This chapter presents a multidisciplinary solution that considers as evolution of endogenous natural extreme event deforestation the threats of droughts and fires in the Brazilian Amazon region. The data are collected from social media, such as newspapers and magazines, related to the domain of droughts and fires that could trigger and accelerate the process of deforestation in the period from 2015 to 2020. The data science concepts and natural language processing with sentiment analysis are used and generate the degree of threat that each news presents regarding the possibility of deforestation. This threat degree generates an endogenous time series that will be used to predict the threat evolution of occurrence of drought, fire, and deforestation for a future of three months. The time series prediction is performed using machine learning and deep learning with an LSTM neural network. An analysis of the endogenous time series is performed using the statistical tools of mean, variance, standard deviation, asymmetry, and kurtosis.
TopIntroduction
The end of the twentieth century was marked by the advent of the internet and consequently this favored those massive volumes of information, from the most different fields of knowledge, to circulate through the World Wide Web – WWW (Castells, 2003). The disclosure and sharing of this information by society introduced new understandings of how this volume of data could be used to generate value for the most diverse areas of knowledge and thus bring benefits to society. Among the most varied domains of information circulating on the WWW, natural extreme events deserve attention for a more detailed study and understanding of their causes, consequences, and possible prevention.
An extreme event is characterized by a sequence of small events generated by human emotions or some reaction of nature that can evolve into a larger event reaching up to a catastrophic event (Rosa et al., 2019) (Clauset, 2018) (Ibanez et al. 2022). The natural extreme event model considered is deforestation, as it has a great influence on the life of society (Santos et al., 2017). Due to its complexity, a multidisciplinary solution would assist in understanding its evolution and possible prevention of this natural extreme event model.
Based on the context presented, this work proposes a multidisciplinary solution that considers the threats of droughts and fires in the Brazilian Amazon region as the evolution of deforestation. To carry out the threat analysis, used as case studies for the natural extreme event data collected from social media, such as newspapers and magazines. Considered the social media of large national circulation, about the occurrence of droughts, fires, and deforestation in the years 2015, 2016, 2017, 2018, 2019 and 2020. The collection of this information is carried out using Google's web search engine (Google, 2016) which performs a search for news related to the topics addressed about drought, burning, and deforestation threats. Each collected news is stored and grouped considering the increasing order of its publication date (Ibanez et al., 2022).
The news collected from social media are processed using data science and machine learning techniques that allow identifying some nature reaction present in a text document. As per the context of the chapter, the reaction identified in the analyzed news texts is the threat of the natural extreme event addressed. The machine learning technique used for the identification of the threat in the news is sentiment analysis, being applied in the chapter with the Natural Language Processing technique, which performs text analysis (Ibanez et al., 2022). Sentiment analysis makes it possible to identify how similar a text is to a given context, using a base text with words referring to a domain (Bird et al., 2009), in this case, the threat of the extreme event. Thus, for each news story collected, the percentage of threat existing in its text is analyzed and identified (Ibanez et al., 2022).