Graph Theory on LinkedIn?

The article I choose for this week LinkedIn hits 200 million members, majority outside U.S. I thought it was very interesting since the entire idea of LinkedIn fast growth and expansion as a social network can be explained through the concepts of graph theory.

The article refers to LinkedIn growth in countries other than the United States, which is the host country, for the application. Yet, countries like India, Colombia, Brazil and the UK are the countries with the greatest amount of LinkedIn profiles. It is always interesting seeing how fast social networks flourish. Graph theory is useful to explain LinkedIn propagation. Therefore, when one person adds a contact the application seeks other people within the network of both contacts and refers contacts of both to one another. No wonder this professional network expansion all around the world.

LinkedIn profiles’ structure is just as graph theories suggest. First, graph theory consists of “strong ties” within nodes, these nodes then form paths which relate to LinkedIn levels of connections. First, LinkedIn has 1st degrees connections who are those closer and immediate relationships or “strong ties”. Second, the “weak ties” are arguably those 2nd degree connections with the least frequent interaction. LinkedIn social connections can also be explained by the strong triadic closure property since the third connection between either a strong tie or a weak tie that does not end in an edge is the least frequent connection between the nodes; likewise are third degree connections as LinkedIn labels them.

Doing this, LinkedIn employs a strategy similar to the Breath-First Search Algorithm to suggest contacts with possible matches that could potentially be known to either party. First, LinkedIn searches for contacts within the same circle whether colleagues or classmates. These in the Breath algorithm have a declared distance of 1. Second, LinkedIn suggest people who are friends of those classmates and colleagues which in the algorithm suggests a distance of 2. Third, LinkedIn classifies those contacts that could potentially be acquaintances since these are not in the same path, instead these contacts share a group and/or have similar backgrounds. In this case the algorithm suggests that those contacts have a distance of 3.

Employing the theory and the models one can test how graph theory allows LinkedIn to grow its network among contacts within the same country and with similar backgrounds. Probably this gives some insight of how has LinkedIn been able to grow and expand so fast overseas as well.

Sources click here.


Until next post!


2 thoughts on “Graph Theory on LinkedIn?”

  1. I really like this post because it investigates how LinkedIn is getting one of the largest business social networks in the world and which patterns are used to do so. It is also interesting to see how the process is correlated to the expansion of Facebook in the world. We also know that at some point if LinkedIn is the largest platform for business social networks on-line, it will have its own dynamic as well as dominate and easily eliminate similar pages. Nevertheless, I know that LinkedIn has not reached Europe to a large extend yet and is almost unknown in countries like Germany, for example.

  2. I like how you broke this down into the Breath-First Search Algorithm. One of my concerns about LinkedIn in the long-run draws on the classification of the third level which are meant to be acquaintances. Whenever I am on LinkedIn I am hesitant to connect to those removed from my direct network, and instead try to use LinkedIn as if it were Facebook. If the best way to connect to employers is through weak links, then is LinkedIn serving it’s purpose?

Comments are closed.