Google has announced a change to the Omnibox, the search bar in its Chrome browser. It should be optimized with machine learning.
While browsing Chrome, the search bar of the Omnibox allows us to come across previous searches, which we can do regularly, or browser suggestions, which is based on the first letters typed. A rather convenient service, which should be even more relevant in the future thanks to the integration of machine learning models in this search bar.
From hand-optimized formulas to machine learning
The way Omnibox operates is changing profoundly. While the search bar’s prediction system used to work with “hand-built and adapted formulas,” it will now benefit from machine learning.
The goal is to make the Omnibox more flexible and able to “pick up fresher signals, reform, evaluate and implement new models,” in the words of the Mountain View company. As a result, the quality of the bar’s offerings is expected to improve over time.
A change that starts with Chrome 124
And to do this, the Omnibox will take into account the user’s behavior. If the search bar, after suggesting a URL, registers that the user stays on the proposed page for only a short time, it will give that page a low score, which means that it was not the right result to offer.
For Google, machine learning in the Omnibox is a valuable tool for the future. It would open up “many new opportunities to improve the user experience by potentially incorporating new signals, such as differentiating time of day to improve relevance.” And this change is built into the latest Chrome 124 update. Will we see something new in a few weeks?