In general, an algorithm is a set of instructions to be followed. While mostly used in the context of computers and programming, it can be applied to many other things, such as..
Baking a Cake
Solving a Rubik's
ALGORITHMS IN SOCIAL MEDIA
To keep you constantly engaged, social media sites need to offer you interesting content. This content comes largely from the pool of posts, likes, and photos created by your friends, and what you like.
A post is created (text, photo, video)
Post is categorized into a personalized
When you log in, the algorithm iterates over your friend's index
streams, and ranks them
Now, the system goes through a separate ranking process based on the posts in the index streams.
Clicks, likes, comments, shares, and other interaction is used to rank potential
posts in your feed. These are split up in two ways:
(eg. likes, shares, comments, clicks)
Passive interaction (eg. time spent on a webpage)
Interaction with posts on your feed is now compared to each posts ranking, through a machine learning program. This helps predict future content, as well as the best pattern of content, like
Every time you log in, these three steps are repeated to give you the most personalized content, while also learning from your clicks. To sum it up:
Based on the pool of your friend's likes, shares, comments, and your own personal interests, the algorithm makes a list of possible content to show you.
The algorithm then ranks the list and injects content into your feed, with what it thinks you want to see the most at the top, instead of what was posted chronologically.
Your implicit and explicit interaction with the given content is put back into a machine learning program to better predict content the next time you log in.
Is This Ethical?
With great power, comes great responsibility.
Ben Parker, Spiderman
As these algorithms become smarter, social media sites know more about us. Our interests, passions, likes, dislikes, family, friends, and private info become more available the more we use social media.
While others argue that, if used correctly, these algorithms can help us more than they can hurt us.
Some argue that nothing good can come out of machines knowing this much information.
AN EXAMPLE OF
Because the algorithm favors clicks,
These algorithms take into account how often you interact with different types of posts, the level of engagement the post has received, and its performance among those who viewed it.
More personalized, relevant content
Less time searching
Allows for companies and startups to grow faster
Connects more people (eg. recommended friends)
The more information algorithms know, the more that can be leaked
Algorithms favor clicks, which leads to click bait and fake news
They can choose to hide information you may actually want to see
As technology progresses, and algorithms become smarter, consumers and computer scientists alike will need to look deeply at social media and predictive algorithms to see what's best in the long run. Whether or not they hurt us or harm us will depend on our decisions in the near future.