Taking the Web to Another Level With Machine Learning

“A black box that does magic tricks 🦄”. Maybe that’s the idea that many of us have about machine learning, especially if we have never had an approach to artificial intelligence.

But the reality is that artificial intelligence is becoming more and more relevant in almost every branch of engineering and development, including the web.

But not everything has to be rocket science, right? So let’s take a look at some scenarios where machine learning could take our web applications to the next level 😉

Let’s analyze the data!

This is one of the machine learning applications that comes to mind most quickly: taking the large amount of data we collect and using specialized algorithms to discover patterns or inconsistencies. This analysis of the information can be used to make changes almost in real-time.

It’s time to understand user behavior! 👀

Your web application can use machine learning to accurately understand user behavior. For example, an e-commerce website can apply ML algorithms to monitor and understand a user’s affinity with a product or category. It could even predict expected user actions based on search history and interaction within the results page. Better results and more accurate recommendations can mean more sales and more time the user spends on the website.

Did you know that by using machine learning you could optimize your response times? That’s what the page forecasting model is all about: predicting the next page the user will visit using historical data from Google Analytics. Through this prediction, you can apply techniques to navigate faster.

Where is my 21st-century user experience? ⏳🔊🖖

Web technologies in the 21st century have already evolved to an impressive level. There are already several APIs based on artificial intelligence within browsers* that enable alternative and adaptive experiences.

One example of these technologies is the Web Speech API:

“The Web Speech API adds voice recognition (speech to text) and speech synthesis (text to speech) to JavaScript.” - Eric Bidelman
(Web apps that talk – Introduction to the Speech Synthesis API)

You can create applications that are voice-driven or that integrate voice recognition into forms or search boxes as Google or YouTube do.

Please note that several of these technologies are not fully supported by browsers. For example, Safari supports Speech Synthesis but does not support Speech Recognition.

But wait… audio isn’t everything. The camera can also be used to play/experiment with the user using ml5.js: “machine learning for the web in your web browser”. Through ml5.js we can use a variety of models. For example, PoseNet or Handpose, for real-time pose estimation (let’s play using our body!). The Coding Train has an introductory video that I recommend: ml5.js Pose Estimation with PoseNet.

Artificial intelligence is an exponentially growing trend. Every day we see it more and more in web development. Let’s take advantage of machine learning to make our application an unforgettable experience. Happy hacking!

Join 2000+ Founders and Developers crushing their businesses and careers with monthly advice. You can also follow us on LinkedIn , Twitter & Instagram!