Loading...
8 results
Search Results
Now showing 1 - 8 of 8
- A simple approach to detect anomalies in microservices-based systems using PyODPublication . Landim, Lauriana Patricia Tavares; Barata, Luís; Lopes, EuricoEase of scale is one of the defining characteristics of microservices. However, with scalability comes the problem of diversity of services, making it very important to detect anomalies the soonest possible. Because it is recent, there are still few studies on the best approaches to detecting anomalies in microservices. This paper proposes the Python toolkit, PyOD, as an approach for microservice anomaly detection. This toolkit is composed of a set of anomaly detection algorithms, including classical LOF (SIGMOD2000) to the latest ECOD (TKDE2022). To evaluate the approach, we used two of its algorithms, k Nearest Neighbors (kNN) and Histogram-based Outlier Score (HBOS) to detect anomalies such as application bugs, CPU exhausted, and network jam on the TraceRCA dataset. This dataset contains logs from a real microservices system. The preliminary results show that HBOS algorithm performs better than kNN, with Recall and F1-Score of 93% and 89%, respectively, while for kNN these metrics were 92% and 85%, respectively.
- Desenvolvimento ágil com o método SCRUMPublication . Silva, Eduardo; Batista, Pedro; Barata, LuísEste trabalho de investigação tem como objetivo aprofundar um método de desenvolvimento ágil muito utilizado nos dias de hoje, o SCRUM. O mais relevante neste artigo é saber o que é realmente o SCRUM e os seus valores de uma forma simples de entendimento. O SCRUM é uma estrutura que gere e desenvolve a entrega de um produto. O SCRUM é bastante útil no desenvolvimento onde os sistemas são mais complexos e incomuns. Após este trabalho de investigação, conclui-se que o SCRUM é um método eficiente de desenvolvimento ágil, que ajuda a gerir o desenvolvimento de projetos de forma mais eficiente.
- A informática ao serviço dos mais velhos: uma teoria emergentePublication . Barata, LuísEste trabalho analisa as motivações que levam os mais velhos a aprenderem informática, procurando descortinar quais os interesses subjacentes. Dessa forma procura-se compreender a forma como os mais velhos vêm a informática e o uso dos computadores, como a forma de integração na sociedade. Pretende-se ainda investigar quais as modificações ao nível das estruturas relacionais, nomeadamente na forma como o computador auxilia e aproxima as redes sociais do idoso e a forma como a informática poderá facilitar e ajudar, no processo de envelhecimento. A investigação centrou-se sobre a vivência de aprendizagem dos alunos da Universidade Sénior Albicastrense, recorrendo a técnicas de observação e entrevista, os dados recolhidos foram analisados à luz da metodologia Grounded Theory, usando para isso a codificação por conceitos-chave. A primeira conclusão é o facto de se verificar uma nítida divisão de grupos, fruto dos interesses individuais, nomeadamente: • Aqueles que pretendem adquirir e/ou incrementar a sua base de conhecimentos, ao nível das tecnologias da informação e simultaneamente desmistificar o seu uso. • Os que pretendem uma maior integração, junto das diferentes redes sociais, nomeadamente ao nível familiar, verificando-se ainda uma inversão de papéis, em que as gerações mais novas são a base do conhecimento. • Aqueles que pretendem utilizar as tecnologias da informação e comunicação, como forma de aproximação virtual aos diferentes sistemas (família, amigos,…) alargados e dispersos, com recurso à internet • Outra das conclusões a destacar é o facto de a internet ser considerada o principal motivo da aprendizagem, nomeadamente pelo acesso facilitado à informação, a aproximação aos seus e a manutenção das capacidades cognitivas dos mais velhos. • Durante o decorrer do trabalho serão apresentadas uma série de ideias, que possibilitarão a simplificação do processo de integração com a informática.
- Projeto de desenvolvimento de software GESCORR: sistema de gestão documental e processos empresariaisPublication . Barata, LuísA gestão de documentos e processos é essencial ao bom desempenho de uma empresa e um reflexo da sua organização interna. A existência de grandes volumes de informação, tanto oriundos do exterior como internos potencia a perda de documentos, aumenta a dificuldade em aceder rapidamente à informação e à sua localização física e dificulta o controlo dos fluxos de informação. O GesCorr é uma plataforma de gestão documental que permite o seguimento, consulta, organização e arquivo da informação interna e externa da empresa, em formato eletrónico. Este sistema permite concentrar toda a informação de natureza documental trocada com clientes, fornecedores ou cidadãos nos pontos de contacto com a organização, contribuindo fortemente para a melhoria do nível de serviço prestado. Este trabalho pretende apresentar todo o processo de desenvolvimento e implementação da aplicação de Gestão Documental desenvolvida pelo autor no âmbito da sua atividade profissional na BEIRANET – Soluções Informáticas, Lda e posteriormente na BEIRANET II – Serviços Empresariais, Lda (spin-off da BEIRANET). Mostra-se o processo de modelação do software, o processo de modelação de dados e mostrar-se-ão alguns dos principais módulos e funcionalidades do sistema.
- Predictive maintenance based on log analysis: A systematic reviewPublication . Barata, Luís; Sequeira, Sérgio; Lopes, EuricoIn today’s industries, the Maintenance process of machines and assets implies a significant part of the total operating cost. Many efforts have been made to reduce this cost by optimizing the process and evolving methods that allow information collection on equipment status, avoiding redundant interventions, and predicting the exact moment to perform a maintenance intervention. Using “intelligent” systems that collect data from the operation and remote management systems allows us to gather all the data and apply some methodologies capable of identifying expected behaviors based on past operations. We present a survey of technologies, techniques, and methodologies to give the knowledge background to develop a framework to minimize the occurrence of failures and optimize the process of Predictive Maintenance (PdM) based on the analysis of Log files collected from the various industrial equipment. Generally, these logs contain many records, and many of these records do not directly contribute to evaluating the operation’s machine status. Most of the studies included in this survey use machine learning techniques and focus a significant part of their research on data preprocessing, uniformization and clarification.
- Big Data analysis applied to the retail sectorPublication . Barata, Luís; Rosa, Gonçalo; Martins, Tiago; Silva, Pedro Tiago Salgueiro da; Eurico, LopesBig Data and Artificial Intelligence are rapidly evolving and transforming the world and e-commerce. To explore the most used techniques for building personalized recommendation systems and inventory management, a systematic review was conducted. It was found that hybrid Machine Learning techniques and the Hadoop system are widely used in these areas, delivering satisfactory results while maintaining scalable and efficient systems.
- Classification of anomalies in microservices using an XGboost-based approach with data balancing and hyperparameter tuningPublication . Barata, Luís; Lopes, Eurico; Inácio, Pedro; Freire, MárioMicroservice architecture has emerged as a leading paradigm for decomposing large monolithic applications into smaller, autonomous services. Although this approach offers many advantages, its complexity, distributed nature, and substantial scale create significant challenges for monitoring and anomaly detection. The vast volume of generated data further exacerbates computational load and detection latency, complicating the identification of anomalies. This study analyses the impact of data balancing and hyperparameter tuning on anomaly classification in microservices and introduces ADMXGB - Anomaly Detection in Microservices using XGBoost, a XGBoost-based framework tailored for anomaly detection in microservices that seamlessly integrates data balancing with hyperparameter tuning. We propose guidelines to determine appropriate threshold values that balance sensitivity with false positives, and show that the framework is model-agnostic, enabling integration with different machine learning algorithms beyond XGBoost. Validation was performed using a four-stage process encompassing preprocessing, training, validation, and testing. ADMXGB demonstrated improvements in both Accuracy and F1-Score, reaching 99.96% in both metrics on the TraceRCA dataset, outperforming the baseline XGBoost method by a margin of 1.46% in Accuracy and 45.62% in F1-Score. Moreover, ADMXGB achieves reductions in execution time (-86.3%) and memory usage (-21.7%), while maintaining an acceptable CPU overhead. These findings highlight the robustness of ADMXGB in delivering high-accuracy classification in a microservice environment.
- Big Data-based recommendation systemsPublication . Barata, Luís; Mata, Diogo; Nunes, Jorge; Rodrigues, LucasIntroduction: With the exponential growth of data, digital platforms increasingly rely on Big Data technologies to personalize user experiences and improve the accuracy of item recommendations. Recommendation systems play a critical role in e-commerce, entertainment, and social media by analyzing user interactions, behaviors, and preferences. However, the complexity of large-scale data processing and the diversity of filtering techniques pose significant challenges to achieving high performance and scalability. Objectives: This study aims to analyze how Big Data technologies are applied in modern recommendation systems, emphasizing their role in enhancing personalization and system performance. The work seeks to identify the main data collection strategies, algorithms, and tools adopted in recent research, as well as to assess how these systems address challenges related to scalability and real-time processing. Methods: A systematic literature review was conducted using the IEEE Xplore database, focusing on articles published between 2015 and 2025. The search targeted studies combining Big Data and recommendation systems within e-commerce contexts. Out of 49 retrieved publications, 10 met the inclusion criteria, and 5 were ultimately selected after applying exclusion filters. Each selected paper was analyzed regarding its objectives, employed algorithms, data sources, and achieved outcomes. Results: The reviewed studies demonstrate a wide variety of approaches and technologies, including Hadoop and Spark frameworks for large-scale data processing, deep learning models such as DeepLearning4j for real-time prediction, and classical data mining algorithms like K-Means, Apriori, and FP-Growth. Hybrid methods combining collaborative and content-based filtering were shown to overcome limitations such as the cold start and data sparsity problems. Scalability was addressed through distributed processing and optimization techniques like network pruning and parallel computation. These systems achieved higher recommendation precision and responsiveness across different e-commerce and media applications. Conclusions: The analysis confirms that Big Data–driven recommendation systems are essential for enhancing user engagement and conversion in digital platforms. By integrating data mining, machine learning, and distributed processing technologies, these systems deliver efficient, adaptive, and context-aware recommendations. Future developments should continue exploring hybrid and deep learning–based approaches to further improve scalability, personalization, and computational performance in increasingly complex digital ecosystems.