AUTHOR=Ribeiro Yuri Geraldo Gomes , Bastos Rodrigo Matta , Silva Beatriz Oliveira , Marchini Silvio , Morais Rafael Batista , Catapani Mariana Labão , Corrêa Pedro Luiz Pizzigatti , da Rocha Ricardo Luís Azevedo , da Silva Ariana Moura , Ferraz Katia Maria Paschoaletto Micchi Barros 

TITLE=Social media data from two iconic Neotropical big cats: can this translate to action?

JOURNAL=Frontiers in Conservation Science

VOLUME=Volume 4 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/conservation-science/articles/10.3389/fcosc.2023.1101531

DOI=10.3389/fcosc.2023.1101531

ISSN=2673-611X

ABSTRACT=There is a gradual increase in studies of social media data usage in biodiversity conservation. Social media data is an underused source of information with the potential to maximize the outcomes of established conservation measures. In this study, we assessed how structured social media data can provide insight into species conservation through a species conservation plan, based on pre-defined actions. We established a framework centered on defined steps that goes from defining social media platforms and species of interest to applying general analysis in data based on data dimensions -3 W's framework (what, when, who), and the public engagement that posts received. The final and most important step in our proposed frameworks is to assess the overlap between social media data outcomes and measures established in conservation plans. In our study, we used the Brazilian National Action Plan (BNAP) for big cats as our model. We extracted posts and metrics about jaguar (Panthera onca) and puma (Puma concolor) from the two social media platforms, Facebook and Twitter. We obtained 159 posts for both jaguar and puma on Facebook (manually) and 23,869 posts for the jaguar and 14,675 posts for the puma on Twitter (through an application user interface). Data were categorized for content and users (only Facebook data) based on analysis of the content obtained and similarities found between posts. We used descriptive statistics for analyzing the metrics extracted for each data dimension (what, when, who, and engagement). We also used algorithms to predict categories in the Twitter data base. Our most important findings were based on the development of a matrix summarizing the overlapping actions and dimensions of the data. Our