3.6. Keywords and Network Analysis
Constructing the keywords co-occurrence is an informative tool that recognizes the research composition. This method refers to the use of correlation measures between words, revealing their appearance relationships. The co-occurrence network highlights the link between keywords, forming a network map.
To investigate related keywords to food security and price dynamics, we present the first ranked 10 keywords based on their number of associations and the intensity level of their relationship. The first keyword was “food security”.
Figure 5 identifies the keyword network. The color represents the cluster and links between rectangles show the strength level of the relationship between these keywords; a shorter distance signifies more intense association.
As highlighted in
Figure 5, food security occupies the network hub, presented by the biggest rectangle. The keyword “Agriculture” is superposed on “food security”, reflecting the closeness between them, in other words, we cannot separate agriculture from food security.
The research’s main focus was placed on socio-economic and environmental aspects of food security such as climate change, health, poverty, price, and nutrition.
Following
Figure 5,
Table 8 identifies the first ranked keywords based on their co-occurrence. Food security is in first place with 77 co-occurrences, followed by the word security. Agriculture has 55 co-occurrences. The keyword price is among the first ten words, ordered by their appearance, with 44 co-occurrences. Based on these findings, food demand and consumption, income, health, and nutrition are emerging keywords in research studies. This highlights the importance of exploring the relationship between food price variability, available income, food nutritional value, and food security.
The dendrogram in
Figure 6 displays the hierarchical rank and the connection between the ordered keyword groups. Multiple Correspondence Analysis (MCA) has been employed. It transforms a considerable number of variables into small structural groups of similar keywords. It helps to investigate and interpret the different identified clusters. It does not intend to find the perfect level of associations between clusters, but it aims to estimate the approximate number of clusters for further discussion.
The dendrogram of keywords approves what is presented in the keyword co-occurrence. Two clusters are identified. For the blue group, food security is linked to production and policies. For the red cluster, the word insecurity is related to quality, health, and income. Future research focuses on the production, household income, and access to healthy food, which affords them the necessary nutritive needed value to protect their health status.
As highlighted in
Figure 7, the food security field is noticeably increasing over the years. This means that it is always the core interest of many researchers. The food security composite keywords are joined with the word "price" and "impact", which also rise with an important speed. This signifies the growing interest in research in this field.
In parallel with the decreased rate of under nourished people, food security research has followed an exponential movement. It confirms that research considers food security a crucial issue for the whole world. The keyword map displays that food security presents a complicated issue related to different fields, such as human health and nutrition and climate (environment).
Figure 8 shows that agriculture, food security, and climate change are always the core of the analyzed research papers. We notice as well that "consumption" and “price” are present with high importance. From this point, a relationship has been exposed in previous research using these keywords in the same research topic. A relationship could be established linking agriculture, security, food, price, and consumption. This proves the importance of research in food security and its impact and the word "price". The increasing curve of these keywords, "food security", "impact", and "price" confirms the rising research linking these three keywords.
In their seminal work, Muñoz-Leiva et al. (2012) [
46] have analyzed the relative position of different topics according to their relationship within a network of a set of articles. In this way, combining the network and cluster analysis, they have defined four types of topics depicted in a two-dimensional coordinate system. These dimensions are suitable to describe the centrality and density of different topics.
The so-called motor themes are at the center of the research field: they are well developed and have a wide and intense relationship with other topics. The basic or transversal themes appear in the lower right part of the graph and can be considered as the backbones of the given set of publications. They can be characterized by high centrality and low density.
The highly developed but isolated topics can be characterized by high density and low centrality. There are intense academic communications within these topics, but their contribution is relatively marginal to the set in general. The emerging or declining themes can be characterized by low centrality and density.
We have divided the investigation period into two parts for convenience: up to the world economic crisis in 2008 and after this period. We have applied this approach to reflect the changes in economic thinking: the crisis has highlighted the vulnerability of the global economic system [
47], and on the other hand, it has shown the importance of new market players, first of all, China [
48].
The basic themes in the first period (
Figure 9) were the supply of a rapidly developing population with rice and stabilization. Other important basic topics were the set of costs—price relationships, access, and availability. The application of cointegration has been on the border between basic and motor themes. The most important emerging topics in this period were the effects of climate change on hunger, health, and the environment. Interestingly, the studies analyzing the problems in Africa were rather peripheral ones.
In the second period (
Figure 10), practically in the last decade, some general topics (e.g., cost-price analysis, cointegration, land use policy analysis) became standard categories and became basic themes. Such relatively new topics, e.g., economic aspects of bioenergy production, have gained importance. The price distorting effects of subsidies became central topics. In recent years, it became obvious that obesity cannot be considered a local problem anymore. Under these conditions, the economic aspects of obesity became motor themes. Some relatively new topics emerged as new problems, such as price transmission, network analysis, and supply chain management.
Figure 11 exhibits the link between countries, keywords, and sources. The rectangle’s size describes the importance of the word: the bigger the rectangle, the more important the country, keyword, and journal.
This supports what was analyzed in the previous sections. The USA has published more articles on food security. Food security is the most prominent topic that the authors’ keywords have mentioned. The analysis of the top countries, sources, and keywords shows that there were four main countries (i.e., the USA, Australia, and the UK) and four main sources (i.e., food security, food policy, sustainability, and global environmental change-human and policy dimensions) highlighting solid links with the literature’s main research topics (i.e., food security and food prices).
This shows the relations between the countries, the principal authors’ keywords, and sources. This diagram demonstrated which countries treat the food security and food price change of publications, the core author keywords used in these publications, and the sources they have published. As presented in the figure, the USA headed the country list, which focuses on food security as a keyword; their publications are mainly in food security and food policy journals. Climate change was the second important keyword used mainly by American authors.