Pagerank and Extinction: How Google’s algorithm is being used to analyze Ecosystems
Every year, a disturbing amount of species around the globe go extinct, and despite our vast technological progress, there has been very little progress in stopping this extinction epidemic. One strategy has been trying to target species that many other species are dependent on: species which either prey on a multitude of others, or are prey for a variety of others. If a species that has a lot of other species dependent on it goes extinct, then there is a high chance that all of the dependent species will also become extinct. Therefore, it has become critical to try and target these species and focus on preventing their extinction. However, it is extremely difficult to track interactions between species, due to obstacles like ecosystem weather and accessibility, and therefore to identify species on which to focus. Scientists are always trying to get a leg up and find new, innovative solutions to try and track these interactions, and one of those is using computer programs to try and model ecosystems and predator-prey chains.
In order to optimize their programs and get faster and easier to process results, two scientists, Stefano Allesina and Mercedes Pascual, tried searching for solutions in places that weren’t so obvious. One of the places they searched was the world wide web, and one of the solutions they found was the unlikely Google pagerank algorithm. As we discussed in class, Google’s algorithm operates on the principle that the more references to a given page there are, the more important that page is. In order to translate that algorithm to suit their purposes, they chose to represent species as pages. The way they were able to use the algorithm in a similar manner to Google was by increasing the importance of a species if there were a large amount of species that were dependent on it as prey. After this modification, they were able to determine which animals were the most important in the ecosystems they were studying faster and more clearly than they previously had been able to. This shows that, while we only focused on web applications of the algorithm in class, there are an infinite amount of problems that this algorithm, and many more like it, could be applied to.
Resource Used: http://e360.yale.edu/features/global_extinction_rates_why_do_estimates_vary_so_wildly