Machine learning and AI in modern web platforms

Today, various types of machine learning technologies are used, and we may have become so accustomed to them that we hardly notice them. For example, apps that recognise objects or your face and adjust the interface to that object. Artificial intelligence is used in agriculture, medicine, banking and various other industries.

As the speed of the internet increases every day, and we are now rolling out 5G, you have to ask which platforms offer some type of machine learning?

Natural language understanding is a type of a machine learning technique that gives insight into the human understanding of phrases and terms. For example, the word “shooting” can appear in articles about shooting from a gun or on the other hand, shooting in football. The first article would be undesirable for advertisers but the second would be perfect (as advertisers love sports).

If you are analysing a single article about the context of the terms used, it could take a few milliseconds from a fast processing computer. But what if I told you that there are hundreds of thousands of articles analysed every millisecond all around the world right now?

How they do that? Well fortunately there are some platforms like AWS’s Amazon Comprehend or IBM’s Watson. They can process huge loads in matter of seconds. Cloud providers like the ones mentioned above also provide other services related to machine learning and AI.

Text recognition in images is another technique that is mostly used on mobiles. For example, take a look at Firebase’s text recognition service. They also offer image labelling, face and object detection. With it, you can use your iPhone app to take photos and automatically label them, or scan a product in a shop and add it to your wish list.

Firebase’s Smart reply service, can empower your chat application to send automatic replies according to the question that your users post.

So why are all these services considered to be machine learning services? Well they all have one thing in common, they have the ability to learn from a past occurrences of events. They adjust their response every time someone sends a request to their endpoint, and some offer fine tuning to the “artificial intelligence”. For example, IBM’s Watson also offers you to insert your own entities and connect them in various relationships, instead of relying on the shared intelligence.

One thing is certain, modern platforms and cloud providers are leading the way in offering machine learning and AI services to the public. Without them smaller companies don’t stand the chance to build their own ecosystem. Thanks to them, they can build their websites and apps using various techniques and offering outstanding services to their users.

Leave a Reply

Your email address will not be published. Required fields are marked *