Visualizing the Hyper-Communities of the Bluefin Tuna Controversy using Gephi
Figure One: Behold the aggregate mass!
To begin visualizing the Hyper-Communities that surround the Bluefin Tuna controversy, we used NaviCrawler and Gephi to visualize how websites involved with the Bluefin Tuna controversy connect with one another.
Methodology:
We crawled fifty four In-Sites using NaviCrawler. These crawls consisted of a Max. Depth of 1, a Crawl Distance of 0, a Tabs Count of 3, and Timing of 32 m/s. Before crawling, we tagged the In-Sites as "NGOs," to represent sites for larger, non-governmental organizations; "Production," to represent sites for organizations or companies focused on tuna production; "Government," to represent sites for government entities; and "Blog," to represent smaller, privately operated sites. The In-site total was comprised of 20 NGOs, 7 Production, 19 Government, and 8 Blog sites.
Crawling through the In-Sites produced "Next-Sites," or those sites connected to In-Sites. This data was exported through a .gdf file, and opened with Gephi visualization software. The data-set was assessed and superfluous nodes (those without any relevance to the Bluefin Tuna controversy) deleted. The data-set, after accounting for superfluousness, comprised 1590 nodes and 2177 edges. These nodes were then ranked via size according to their degree of connectivity and the network represented with a variety of algorithms to analyze and assess the hyper communities surrounding and within the Bluefin Tuna controversy.
Results:
Figure One represents the initial visualization of all In-Sites and related Next-Sites.
Figure Two represents the same data-set as visualized through Gephi's Force Atlas 2 algorithm that displays data as based on its complementarity, and repulses nodes from the center based on their connectivity.
Methodology:
We crawled fifty four In-Sites using NaviCrawler. These crawls consisted of a Max. Depth of 1, a Crawl Distance of 0, a Tabs Count of 3, and Timing of 32 m/s. Before crawling, we tagged the In-Sites as "NGOs," to represent sites for larger, non-governmental organizations; "Production," to represent sites for organizations or companies focused on tuna production; "Government," to represent sites for government entities; and "Blog," to represent smaller, privately operated sites. The In-site total was comprised of 20 NGOs, 7 Production, 19 Government, and 8 Blog sites.
Crawling through the In-Sites produced "Next-Sites," or those sites connected to In-Sites. This data was exported through a .gdf file, and opened with Gephi visualization software. The data-set was assessed and superfluous nodes (those without any relevance to the Bluefin Tuna controversy) deleted. The data-set, after accounting for superfluousness, comprised 1590 nodes and 2177 edges. These nodes were then ranked via size according to their degree of connectivity and the network represented with a variety of algorithms to analyze and assess the hyper communities surrounding and within the Bluefin Tuna controversy.
Results:
Figure One represents the initial visualization of all In-Sites and related Next-Sites.
Figure Two represents the same data-set as visualized through Gephi's Force Atlas 2 algorithm that displays data as based on its complementarity, and repulses nodes from the center based on their connectivity.
Figure Three represents the hyper-community as visualized through Force Atlas 2 and partitioned to display NGO vs non-NGO nodes.
Figure Four represents the hyper-community as visualized through Force Atlas 2 and partitioned to display Production vs non-Production nodes.
Figure Five represents the hyper-community as visualized through Force Atlas 2 and partitioned to display Government vs non-Government nodes.
Figure Six represents the hyper-community as visualized through Force Atlas 2 and partitioned to display Blog vs non-Blog nodes.
Discussion
Within the analyzed sample, NGO sites had the greatest complementarily, while Blog sites had the least . However, Blog sites had the greatest Weighted Degree. This would lead one to believe that, given the sample, NGOs provide most information--as they are, relatively, the most linked-to and linked-from--but that Blog sites provide more opportunities for connection outside the sample, as they have the greatest number of hyperlinks. Government sites appear to have the second most complementarity within the sample, demonstrating a high linkage to and from sample sites. Also, it is interesting to note that the few Production sites that were included are all within the network, instead of appearing on the periphery.
Conclusions
This sample seems to indicate that NGOs are the most significant actors in terms of connectivity and complementarity. Blog sites had a high level of connectivity, but not as high a level of complementarity. Government sites resembled the pattern of NGO sites, but were not as connected or complementary. Finally, Production sites had higher complementarity, but lower connectivity in this hyper-community.
While fifty four In-sites provided a large data set, much remains to improve this visualization in terms of scope and accuracy. First, the sample size needs to be continually updated and expanded, especially in terms of Production and Blog sites, so as to better represent the contemporary controversy. Also, the data set needs to be further filtered for superfluous sites or repetitions (i.e. a variety of links to different Twitter posts and accounts). In order to better represent the hyper-community surrounding the Bluefin Tuna controversy, the data should be represented through variety of Gephi algorithms, node ranking, and partitioning. The power of this visualization is the way in which it catalyzes critical thought and provides a potential tool for analysis; this capability can be improved through continued and varied use.
Within the analyzed sample, NGO sites had the greatest complementarily, while Blog sites had the least . However, Blog sites had the greatest Weighted Degree. This would lead one to believe that, given the sample, NGOs provide most information--as they are, relatively, the most linked-to and linked-from--but that Blog sites provide more opportunities for connection outside the sample, as they have the greatest number of hyperlinks. Government sites appear to have the second most complementarity within the sample, demonstrating a high linkage to and from sample sites. Also, it is interesting to note that the few Production sites that were included are all within the network, instead of appearing on the periphery.
Conclusions
This sample seems to indicate that NGOs are the most significant actors in terms of connectivity and complementarity. Blog sites had a high level of connectivity, but not as high a level of complementarity. Government sites resembled the pattern of NGO sites, but were not as connected or complementary. Finally, Production sites had higher complementarity, but lower connectivity in this hyper-community.
While fifty four In-sites provided a large data set, much remains to improve this visualization in terms of scope and accuracy. First, the sample size needs to be continually updated and expanded, especially in terms of Production and Blog sites, so as to better represent the contemporary controversy. Also, the data set needs to be further filtered for superfluous sites or repetitions (i.e. a variety of links to different Twitter posts and accounts). In order to better represent the hyper-community surrounding the Bluefin Tuna controversy, the data should be represented through variety of Gephi algorithms, node ranking, and partitioning. The power of this visualization is the way in which it catalyzes critical thought and provides a potential tool for analysis; this capability can be improved through continued and varied use.