Skip to main content



Discussion of Facebook Friend Suggestion

https://blog.udemy.com/how-does-facebook-suggest-friends/

This article talked about anecdotes of Facebook suggestion system and analyzed how it works and makes profit.

People might feel weird when suggestion of “unfamiliar” or “dislike” friends popped up. The user Miranda has questioned this. This article discussed a situation that user A stalked B and then B might receive the suggestion of A. Why would this happen? It is up to the Facebook suggestion algorithms.

The article treated Facebook as a savvy data miner due to its clever algorithm. For example, it grasps the information like education and suggests the people with same background, or connects a user to other users through online activities. How to reconnect the guy you want to date? This article gives us the answer.

This friend suggestion system can also be utilized in marketing. Consumers with similar interests are target. Advertises can focus on users whose interests are the most closely to their products. Facebook also suggests pages and posts.

As we learned from the class, the relationship can be either positive, which is friendship, or negative, which represents the enemy. The relation between enemies can also be strong. For example, the competitors in the same field know each other, and have many mutual friends. Even though the probability that they become friends is high, it’s also possible that they don’t like each other because of the competition. Thus, if Facebook suggests the “dislike” friend, user couldn’t feel joyful.

The basic idea of Facebook suggestion is based on the triadic closure. If A and B have common friend, A and B have increased likelihood to be friends. Friend-of-Friend is potential friend.

In class, it refers to the algorithm Facebook used. The prediction of friendship is not only related to mutual friends, time discounted mutual friends and new mutual, but also based on the features between users, like the gender, age and country. Meanwhile, different activities have varied weights. People have same interests and features tend to be friends then without common interests. Talk back to the stalk case. If user A often view B’s profile and B’s interested page, A and B have more common activities, such as view the same public pages and posts.

Comments

Leave a Reply

Blogging Calendar

September 2015
M T W T F S S
 123456
78910111213
14151617181920
21222324252627
282930  

Archives