For those people who don’t know what exactly the term ‘Machine Learning’ mean, let me take you to a quick tour.


MACHINE LEARNING is a sub category of Artificial Intelligence(AI) which comes under the field of computer science. Machine learning is generally used to understand the structure of data and to modulate it in a way, which is understood and utilize by people. In other words, machine learning is an idea to learn from examples and experience without being clearly programmed . We provide data in the form of examples to the generic algorithm instead of writing codes, so that the algorithm builds logic, based on the given data.

In this fast-paced world we want machines to work harder than us. After all who wants to work twice as hard when we have options? By applying AI, we can build intelligent machines. Back in the days, with this grand AI exposure, it was difficult to program complex tasks as the challenges kept coming in different ways. People realized that the only way to achieve this task was to let the machine learn from itself. Like a kid! So, it can be said that machine learning is the new ability for computers.

Advantage of it:

Machine learning benefitted every technology user today. Facial recognition technology allows social networking sites to help users tag and share photos with friends easily. Optical character recognition (OCR) technology also helps to convert images into text messages, that too into movable type.

Based on our interests, entertainment platforms show us what movies or shows we should watch next. It also plays an important role when it comes to recognise the difference between spam and no-spam emails. Also there is a lot of improvement in the ads we discover while scrolling down our wall on social media. All this is with the help of recommendation engines, based on the concept of machine learning.

To spot photoshopped images, Adobe is using machine learning:

Back to our news, AI tools are editing images and videos easier than ever, which is the reason of anxiety among the experts around the world. Without checking facts, content can get quickly viral with the help of social media.

It is known that, some of these tools are developed by Adobe itself, but the company is also working on their antidote by researching how machine learning can be used to automatically spot edited pictures. However, Adobe has never released any software before which can spot fake images.

The recent research paper shows that machine learning can be used to identify three types of image malfunction.

  1. Splicing- parts of two different images are combined.
  2. Cloning- objects are copied and pasted in a picture.
  3. Removal- when the whole object is edited out.

In this type of manipulation, digital forensic experts look for hints in the hidden layers of the image. These edits leave behind digital defects and inconsistencies such as the random variations in brightness, color or contrast created by image sensors.

Like other machine learning systems, Adobe uses a large dataset of edited images. With this technology, it learned to spot patterns that indicate interference. This fresh version of Adobe is under developing.

Facebook is using machine learning to fact-check articles:

This is not the first time Facebook has announced that it’s using machine learning to target misleading content. We all know that Facebook is an important target for spammers to spread fake news and hoaxes. The company is doing its best to keep the site secure with the help of AI.

Here’s a thing, machine cannot automatically fact-check and judge the misleading information, but it can easily detect identifiable signals of spam.

According to Facebook’s blog post:

Machine learning helps us identify duplicates of debunked stories. For example, a fact checker in France debunked the claim that you can save a person having a stroke by using a needle to prick their finger and draw blood. This allowed us to identify over 20 domains and over 1,400 links spread that same claim.

In a recent interview, Facebook product manager Tessa Lyons informed about machine learning in detail. She said that the filters are now trying to predict which pages are likely to share bad content. This includes looking for page admins who live in one country but targets users in another, which is a common way for spammers to make money.

Facebook is also working with third-party fact checkers in many countries (now in 14) and is ready to expand this in other nations. It seems like there’s more involvement of human fact checkers ( as they are the most effective) but machine learning is a useful backup for future.