Google Chrome has received a new update, and among its improvements we found one that may be of great interest to users of this browser. To be precise, the main improvements brought by this update (M124) affect Mountain View's Omnibox address bar, and therefore all of us who use it in our daily lives. effect.
The way this can be significantly improved is by allowing users to receive better recommendations through machine learning. This way, you'll be able to receive better recommendations for what you need or are looking for. This new feature replaces the previously mentioned algorithm that the Chrome address bar previously used to provide recommendations.
Google Chrome's address bar suggestions are about to change permanently
The previous operation of this Omnibox to view recommendations used an algorithm that assigned a score to each recommendation, but according to the developers themselves, this method is outdated for modern needs. Therefore, this is why the recommendations themselves have not changed for a long time, and what led to the implementation of these machine learning models.
However, while the scoring system remains, this new update increases or decreases the score based on user actions, allowing Omnibox to provide more accurate and relevant recommendations. In other words, Google can now “collect more recent signals and periodically retrain, evaluate, and implement new models over time,” as stated in a statement on the official Chromium blog. .
The behavior of this new method can be easily illustrated with the following example. If you type the suggested address into the address bar and immediately click the bar again, Chrome will know that this page is not the one you were looking for. In this case, your score will decrease. Therefore, the suggestion will appear further down in the recommendations and may not even appear if your score is very low.
They also talked about the future and potential of this new machine learning method. For example, Chrome developers say that in the future, they “may be able to incorporate new signals, such as differentiating between time of day for better relevance,” so they can provide more accurate recommendations. states that it is possible. It also indicated that they would like to “consider specialized versions of training models for specific environments, such as mobile, business, academic users, or perhaps different locations.”
As long as you're using the latest version of Mountain View's browser, this new machine learning system for the Google Chrome address bar should already be active. If you don't see any changes, it's probably only a matter of time before you start noticing changes in the recommendations.
