PMPedro Olmo Stancioli Vaz de Melo

Pedro Olmo Stancioli Vaz de Melo

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Pedro Olmo Stancioli Vaz de Melo has a total of 223 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Top Cited Papers

    A reporting checklist for large language models in behavioural science

    by Stefan Feuerriegel, Christopher Barrie, M. J. Crockett, Laura K. Globig, Killian L. McLoughlin, Dan-Mircea Mirea, Arthur Spirling, Diyi Yang, Tim Althoff, Maria Antoniak, et al.


  • Open Access

    Identifying Potentially Irregular Electoral Ads in Facebook during the Brazilian Elections

    by Marcio Inacio da Silva, Lucas de Oliveira, Pedro Olmo Vaz de Melo, Oana Goga, Fabrício Benevenuto

    The 2016 United States presidential election was marked by the abuse of targeted advertising on Facebook. Concerned with the risk of the same kind of abuse to happen in the 2018 Brazilian elections, we designed and deployed an independent auditing system to monitor political ads on Meta in Brazil. To do that we first adapted a browser plugin to gather ads from the timeline of volunteers using Facebook. We managed to convince more than 2,000 volunteers to help our project and install our tool. Then, we use a Convolution Neural Network (CNN) to detect political Meta ads using word embeddings. To evaluate our approach, we manually label a data collection of 10k ads as political or non-political and then we provide an in-depth evaluation of proposed approach for identifying political ads by comparing it with classic supervised machine learning methods. Finally, we deployed a real system that shows the ads identified as related to politics during the 2018 National Brazilian elections. We also investigated early electoral advertisement before the 2020 local Brazilian elections using our model on unsponsored content (regular posts in groups and pages). We noticed that not all political ads we detected were present in the Meta Ad Library for political ads on 2018. Additionally, we found possible early electoral advertisements in 2020, which is forbidden in Brazil. Our results emphasize the importance of enforcement mechanisms for declaring political ads and the need for independent auditing platforms.

  • Distance Transform Loss: Boundary-aware segmentation of seismic data

    by Rafael Henrique Vareto, Ricardo Szczerbacki, Luiz A. Lima, Pedro O.S. Vaz-De-Melo, William Robson Schwartz


  • An interpretable and probabilistic competitive balance metric for sports leagues

    by Estéfano Bortolini Vassoler, Pedro O. S. Vaz-De-Melo

    We propose a new probabilistic and interpretable approach to quantify competitive balance in sports leagues based on the precise moment when a tournament can no longer be perceived as perfectly balanced: the longer it takes, the more balanced it is. We analyzed 1,539 seasons from 175 sports leagues in basketball, soccer, handball, and volleyball and observed that only 5 % of the seasons could be seen as perfectly balanced throughout their entire duration. For the others, there is a turning point round after which the points distribution permanently diverges from a range of behaviors likely to occur in perfectly balanced tournaments. Our initial results agree with the general literature since soccer has a considerably more random behavior, i.e., its perceived balance is higher overall. Given the explicit temporal dependence, we also proposed a modification to remove the bias that the order of matches could impose, thereby enabling fair comparisons across different leagues and sports. Our modified coefficient highlights the substantial imbalance in heavily debated leagues, such as the Premier League and the NBA. Furthermore, combining both metrics enables the discovery of anomalous tournament schedules where even simple random changes to the match order could have improved the perceived competitive balance.


  • Open Access

    Uso de antirretrovirais em gestantes vivendo com HIV no Brasil: Coorte retrospectiva, 2014-2019

    by Solange Cristina Teixeira de Lima, Maria das Graças Braga, Jullye Campos Mendes, Matheus Marchesotti Dutra Ferraz, Pedro Olmo Stancioli Vaz de Melo, Micheline Rosa Silveira

    No Brasil, o HIV em mulheres ocorre principalmente na fase reprodutiva, com aumento das taxas de detecção entre gestantes. Alterações nos protocolos de terapia antirretroviral (TARV) têm influenciado o manejo das gestantes vivendo com HIV (GVHIV). Objetivo: Avaliar o uso de antirretrovirais (ARVs) e o perfil sociodemográfico das GVHIV no Brasil entre 2014 e 2019. Métodos: Estudo de coorte retrospectivo com dados do Sistema de Controle Logístico de Medicamentos (SICLOM). Foram incluídas mulheres com 14 anos ou mais e ao menos uma dispensação de ARV durante a gestação. As gestantes foram classificadas entre usuárias prévias de TARV e iniciantes na gestação. Análises descritivas e analíticas foram realizadas. Resultados: Incluíram-se 34.659 gestantes: 53,3% usaram TARV previamente e 46,7% iniciaram na gestação. A maioria tinha 30-39 anos (32,6%), era não branca (46,7%), sem parceiro (38,5%) e com 8-11 anos de estudo (27,7%). O tempo médio de TARV foi de 174,2 dias, com 14,3% relatando reações adversas. Os esquemas de ARVs variaram: lopinavir + ritonavir/zidovudina + lamivudina até 2016, tenofovir + lamivudina + efavirenz em 2017, e tenofovir + lamivudina/raltegravir a partir de 2018. Conclusão: O uso de TARV entre GVHIV seguiu os protocolos nacionais, refletindo práticas consistentes. Contudo, os dados destacam vulnerabilidades sociais e a necessidade de ampliar o acesso à saúde. Melhorar os serviços de saúde reprodutiva é essencial para reduzir a transmissão vertical e garantir cuidado integral.


  • Non-adherence, discontinuation, and seroconversion among people on daily HIV pre-exposure prophylaxis (PrEP) in Brazil: a nationwide cohort study

    by Kennedy Crepalde-Ribeiro, Maria das Graças Braga, Micheline Rosa Silveira, Alexandre Sampaio Moura, Pedro O.S. Vaz-De-Melo, Matheus Marchesotti Dutra Ferraz, Sallie-Anne Pearson, Juliana de Oliveira Costa

    The goal of this study was to evaluate characteristics associated with Pre-exposure Prophylaxis for HIV infection (PrEP) non-adherence or discontinuation in Brazil and assess the association between these outcomes and HIV seroconversion. We used linked national dispensing and pathology data to identify people aged 14+ years initiating PrEP in 2018. We estimated non-adherence using the proportion of days covered (PDC), defining non-adherence as PDC < 60%. We defined discontinuation as a gap in therapy of at least 120 days. We used logistic regression models to assess characteristics associated with non-adherence or discontinuation within 365 days and to assess HIV seroconversion and its association with non-adherence or discontinuation. In 2018, 5100 people initiated PrEP; 34.4% (95%CI 33.1%; 35.7%) were non-adherent and 34.3% (95%CI 33.0%; 35.6%) discontinued. People aged 34 years or less, women, people with lower education levels, or living in the Northern region of Brazil were more likely to be non-adherent or discontinue therapy. Thirty people (0.6% 95%CI 0.4%; 0.8%) seroconverted. HIV seroconversion was associated with PrEP non-adherence or discontinuation (OR = 6.2 for both). Non-adherence and discontinuation were common in people initiating PrEP and strongly associated with HIV seroconversion. Our work identifies populations at higher risk of non-adherence or discontinuation which could be targeted in interventions to reduce new HIV cases in Brazil.


  • Open Access

    Spatial disease anomaly detection under sparse data with stability metrics

    by Raphaella Carvalho Diniz, Renato Assuncao, Pedro Olmo Vaz de Melo

    The stability concept is a widely employed criterion for determining the optimal number of clusters in non-spatial datasets. This principle posits that an algorithm should successfully identify similar clusters across various perturbed instances of the data. In this study, we extend the application of stability concepts to spatial anomaly detection algorithms in the context of disease rate data. Our approach involves introducing novel methods for introducing noise to a singular instance of a disease map while preserving its spatial pattern, even in the absence of knowledge about the underlying distribution. Additionally, we present innovative metrics to assess the stability of spatial anomaly detection algorithms. Furthermore, we demonstrate the utility of the stability concept in analyzing real-world datasets. This involves aiding in the selection of the most suitable spatial anomaly detection algorithm for the given data, pinpointing the epicenter of the spatial anomaly, and quantifying the confidence level of the method for each area on the map. Importantly, our proposal is agnostic to the spatial configuration of the map, the underlying distribution of the data, the chosen spatial anomaly detection algorithm, or any prior information.


  • On Representation Learning-based Methods for Effective, Efficient, and Scalable Code Retrieval

    by Celso França, Rennan C. Lima, Claudio Andrade, Washington Cunha, Pedro O.S. Vaz de Melo, Berthier Ribeiro-Neto, Leonardo Rocha, Rodrygo L.T. Santos, Adriana Silvina Pagano, Marcos André Gonçalves


  • Open Access

    Unraveling the Dynamics of Stable and Curious Audiences in Web Systems

    by Rodrigo Alves, Antoine Ledent, Renato Assunção, Pedro Vaz-De-Melo, Marius Kloft