Using Network Types to Optimize Outcomes
A network, though in its simplest form is a graph of connected nodes, can come in many different forms. A network can be modular – containing many subgroups, centralized – containing a few central nodes and many peripherals, or any combination and permutation of the two. This article analyzes the two main forms of networks, modular and centralized, and their relation to the spread of infection by observing primate groups.
The article observes that a highly modular network of primates that contains several subgroups connected by weak local bridges tends to limit the spread of infection. Often, the outbreak of an infection is limited within each subgroup and dies before it can spread to the other subgroups in the network. This observation is an opposite parallel to the study of the spread of job opportunities as observed in class. Job opportunities seemed to die down within subgroups and rather be spread between local bridges or acquaintances. This can attribute to the different species of primates that make up sub groups and the different kind of friends that may make up a human’s social network leading to new information.
On the other hand, highly centralized networks are more susceptible to infections once the infection has reached the network. The central nodes, in fact, are referred to as super spreaders as they have the most impact in spreading the infection to peripherals. We can compare this again to the spread of job opportunities, where a person’s highly centralized social network may be his coworkers and bosses for example who may not be able to offer him novel job opportunities.
The conclusion to be made from this article is that living in a highly modular network can offset the danger of living in large network in terms of disease spread. However, if another outcome is desired such as a new job opportunity or novel information, then a highly modular network can improve the chances of information spreading. We can deduce, therefore, that it is imperative to analyze the structure of a network in order to optimize an outcome.
Source: http://dash.harvard.edu/bitstream/handle/1/8715733/Griffin_CommunityStructure.pdf?sequence=1