Google in an endeavor to offer more personalized search results for the users and to remain the choicest of search engines has launched certain features. The first is to collect and use data from Gboard on Android phone to offer personalized search results while the other is the “similar items” suggestions panel to help find products that users are looking for keeping their personal choices in mind. Both these features will provide enhanced user search based on their preferences making the search results more accurate. A majority of people are using their smartphones to place a search query or find a particular product. These features are going to simplify the process of search for users.
Providing Personalized Search Results through Gboard
If you are wondering about how Google will provide a more personalized search result through Gboard on Android mobile, then be informed that it is training its AI or artificial intelligence algorithm to collect data from Gboard to offer personalized search results. When an individual searches for result on the application, the app will remember which are the ones that were chosen and which are the ones that got rejected. Accordingly, the personalized result will be provided by the application.
Users using Gboard currently on their Android phones can expect a number of updates to the existing app. Google, in order to provide more personalized search results will utilize the aggregate of changes for the application. However, users will not have to wait for long to experience the new more personalizes search suggestions, as they will immediately provide result post data collection. In fact, Google states that collecting information in this way to provide personalized results will make the affair more private and secure.
Leveraging Federated Learning to Provide Personalized Results
Federated Learning is a new method of training the AI to offer more personalized results. This aids mobile phones to collectively learn a shared prediction model while retaining all the training data on the device itself. This enables machine learning to be able to store data in the cloud. According to Google, this is only the tip of the iceberg and they are all set to achieve even more with Federated Learning. Their spokesperson said, “Beyond Gboard query suggestions, for example, we hope to improve the language models that power your keyboard based on what you actually type on your phone (which can have a style all its own) and photo rankings based on what kinds of photos people look at, share, or delete.”
“Similar Items” Suggestion Benefit
The process of data collection and the related alterations will not affect Gboard’s performance or the phone’s battery life. The updates will take place only when the phone is on a free wireless connection, left unused or is plugged in.
Now, for the second update providing more personalized search; Google’s Image Search has launched “similar items” suggestions on mobile and Android Search application. This will really help shoppers because they will receive suggestions on complimenting products based on their original search thus making the experience more personal.
The suggestions are currently provided for sunglasses, shoes, and handbags. More products will be added to the existing list such as apparels, garden and home decor items. Utilizing the machine vision technology, the “similar items” section recognizes products in lifestyle images to show results for matching products to the user. Additionally, Google is also offering 2 more features like product availability and price which was much requested by users earlier.
Now, from the perspective of the seller. In case, you are a vendor, then adding schema.org product metadata to your products page will help in featuring your products in the ‘similar items’ section. All the products with their image, price, name, availability and metadata can be included. To run this correctly, verify whether the markup has been incorporated properly. To know this, test the pages with Structured Data Testing Tool. In case, the markup has been added of late, it may take up to a week to get crawled and added to search results. This can also be tested by typing “site:yourdomain.com” into Image Search.