You might be wondering why we should filter our network graphs in the first place. After all, it is possible to view entire networks using Gephi, and even to further navigate the same graphs using interactive tools. That's all well and good, but if you have been exposed to one of the hundreds (thousands?) of examples of overly dense, hairball-like graphs with no clear meaning, I believe you will quickly come to appreciate Gephi's ability to reduce visual complexity through the use of filters.

One of the beauties of filters lies in their ability to reduce complexity while not affecting the underlying data structure. All we are doing when we filter is removing unnecessary clutter from various stages of our exploration, including the final output. Filtering permits us to reduce the amount of guesswork and speculation about patterns and relationships by removing peripheral elements, thus permitting us to focus on the pieces that matter the most to our story. Think of a complex...