Desire Paths & Triadic Closure
Recently I was reading about a concept called desire paths, defined by Wikipedia as a “path created as a consequence of erosion caused by human or animal foot traffic.” Essentially, these unpaved paths between already paved ones reflect an optimization arising from individual choices, and it reminded me greatly of the concept of triadic closure.
Take the picture above for example. This is a common sight on college campuses, where a dirt path will form between points of a previously paved brick path as students walk over a conceived path for convenience. We can consider three nodes, the ends of the tan walkway as well as the middle of the semicircular path, and define them as A, B, and C. The AB and BC edges were already predefined around the tree by the pavement. Thus, the AC edge was more likely to arise because of the nodes’ commonalities with node B.
I was curious to how triadic closure, specifically through desire paths, manifested itself in digital spaces as well. Consider the internet a lawn with some already defined “paths” through links between webpages. Triadic closure works to increase the links between closely related pages, forming desire paths for previously unmet needs.
In fact, search engines such as Google in particular are known from creating desire paths for users, even when it may not be obvious what “nodes” are being linking together through edges. For example, when a user begins typing in a question such as “capital of georgia,” Google suggests the answer Atlanta even before they can hit enter. In this situation, the first node is the user’s question about Georgia’s capital, call this A, another node is the act of clicking enter to look up the question on the search engine, call this B, and the last node is the answer to the question itself, call this C. What Google does is create the AC edge and fulfill triadic closure, providing an answer without needing to cross B to ease a user’s experience when using the product.
It’s really interesting to examine the different applications of networks, from dirt paths to complex search algorithms. Triadic closure itself is a concept with broad repercussions. In any instance, a commonality between two pairs of three nodes greatly increases the change of a new relationship between them forming.
Sources:
https://dotsandspaces.medium.com/desire-paths-in-digital-spaces-7a2aed4f7f47