The Social Network Analysis to Study Discourse …
133
different ideals of how to detect a group, and these techniques are very opposed to
measures the centrality [50].
A traditional way is to reduce group detection to a classification or clustering.
Within these techniques are the traditional k-means, genetic algorithms, modularity
analysis (number of links between groups is small, high within groups), among
others. In addition, these techniques are parametrizable (number of classes in kmeans, minimum modularity, etc.), which allows analyzing the quality of the classification. Here, hierarchical trees can be used to decide when the ranking is good.
Another traditional way to detect a group is to apply graph theory, for example, by
using different colors to classify groups in graphs. In this case, it is also possible to
see the problem as one of structural equivalence transformed into another of regular
equivalence: “in a group of friends, the friends share the same friends” (this defines
an iterative algorithm). More recent forms include the use of measures of centrality:
centrality of proximity, “in one group, all actors are close;” centrality of intermediation, “one group is a network more or less isolated from the rest;” and democratic
criteria, which consist of counterbalancing two or more different group detection
criteria.
Fourth, the visualization of social networks also serves as a method of approach
that allows discovering properties, but has less theoretical weight in the analysis.
Observing complex networks is a big challenge. In general, much information as
posible is visually presented so that it is understandable. There are many ways to
look at the data, and each can illustrate different analytic properties: such as centrality,
communities, key actors, etc. When it comes to detecting communities, there are a
large number of algorithms that allow their visualization, each one obeys a different
idea and purpose. But often the goal is to gain an immediate presentation of the
network and to have an efficient algorithm. In this study, we used SonecTv software.
4 Practical Application of Network Analysis in Nexus
Issues—The Case of Brazilian Biofuels
Brazil is a leader in the export of various agricultural products such as soybeans,
corn, and coffee [31]. However, the country also leads the list of countries with
the largest losses of forests in the world in the last years. In 2019 alone, Brazil
was responsible for more than a third of all global deforestation. The expansion
of the agricultural frontier is the main driver of deforestation in the country and
has compromised springs, bodies of water, and the ability of forests to generate
rain. The deterioration of water resources directly affects the electric energy sector
in Brazil, given that 65% of its energy is generated by hydroelectric plants [14].
Due to the exporter of agricultural commodities conditions, the nexus approach is
needed to explore to the case of Brazil, for example, connections between demands
for resources and the supply of exports. These commodity exportation practices
directed toward supplying external markets, generate pressure on local resources,
133
different ideals of how to detect a group, and these techniques are very opposed to
measures the centrality [50].
A traditional way is to reduce group detection to a classification or clustering.
Within these techniques are the traditional k-means, genetic algorithms, modularity
analysis (number of links between groups is small, high within groups), among
others. In addition, these techniques are parametrizable (number of classes in kmeans, minimum modularity, etc.), which allows analyzing the quality of the classification. Here, hierarchical trees can be used to decide when the ranking is good.
Another traditional way to detect a group is to apply graph theory, for example, by
using different colors to classify groups in graphs. In this case, it is also possible to
see the problem as one of structural equivalence transformed into another of regular
equivalence: “in a group of friends, the friends share the same friends” (this defines
an iterative algorithm). More recent forms include the use of measures of centrality:
centrality of proximity, “in one group, all actors are close;” centrality of intermediation, “one group is a network more or less isolated from the rest;” and democratic
criteria, which consist of counterbalancing two or more different group detection
criteria.
Fourth, the visualization of social networks also serves as a method of approach
that allows discovering properties, but has less theoretical weight in the analysis.
Observing complex networks is a big challenge. In general, much information as
posible is visually presented so that it is understandable. There are many ways to
look at the data, and each can illustrate different analytic properties: such as centrality,
communities, key actors, etc. When it comes to detecting communities, there are a
large number of algorithms that allow their visualization, each one obeys a different
idea and purpose. But often the goal is to gain an immediate presentation of the
network and to have an efficient algorithm. In this study, we used SonecTv software.
4 Practical Application of Network Analysis in Nexus
Issues—The Case of Brazilian Biofuels
Brazil is a leader in the export of various agricultural products such as soybeans,
corn, and coffee [31]. However, the country also leads the list of countries with
the largest losses of forests in the world in the last years. In 2019 alone, Brazil
was responsible for more than a third of all global deforestation. The expansion
of the agricultural frontier is the main driver of deforestation in the country and
has compromised springs, bodies of water, and the ability of forests to generate
rain. The deterioration of water resources directly affects the electric energy sector
in Brazil, given that 65% of its energy is generated by hydroelectric plants [14].
Due to the exporter of agricultural commodities conditions, the nexus approach is
needed to explore to the case of Brazil, for example, connections between demands
for resources and the supply of exports. These commodity exportation practices
directed toward supplying external markets, generate pressure on local resources,