Posts tagged Ant Colony

In a previous article, I have compared social media to ant colonies: people rate things and the cumulative rating is an indication of the importance of that thing compared to other things. Well, this is not all true. There is one big difference between ant colonies and ‘human colonies’: ants don’t do politics.

By politics, I mean that an ant’s actions are not driven by self-interest. The ant merely executes things that it’s programmed to do and it’s not even conscious about it. The picture is quiet different for humans who are well known to be thrived by self-interest. People tend to post things and vote for things for which they have something to gain. I would go further and say that some use social psychology concepts such as the ‘group effect’ to catalyze interest into something they are promoting. Add to this the fact that a small proportion of users create content on social media and it becomes easy to see that there is great bias on information found in social media.

When analyzing this phenomenon, we realize that it is a perversion from the initial intent of social media, which was a set of services build for the well-being of a community. It seems that it has become a set of services built for the well-being of certain community members. In this new reality, it is more important than ever for the consumer to be aware and critical to the information the is provided on social media.

Ant colony optimization is a swarm intelligence heuristics method used in optimizing complex problems for which it is not possible to investigate all possible solutions in timely fashion. The traveling salesman’s problem is an example of this kind of problems. In this problem, a salesman’s objective is to find the shorted path that will connect a set of cities that he has to visit. As the number of cities grows, it becomes increasingly time-consuming to determine the best path because the number of possible paths (a combination of all the different paths) grows in a polynomial way.

Heuristics are used to find a partial solutions to this kind of problems. By partial solutions, I mean a solution that is not the best but one that is acceptable enough given a time constrain that we have to make a decision. Ant colony optimization is an heuristic method which is inspired by ant colonies: while looking for food, ants lay down pheromone (a chemical signal that triggers a natural response in another member of the specie) on the path that they take randomly. As a “convention” between all ants in the colony, the path that has the strongest pheromone signal is the one that has to be prioritized for the next ant to take. In fact, more pheromone means that more ants have taken that path and that more food is likely to be found from that path. Social media exhibit similar characteristics and I believe they operate in the same way as ant colonies.

All social media have an “object of interest by the community” that is analogical to the food in ant colonies. On Facebook, for example, People are the object of interest. Professionals are the object of interest on LinkedIn. Links to other sites are the object of interest on Digg or Reddit. Like ant colonies, the purpose of the social networks’ community is to find the most valuable “object of interest”. Similarly to the pheromones in ant colonies, members of a social network vote on objects to let others know that it might be interesting to the rest of the community. The more votes an object gets, the more valuable it becomes to the community. Using the ant colony metaphor, the objects that get a lot of votes are the path that the rest of the colony should be following. On Facebook, the number of friends are the pheromone; those people are the “cool” people. On linked in, it is the number of connections or recommendations; those professionals are the “competent” ones. On Digg, it is the number of “diggs” voted on the a certain link; those websites have the most “interesting” content.