Twitter and Mobile Recommendations
Twitter’s new mobile update for iOS and Android takes recommendations of “Who to Follow” to a mobile framework. The company recognized the market’s need to continue to develop the mobile format to boost engagement rates amongst Twitter users. One piece that is particularly interesting in Twitter’s algorithms is the aspect of proximity and the resulting increased probability of becoming friends or following someone who is geographically desirable. This is important in instilling this new mobile recommendation format because people are more likely to bring their phones with them when traveling to different locations, and therefore Twitter can take full advantage of their location data with its latest update.
This relates to our class in several ways. One problem we have been working on in class includes choosing the correct node that a network such as Facebook or Twitter should recommend to a user. We perused the question of whether it is more important to recommend an actual friend of the user who is potentially a weak tie, or a part of a cluster of strong ties also containing the user. Similarly, Twitter utilizes an algorithm that incorporates and centralizes other existing networks through syncing users’ contact data. In doing so, they have a strong method to reach the correct contact and bring him or her over to their particular social network through the “Who to Follow” section. This also relates to our guest lecture this past Friday about Facebook and shows how a similar, yet different social network utilizes networks to maximize the user’s experience and network of friends and followers. In the future, it would be interesting to compare Facebook’s and Twitter’s algorithms on a mobile vs. desktop format to see which network recommendations have the most success.
Sources: