-specific subsets. Conclusion. The proposed synergy of NLP tools and refined dummy templates increases coverage for complex KBQA, though further improvements in morphological handling and specialized embeddings may be needed to address challenging multi-hop or niche queries comprehensively.

A detecção e a previsão precisas de eventos são fundamentais para sistemas de alerta precoce, visando prevenir desastres e mitigar riscos. Este estudo propõe um framework que combina integração de dados de sensores baseada em ontologia com aprendizado profundo para aprimorar a previsão de eventos críticos. Uma extensão da ontologia Semantic Sensor Network (SSN) é utilizada para padronizar dados heterogêneos de sensores e enriquecê-los com informações semânticas e contextuais, que então alimentam uma rede Long Short-Term Memory (LSTM) para realizar a previsão. Um estudo de caso com dados reais hidrométricos e pluviométricos demonstrou que essa abordagem combinada melhora a acurácia das previsões e possibilita a detecção antecipada de eventos críticos em comparação com análises isoladas de sensores.

Intelligent environments are complex interaction spaces between people, sensors, devices, and systems. Software Engineering requires specific techniques to deal with the development of these systems, addressing intrinsic characteristics of devices and sensors, complex interactions in intelligent environments, quality aspects, scalability and interoperability, and the ability to consolidate good development practices. Using computational resources can help support the construction of contemporary IoT systems, dealing with the complexities and quality requirements presented. The main objective of this work is to present the results of a systematic mapping considering the allocation of resources in an intelligent e-health environment with the help of artificial intelligence. With this, it was possible to observe how data collection, environmental monitoring, prediction of the use of computational resources, and environmental management optimization can occur.

Context: In public institutions, there is great concern about the average number of students graduating at the end of undergraduate studies, which substantially impacts public higher education policies and public investments in Brazilian universities. Moreover, dropout imposes a financial and human burden, preventing students from learning. Problem: Brazil witnessed a university dropout rate of almost 55%. This problem affects society in general, which needs more suitably qualified professionals to face the challenges of the job market. Solution: This work aims to analyze, through AI explainability algorithms, the specific factors that lead to student dropout, considering specific courses from the great areas of science. We strive to explore the profile of students who have dropped out in recent years, stratified by course. Explainability algorithms allow the formal inspection of each factor that led to the dropout. Method: We used the Design Science Research methodology to conduct our study. An analysis with data from a specific university, considering the GDPR, was conducted to verify the proposal’s feasibility. Results: Our results show that the solution can help identify key factors that lead to dropping out, stratified by areas, helping to provide specific actions to deal with this problem in universities.

Context: The Brazilian Information Systems (IS) community has grown in complexity involving different research topics. Characterizing the IS community research topics is vital to bringing knowledge to advance IS research maturity. Problem: The complexity, diversity, and plurality of research topics in IS present a challenging task involving collecting adequate data, analyzing, and exploring different techniques to understand the evolution in the area. Solution: Using research data published by the Brazilian Symposium on Information Systems (Simpósio Brasileiro de Sistemas de Informação - SBSI), we provide the NetO+ ontology to enable automated reasoning about SBSI research topics characterization. IS Theory: This study uses the language action perspective, highlighting what people do while communicating and how they create scientific reality using language. We focus on the SBSI research topics with automatic characterization using the Large Language Model (LLM) and ontology techniques. Method: We analyzed the technological research topics employing semantic analysis using an LLM to identify research topics with keywords related to articles’ titles. Also, the ACM taxonomy CCS Concepts in the IS branch were allied to ontological inferences. Summarization of Results: The findings with LLM to extract keywords from articles, and the ontology inference allows us to find research topics. This approach provides an overview of SBSI research, highlighting the topics’ evolution and their interconnections with ACM taxonomy. Contributions and Impacts on the IS area: This work provides IS research topics to bring knowledge enabling decision-making for sustainable community growth based on the socio-technical approach concerning excellence in IS conception, development, and maintainability.

Electronic medical record systems are essential for the functioning of healthcare institutions today. However, the software that runs these systems often has availability and integrity failures with patients’ personal data. To solve this problem, this paper presents an architectural proposal for decentralized electronic medical record systems that integrate users’ medical and personal records. A provenance mechanism was also introduced to ensure data tracing and auditing. A prototype was developed to evaluate the solution in study cases and verify its functionality to demonstrate its feasibility. The results show that the solution is capable of operating the functions required.

Global concerns about the impact of agriculture, particularly regarding methane emissions from enteric fermentation in livestock, highlight the need for more effective strategies to mitigate these emissions. The CarbonSECO platform, designed to generate carbon credits in Brazilian rural areas, emerged in response to the growing importance of carbon credits to offset greenhouse gas (GHG) emissions. However, additional solutions are required to comprehensively address GHG emissions in the agricultural sector. This article extends the CarbonSECO platform, aiming to quantify, monitor, and control carbon emissions from enteric fermentation in livestock. Through ontologies and machine learning techniques, the platform provides solutions to assess and manage the environmental impact of livestock farming. These advancements directly address mitigating carbon emissions in Brazilian dairy farming. A case study involving 25 monitored farms demonstrates the platform’s ability to predict future emissions and milk production, offering decision-making tools for sustainable agricultural management. The results underscore the effectiveness of these services in promoting environmentally sustainable practices in livestock farming.

Global warming and climate change have been subjects of great interest in recent years. They are understood to be related to greenhouse gas (GHG) emissions. Although agriculture suffers the consequences of these changes, it is one of the top global emitters of GHG. While it is complex in environmental, social, and economic aspects, there is a need to advance solutions for more sustainable agriculture. In the farm environment, an important step is the generation of GHG inventories. A GHG inventory is a systematic process for measuring and recording gas emissions and sequestration. However, generating inventories on rural properties presents many challenges due to the many variables involved, such as land use, animal husbandry, use of electricity, and fuels. One of the main challenges is to deal with this data heterogeneity. The data landscape presents obstacles such as managing source diversity, handling massive data volumes, adapting to various data formats, and ensuring real-time integration for decision-making. In this complex data landscape, ontologies emerge as a potential solution. An ontology is defined as a formal representation of a set of concepts within a domain and the relationships between those concepts. This study proposes an ontological model called CarbOnto for the syntactic and semantic integration of heterogeneous databases. Using an ontology, we intend to contribute to the standardization and interpretation of domain concepts and the addition of semantic information to generate complete GHG inventories. CarboOnto provides the means to generate farm inventories, identify imbalances, and search for solutions to neutralize gas emissions. We report a Case Study and argue that using this ontology can support balancing these gases.

Student dropout from higher education is still a challenge, imposing a financial and human burden and refusing students to learn. Brazil witnessed a university dropout rate of almost 55%. This work aims to analyze the factors that lead to student dropout from Information System courses, exploring the profile of students, using intelligent techniques. The information obtained can help reduce the evasion rate and identify key actions to control the problem. We used the Design Science Research methodology to conduct our study. An analysis with data from a university, considering the LGPD was conducted to verify the proposal's feasibility. Our results show that the solution can help identify key factors that lead to dropping out.

This study focuses on event detection and prediction using long short-term memory (LSTM) algorithms implemented in ontology-based sensor data integration. By integrating data from various sources, we facilitate the creation of semantically enriched and contextually integrated data, thereby enhancing the capability for more holistic predictions. The study introduces a framework that leverages ontology models for data integration, fostering the development of semantic context within a sensor network. A feasibility study, conducted with actual sensor data obtained from hydrometric and hydrological stations, underscores the framework's proficiency in abstracting the data integration process, constructing context, correlating events, and predicting their occurrences. The feasibility study results demonstrate the framework's effectiveness and highlight the potential of combining ontologies and artificial intelligence to enhance data interpretation.

Finding software developers with expertise in specific technologies that align with industry domains is an increasingly critical requirement. However, due to the ever-changing nature of the technology industry, locating these professionals has become a significant challenge for companies and institutions. This research presents a comprehensive overview of studies exploring suitable recommendation systems that can assist companies in addressing this pressing need. To conduct this study, we employ a hybrid systematic mapping approach with an initial number of 1251 studies and a final selection of 21 studies. Our work focuses on collecting data on key technologies, methodologies, and data sets utilized in proposed recommendation systems, to design a new recommendation system that can effectively identify specialists capable of aligning specific technical knowledge with industry domains. The outcomes of this study include insights into the current research trends in this field, alongside a practical overview of considerations necessary for developing a recommendation system that successfully meets the criteria for aligning technical skills with industry domains. By following a hybrid Systematic Mapping methodology and presenting the outcomes in the form of insights, this research addresses the challenge of finding software developers with domain-specific expertise in a rapidly changing technology industry, laying the groundwork for aligning technical skills with industry domains.

Diseases caused by slow and progressive damage are the leading cause of mortality. The stress experienced throughout the day can cause many illnesses, as it is responsible for diminishing the body's defenses. In such cases, prevention is a fundamental component achieved through monitoring individuals, usually through a process that heavily depends on human intervention. Therefore, developing solutions capable of automated monitoring becomes necessary to assist individuals in their daily lives. This work aims to promote individuals' health by monitoring them through smart wearable devices and providing notifications that enable them to learn more about themselves. The work focuses on developing a computational environment composed of wearable devices and an application integrated with a machine-learning model. This model predicts the user's heart rate data and generates notifications accordingly. The results show that real-time user monitoring is possible, and moments of stress can be identified using machine learning, leading to generating notifications.