Licenciatura em Computação (Sede)

URI permanente desta comunidadehttps://arandu.ufrpe.br/handle/123456789/19

Siglas das Coleções:

APP - Artigo Publicado em Periódico
TAE - Trabalho Apresentado em Evento
TCC - Trabalho de Conclusão de Curso

Navegar

Agora exibindo 1 - 3 de 3

A comprehensive software aging analysis in LLMs-based systems
(2025) Santos, César Henrique Araújo dos; Andrade, Ermeson Carneiro de; http://lattes.cnpq.br/2466077615273972; http://lattes.cnpq.br/9618931332191622
Large language models (LLMs) are increasingly popular in academia and industry due to their wide applicability across various domains. With their rising use in daily tasks, ensuring their reliability is crucial for both specific tasks and broader societal impact. Failures in LLMs can lead to serious consequences such as interruptions in services, disruptions in workflow, and delays in task completion. Despite significant efforts to understand LLMs from different perspectives, there has been a lack of focus on their continuous execution over long periods to identify signs of software aging. In this study, we experimentally investigate software aging in LLM-based systems using Pythia, OPT, and GPT Neo as the LLM models. Through statistical analysis of measurement data, we identify suspicious trends of software aging associated with memory usage under various workloads. These trends are further confirmed by the Mann-Kendall test. Additionally, our process analysis reveals potential suspicious processes that may contribute to memory degradation.
Analisando o Backup-as-a-Service como uma estratégia de recuperação de desastres
(2021-06-02) Queiroz, Ewerton Cleyton Silva de; Andrade, Ermeson Carneiro de; Mendonça Neto, Júlio Rodrigues de; http://lattes.cnpq.br/7849727159222731; http://lattes.cnpq.br/2466077615273972; http://lattes.cnpq.br/1234353605805269
In modern environments, failures in information and communication technology (ICT) systems can have several consequences for a business, like data and revenue loss and customers dissatisfaction. Disaster recovery (DR) solutions, as BackupasaService (BaaS), has been adopted by companies as a way to avoid these problems and assure business continuity. Nevertheless, there are plenty of variables to consider during the adoption of a DR solution. Then, in this work, we present an integrated approach using experiments and models to evaluate a BaaS environment designed for DR. In our analysis, we consider relevant DR metrics like availability, downtime, RTO (Recovery Time Objective), and RPO (Recovery Point Objective). The results demonstrate that once BaaS is applied, the environment availability can vary according to the amount of data needed to be backed up or restored. Furthermore, sensitivity analysis indicates that the time needed to recover the data center and the backup interval are the most important parameter values for metrics like RTO and RPO. The proposed approach can help companies or individuals involved in the decisionmaking process for purchasing a DR solution.
Aprendizagem colaborativa apoiada por computador durante a pandemia do Covid-19: uma revisão sistemática
(2022-10-07) Melo, Daniel Lemos de; Araújo, Carlos Julian Menezes; Andrade, Ermeson Carneiro de; http://lattes.cnpq.br/2466077615273972; http://lattes.cnpq.br/3156174527107999; http://lattes.cnpq.br/4495434877384543
With the social isolation measures caused by the pandemic in 2020, educational institutions had to plan alternatives to ensure remote teaching. Among the various approaches, we cite the Computer Supported Collaborative Learning (CSCL) strategy, which seeks, through computers, to promote collaborative teaching remotely. This study seeks to evaluate how the use of the CSCL was during this period of isolation, through a systematic review. From the query in several databases, it was possible to identify 120 works that reported experiences using CSCL. The results indicate considerable success in adopting the CSCL strategy, as well as trends in technology use and approaches in each region of the planet.