The first thing that came to mind when reading Demystifying Networks was infamously creepy social network LinkedIn. Weingart introduces the basics of networks along with the inevitable challenges that come with them. The blog post is directed, as Weingart notes, to digital humanists. Therefore, the issues are directed at humanist scholars who face the challenge of dealing with data that is “uncertain, open to interpretation, flexible, and not easily definable”.
This is where I began thinking about social networks. Weingart warns of the dangers of using networks to analyze data. First, “networks can be used on any project. Networks should be used on fewer”. Second, “methodology appropriation is dangerous”; scientific approach, as we know, does not always map on neatly to a humanist one. Social networks connect people. I am not sure how nodes and edges work within social networks, but I assume that these are in use for websites’ features like “People You May Know”.
There are many articles online that question LinkedIn’s analysis techniques. For example, David Veldt’s article LinkedIn: The Creepiest Social Network for Interactually.com takes a critical look at some of the site’s functions and features. I don’t personally have a LinkedIn account but know from friends and family that use it that they often see the most random, unexpected people pop up in their LinkedIn “People You May Know” section. Veldt lists a couple examples of his own experience with “People You May Know”. The suggestions it comes up with are often inexplicable – it seems that LinkedIn has no possible way of knowing that this person is your mailman’s cousin! It even sometimes suggests the name of someone you know, but is not actually that person (just the same name).
Veldt attempts to analyze LinkedIn’s established network. Although I am not positive, it is pretty safe to assume that LinkedIn has some system of “edges”, which Weingart defines as descriptive links that connect nodes. I believe this is what Veldt is after – what is LinkedIn using to inform its edges? LinkedIn’s Help Center quotes only two factors that the “People You May Know” section is based on: “Commonalities between you and other members. For example, you may have common connections, similar profile information and experiences, work at the same company or in the same industry, or attend the same school” and “Members you’ve imported from other address books in your Contacts list”. Veldt discovers that there are pre-checked boxes within his account that allow LinkedIn to share his data with third party applications as well as giving information about his site visits to pages that use LinkedIn plugins. However, Facebook (as Veldt suspected) is not listed as one of these plugins. The mystery remains…
I am genuinely interested in how LinkedIn succeeds in such creepiness. This example resonates with Weingart’s opinion on humanist approach to data (as far as I understand how LinkedIn works). Weingart argues, “Unfortunately, given that humanist data are often uncertain and biased to begin with, every arbitrary act of data-cutting has the potential to add further uncertainty and bias to a point where the network no longer provides meaningful results. The ability to cut away just enough data to make the network manageable, but not enough to lose information, is as much an art as it is a science”. Plotting links between human relationships seems so complicated, but LinkedIn somehow masters it to an uncomfortable degree.
http://www.interactually.com/linkedin-creepiest-social-network/