MIT Algorithm Can Anticipate How Popular A Photo Will Be On Social Web

MIT PhD candidate Aditya Khosla has developed an algorithm which can predict how many views your photo will get in any social networking websites before you upload it. He claimed to have two key components of an image that determines it’s popularity, namely the image content and social context.

Hand holding a Social Media 3d Sphere

Millions of pictures are being uploaded to the internet in every minute via different social networking and photo sharing websites. Since some photos are getting tons of likes and huge popularity, others are simply ignored. Before uploading the photo on Facebook or Instagram or Flickr, can you predict which photo is going to be popular or what makes a photo popular?

Aditya Khosla, who is doing PhD at MIT’s prestigious Computer Science and Artificial Intelligence Lab, had beenĀ researching on this questions. Before this, he had also developed another computer algorithm that can subtly modify the image of a person’s face to make it easier or harder to remember. This time, Khosla claims his algorithm can predict exactly how many likes/views your photo is going to receive even before you upload it. This algorithm will work with some social factors like how many friends/followers you have, the numbers of hash-tag and the length of the title of that photo. It will also check the texture, color, gradient and objects exist in that photo.

During a recent interview with The Verge Khosla said about his algorithm, “It actually represents how many views are expected on your image per day, on average, but in a log scale. What this means: if the score is 5, we expect roughly 2^5 = 32 views on your image per day, or if its 6, roughly 2^6 = 64. However, this metric is not entirely accurate, and the most important thing is the relative score i.e. if A has score higher than B, it’s likely to receive more views.”

Before publishing on social media, you can go through Khosla’s website and upload your photo to check the potential popularity. In future, Khosla wants to improve this algorithm with a tool that will automatically edit your photo to make it more popular!

[ttjad keyword=”adobe-software”]

Leave a Reply