Google just filed a new patent that’s designed to provide users with search results, depending on their browsing and email history.
The Android maker outlined the latest means for search inside the context of search engines through an email interface and even voice voice-based assistant. Many people had the issue that they recall everything that they saw, but not where it was seen or how they located it.
So the latest patent solves this issue and is dubbed Generating Query Answers from User’s History. It’s the ultimate solution by giving users the information that they need but couldn’t get their hands on it as they couldn’t find it.
The data is gathered from previous search history or by glancing over emails or webpages. They enable them to request what they’re searching for in everyday terms. So you can ask Google what article you read recently related to coffee to help find what you’re looking for.
The issue it solves is that classic Google doesn’t enable you to look for things by going through your search history using everything terminology. This new service will take your word of mouth, interpret it, and try to relate it to recent search results from personal history. It makes use of filters such as topics, devices, or dates.
What’s new here is that the invention is the system’s ability to comprehend fuzzy natural language questions and match them to the user’s past interactions. This entails seeing page versions as they appeared when the user saw them first.
The system will be able to do this by figuring out the intention of the user through voice or text. They will classify the query and analyze the phrasing part to determine intention. It uses a similar threshold to relate to the past to the present and link them to find what you’re in search of.
The next step involves filtering. It occurs after systems identify the history-seeking part and apply filters to restrict search from the user’s past history that matches the same criteria. There is a time filter to ensure that the search was done during a certain timeframe, so it’s easier to find.
Every filter helps to make the search narrower and match it to look for exactly what the user wants or remembers. Once the search is narrowed down by the filters, it can retrieve the data on emails or on the page explored. There’s even a location filter to limit searches to a certain physical location.
Image: DIW-Aigen
Read next: Meta Refers to its Latest Llama 4 AI Model as Less Politically Biased That Past Versions
The Android maker outlined the latest means for search inside the context of search engines through an email interface and even voice voice-based assistant. Many people had the issue that they recall everything that they saw, but not where it was seen or how they located it.
So the latest patent solves this issue and is dubbed Generating Query Answers from User’s History. It’s the ultimate solution by giving users the information that they need but couldn’t get their hands on it as they couldn’t find it.
The data is gathered from previous search history or by glancing over emails or webpages. They enable them to request what they’re searching for in everyday terms. So you can ask Google what article you read recently related to coffee to help find what you’re looking for.
The issue it solves is that classic Google doesn’t enable you to look for things by going through your search history using everything terminology. This new service will take your word of mouth, interpret it, and try to relate it to recent search results from personal history. It makes use of filters such as topics, devices, or dates.
What’s new here is that the invention is the system’s ability to comprehend fuzzy natural language questions and match them to the user’s past interactions. This entails seeing page versions as they appeared when the user saw them first.
The system will be able to do this by figuring out the intention of the user through voice or text. They will classify the query and analyze the phrasing part to determine intention. It uses a similar threshold to relate to the past to the present and link them to find what you’re in search of.
The next step involves filtering. It occurs after systems identify the history-seeking part and apply filters to restrict search from the user’s past history that matches the same criteria. There is a time filter to ensure that the search was done during a certain timeframe, so it’s easier to find.
Every filter helps to make the search narrower and match it to look for exactly what the user wants or remembers. Once the search is narrowed down by the filters, it can retrieve the data on emails or on the page explored. There’s even a location filter to limit searches to a certain physical location.
Image: DIW-Aigen
Read next: Meta Refers to its Latest Llama 4 AI Model as Less Politically Biased That Past Versions
