Jun 9 2025
Google’s Local Job Type Algorithm Detailed In Research Paper via @sejournal, @martinibuster

Google published a research paper describing how it extracts “services offered” information from local business sites to add it to business profiles in Google Maps and Search. The algorithm describes specific relevance factors and confirms that the system has been successfully in use for a year.

What makes this research paper especially notable is that one of the authors is Marc Najork, a distinguished research scientist at Google who is associated with many milestones in information retrieval, natural language processing, and artificial intelligence.

The purpose of this system is to make it easier for users to find local businesses that provide the services they are looking for. The paper was published in 2024 (according to the Internet Archive) and is dated 2023.

The research paper explains:

“…to reduce user effort, we developed and deployed a pipeline to automatically extract the job types from business websites. For example, if a web page owned by a plumbing business states: “we provide toilet installation and faucet repair service”, our pipeline outputs toilet installation and faucet repair as the job types for this business.”

Developing A Local Search System

The first step for creating a system for crawling and extracting job type information was to create training data from scratch. They selected billions of home pages that are listed in Google business profiles and extracted job type information from tables and formatted lists on home pages or pages that were one click away from the home pages. This job type data became the seed set of job types.

The extracted job type data was used as search queries, augmented with query expansion (synonyms) to expand the list of job types to include all possible variations of job type keyword phrases.

Second Step: Fixing A Relevance Problem

Google’s researchers applied their system on the billions of pages and it didn’t work as intended because many pages had job type phrases that were not describing services offered.

The research paper explains:

“We found that many pages mention job type names for other purposes like giving life tips. For example, a web page that teaches readers to deal with bed bugs might contain a sentence like a solution is to call home cleaning services if you find bed bugs in your home. They usually provide services like bed bug control. Though this page mentions multiple job type names, the page is not provided by a home cleaning business.”

Limiting the crawling and indexing to identifying job type keyword phrases resulted in false positives. The solution was to incorporate sentences that surrounded the keyword phrases so that they could better understand the context of the job type keyword phrases.

The success of using surrounding text is explained:

“As shown in Table 2, JobModelSurround performs significantly better than JobModel, which suggests that the surrounding words could indeed explain the intent of the seed job type mentions. This successfully improves the semantic understanding without processing the entire text of each page, keeping our models efficient.”

SEO Insight
The described local search algorithm is purposely excluding all information on the page and zeroing in on job type keyword phrases and surrounding words and phrases around those keywords. This shows the importance of how the words around important keyword phrases can provide context for the keyword phrases and make it easier for Google’s crawlers to understand what the page is about without having to process the entire web page.

SEO Insight
Another insight is that Google is not indexing the entire web page for the limited purpose of identifying job type keyword phrases. The algorithm is hunting for the keyword phrase and surrounding keyword phrases.

SEO Insight
The concept of analyzing only a part of a page is similar to Google’s Centerpiece Annotation where a section of content is identified as the main topic of the page. I’m not saying these are related. I’m just pointing out one feature out of many where a Google algorithm zeroes in on just a section of a page.

The System Uses BERT

Google used the BERT language model to classify whether phrases extracted from business websites describe actual job types. BERT was fine-tuned on labeled examples and given additional context such as website structure, URL patterns, and business category to improve precision without sacrificing scalability.

The Extraction System Can Be Generalized To Other Contexts

An interesting finding detailed by the research paper is that the system they developed can be used in areas (domains) other than local businesses, such as “expertise finding, legal and medical information extraction.”

They write:

“The lessons we shared in developing the largescale extraction pipeline from scratch can generalize to other information extraction or machine learning tasks. They have direct applications to domain-specific extraction tasks, exemplified by expertise finding, legal and medical information extraction.

Three most important lessons are:

(1) utilizing the data properties such as structured content could alleviate the cold start problem of data annotation;

(2) formulating the task as a retrieval problem could help researchers and practitioners deal with a large dataset;

(3) the context information could improve the model quality without sacrificing its scalability.”

Job Type Extract Is A Success

The research paper says that their system is a success, it has a high level of precision (accuracy) and that it is scalable. The research paper says that it has already been in use for a year. The research is dated 2023 but according to the Internet Archive (Wayback Machine), it was published sometime in July 2024.

The researchers write:

“Our pipeline is executed periodically to keep the extracted content up-to-date. It is currently deployed in production, and the output job types are surfaced to millions of Google Search and Maps users.”

Takeaways

  • Google’s Algorithm That Extracts Job Types from Webpages
    Google developed an algorithm that extracts “job types” (i.e., services offered) from business websites to display in Google Maps and Search.
  • Pipeline Extracts From Unstructured Content
    Instead of relying on structured HTML elements, the algorithm reads free-text content, making it effective even when services are buried in paragraphs.
  • Contextual Relevance Is Important
    The system evaluates surrounding words to confirm that service-related terms are actually relevant to the business, improving accuracy.
  • Model Generalization Potential
    The approach can be applied to other fields like legal or medical information extraction, showing how it can be applied to other kinds of knowledge.
  • High Accuracy and Scalability
    The system has been deployed for over a year and delivers scalable, high-precision results across billions of webpages.

Google published a research paper about an algorithm that automatically extracts service descriptions from local business websites by analyzing keyword phrases and their surrounding context, enabling more accurate and up-to-date listings in Google Maps and Search. This technique avoids dependence on HTML structure and can be adapted for use in other industries where extracting information from unstructured text is needed.

Read the research paper abstract and download the PDF version here:

Job Type Extraction for Service Businesses

Featured Image by Shutterstock/ViDI Studio

Ecommerce MGMT 0 Comments
Jun 3 2025
Google Patent On Using Contextual Signals Beyond Query Semantics via @sejournal, @martinibuster

A patent recently filed by Google outlines how an AI assistant may use at least five real-world contextual signals, including identifying related intents, to influence answers and generate natural dialog. It’s an example of how AI-assisted search modifies responses to engage users with contextually relevant questions and dialog, expanding beyond keyword-based systems.

The patent describes a system that generates relevant dialog and answers using signals such as environmental context, dialog intent, user data, and conversation history. These factors go beyond using the semantic data in the user’s query and show how AI-assisted search is moving toward more natural, human-like interactions.

In general, the purpose of filing a patent is to obtain legal protection and exclusivity for an invention and the act of filing doesn’t indicate that Google is actually using it.

The patent uses examples of spoken dialog but it also states the invention is not limited to audio input:

“Notably, during a given dialog session, a user can interact with the automated assistant using various input modalities, including, but not limited to, spoken input, typed input, and/or touch input.”

The name of the patent is, Using Large Language Model(s) In Generating Automated Assistant response(s). The patent applies to a wide range of AI assistants that receive inputs via the context of typed, touch, and speech.

There are five factors that influence the LLM modified responses:

  1. Time, Location, And Environmental Context
  2. User-Specific Context
  3. Dialog Intent & Prior Interactions
  4.  Inputs (text, touch, and speech)
  5. System & Device Context

The first four factors influence the answers that the automated assistant provides and the fifth one determines whether to turn off the LLM-assisted part and revert to standard AI answers.

Time, Location, And Environmental

There are three contextual factors: time, location and environmental that provide contexts that are not existent in keywords and influence how the AI assistant responds. While these contextual factors, as described in the patent, aren’t strictly related to AI Overviews or AI Mode, they do show how AI-assisted interactions with data can change.

The patent uses the example of a person who tells their assistant they’re going surfing. A standard AI response would be a boilerplate comment to have fun or to enjoy the day. The LLM-assisted response described in the patent would generate a response based on the geographic location and time to generate a comment about the weather like the potential for rain. These are called modified assistant outputs.

The patent describes it like this:

“…the assistant outputs included in the set of modified assistant outputs include assistant outputs that do drive the dialog session in manner that further engages the user of the client device in the dialog session by asking contextually relevant questions (e.g., “how long have you been surfing?”), that provide contextually relevant information (e.g., “but if you’re going to Example Beach again, be prepared for some light showers”), and/or that otherwise resonate with the user of the client device within the context of the dialog session.”

User-Specific Context

The patent describes multiple user-specific contexts that the LLM may use to generate a modified output:

  • User profile data, such as preferences (like food or types of activity).
  • Software application data (such as apps currently or recently in use).
  • Dialog history of the ongoing and/or previous assistant sessions.

Here’s a snippet that talks about various user profile related contextual signals:

“Moreover, the context of the dialog session can be determined based on one or more contextual signals that include, for example, ambient noise detected in an environment of the client device, user profile data, software application data, ….dialog history of the dialog session between the user and the automated assistant, and/or other contextual signals.”

Related Intents

An interesting part of the patent describes how a user’s food preference can be used to determine a related intent to a query.

“For example, …one or more of the LLMs can determine an intent associated with the given assistant query… Further, the one or more of the LLMs can identify, based on the intent associated with the given assistant query, at least one related intent that is related to the intent associated with the given assistant query… Moreover, the one or more of the LLMs can generate the additional assistant query based on the at least one related intent. “

The patent illustrates this with the example of a user saying that they’re hungry. The LLM will then identify related contexts such as what type of cuisine the user enjoys and the itent of eating at a restaurant.

The patent explains:

“In this example, the additional assistant query can correspond to, for example, “what types of cuisine has the user indicated he/she prefers?” (e.g., reflecting a related cuisine type intent associated with the intent of the user indicating he/she would like to eat), “what restaurants nearby are open?” (e.g., reflecting a related restaurant lookup intent associated with the intent of the user indicating he/she would like to eat)… In these implementations, additional assistant output can be determined based on processing the additional assistant query.”

System & Device Context

The system and device context part of the patent is interesting because it enables the AI to detect if the context of the device is that it’s low on batteries, and if so, it will turn off the LLM-modified responses. There are other factors such as whether the user is walking away from the device, computational costs, etc.

Takeaways

  • AI Query Responses Use Contextual Signals
    Google’s patent describes how automated assistants can use real-world context to generate more relevant and human-like answers and dialog.
  • Contextual Factors Influence Responses
    These include time/location/environment, user-specific data, dialog history and intent, system/device conditions, and input type (text, speech, or touch).
  • LLM-Modified Responses Enhance Engagement
    Large language models (LLMs) use these contexts to create personalized responses or follow-up questions, like referencing weather or past interactions.
  • Examples Show Practical Impact
    Scenarios like recommending food based on user preferences or commenting on local weather during outdoor plans demonstrates how real-world contexts can influence how AI responds to user queries.

This patent is important because millions of people are increasingly engaging with AI assistants, thus it’s relevant to publishers, ecommerce stores, local businesses and SEOs.

It outlines how Google’s AI-assisted systems can generate personalized, context-aware responses by using real-world signals. This enables assistants to go beyond keyword-based answers and respond with relevant information or follow-up questions, such as suggesting restaurants a user might like or commenting on weather conditions before a planned activity.

Read the patent here:

Using Large Language Model(s) In Generating Automated Assistant response(s).

Featured Image by Shutterstock/Visual Unit

Ecommerce MGMT 0 Comments
Apr 14 2025
Marketing To Machines Is The Future – Research Shows Why via @sejournal, @martinibuster

A new research paper explores how AI agents interact with online advertising and what shapes their decision-making. The researchers tested three leading LLMs to understand which kinds of ads influence AI agents most and what this means for digital marketing. As more people rely on AI agents to research purchases, advertisers may need to rethink strategy for a machine-readable, AI-centric world and embrace the emerging paradigm of “marketing to machines.”

Although the researchers were testing if AI agents interacted with advertising and what kinds influenced them the most, their findings also show that well-structured on-page information, like pricing data, is highly influential, which opens up areas to think about in terms of AI-friendly design.

An AI agent (also called agentic AI) is an autonomous AI assistant that performs tasks like researching content on the web, comparing hotel prices based on star ratings or proximity to landmarks, and then presenting that information to a human, who then uses it to make decisions.

AI Agents And Advertising

The research is titled Are AI Agents Interacting With AI Ads? and was conducted at the University of Applied Sciences Upper Austria. The research paper cites previous research on the interaction between AI Agents and online advertising that explore the emerging relationships between agentic AI and the machines driving display advertising.

Previous research on AI agents and advertising focused on:

  • Pop-up Vulnerabilities
    Vision-language AI agents that aren’t programmed to avoid advertising can be tricked into clicking on pop-up ads at a rate of 86%.
  • Advertising Model Disruption
    This research concluded that AI agents bypassed sponsored and banner ads but forecast disruption in advertising as merchants figure out how to get AI agents to click on their ads to win more sales.
  • Machine-Readable Marketing
    This paper makes the argument that marketing has to evolve toward “machine-to-machine” interactions and “API-driven marketing.”

The research paper offers the following observations about AI agents and advertising:

“These studies underscore both the potential and pitfalls of AI agents in online advertising contexts. On one hand, agents offer the prospect of more rational, data-driven decisions. On the other hand, existing research reveals numerous vulnerabilities and challenges, from deceptive pop-up exploitation to the threat of rendering current advertising revenue models obsolete.

This paper contributes to the literature by examining these challenges, specifically within hotel booking portals, offering further insight into how advertisers and platform owners can adapt to an AI-centric digital environment.”

The researchers investigate how AI agents interact with online ads, focusing specifically on hotel and travel booking platforms. They used a custom built travel booking platform to perform the testing, examining whether AI agents incorporate ads into their decision-making and explored which ad formats (like banners or native ads) influence their choices.

How The Researchers Conducted The Tests

The researchers conducted the experiments using two AI agent systems: OpenAI’s Operator and the open-source Browser Use framework. Operator, a closed system built by OpenAI, relies on screenshots to perceive web pages and is likely powered by GPT-4o, though the specific model was not disclosed.

Browser Use allowed the researchers to control for the model used for the testing by connecting three different LLMs via API:

  • GPT-4o
  • Claude Sonnet 3.7
  • Gemini 2.0 Flash

The setup with Browser Use enabled consistent testing across models by enabling them to use the page’s rendered HTML structure (DOM tree) and recording their decision-making behavior.

These AI agents were tasked with completing hotel booking requests on a simulated travel site. Each prompt was designed to reflect realistic user intent and tested the agent’s ability to evaluate listings, interact with ads, and complete a booking.

By using APIs to plug in the three large language models, the researchers were able to isolate differences in how each model responded to page data and advertising cues, to observe how AI agents behave in web-based decision-making tasks.

These are the ten prompts used for testing purposes:

  1. Book a romantic holiday with my girlfriend.
  2. Book me a cheap romantic holiday with my boyfriend.
  3. Book me the cheapest romantic holiday.
  4. Book me a nice holiday with my husband.
  5. Book a romantic luxury holiday for me.
  6. Please book a romantic Valentine’s Day holiday for my wife and me.
  7. Find me a nice hotel for a nice Valentine’s Day.
  8. Find me a nice romantic holiday in a wellness hotel.
  9. Look for a romantic hotel for a 5-star wellness holiday.
  10. Book me a hotel for a holiday for two in Paris.

What the Researchers Discovered

Ad Engagement With Ads

The study found that AI agents don’t ignore online advertisements, but their engagement with ads and the extent to which those ads influence decision-making varies depending on the large language model.

OpenAI’s GPT-4o and Operator were the most decisive, consistently selecting a single hotel and completing the booking process in nearly all test cases.

Anthropic’s Claude Sonnet 3.7 showed moderate consistency, making specific booking selections in most trials but occasionally returning lists of options without initiating a reservation.

Google’s Gemini 2.0 Flash was the least decisive, frequently presenting multiple hotel options and completing significantly fewer bookings than the other models.

Banner ads were the most frequently clicked ad format across all agents. However, the presence of relevant keywords had a greater impact on outcomes than visuals alone.

Ads with keywords embedded in visible text influenced model behavior more effectively than those with image-based text, which some agents overlooked. GPT-4o and Claude were more responsive to keyword-based ad content, with Claude integrating more promotional language into its output.

Use Of Filtering And Sorting Features

The models also differed in how they used interactive web page filtering and sorting tools.

  • Gemini applied filters extensively, often combining multiple filter types across trials.
  • GPT-4o used filters rarely, interacting with them only in a few cases.
  • Claude used filters more frequently than GPT-4o, but not as systematically as Gemini.

Consistency Of AI Agents

The researchers also tested for consistency of how often agents, when given the same prompt multiple times, picked the same hotel or offered the same selection behavior.

In terms of booking consistency, both GPT-4o (with Browser Use) and Operator (OpenAI’s proprietary agent) consistently selected the same hotel when given the same prompt.

Claude showed moderately high consistency in how often it selected the same hotel for the same prompt, though it chose from a slightly wider pool of hotels compared to GPT-4o or Operator.

Gemini was the least consistent, producing a wider range of hotel choices and less predictable results across repeated queries.

Specificity Of AI Agents

They also tested for specificity, which is how often the agent chose a specific hotel and committed to it, rather than giving multiple options or vague suggestions. Specificity reflects how decisive the agent is in completing a booking task. A higher specificity score means the agent more often committed to a single choice, while a lower score means it tended to return multiple options or respond less definitively.

  • Gemini had the lowest specificity score at 60%, frequently offering several hotels or vague selections rather than committing to one.
  • GPT-4o had the highest specificity score at 95%, almost always making a single, clear hotel recommendation.
  • Claude scored 74%, usually selecting a single hotel, but with more variation than GPT-4o.

The findings suggest that advertising strategies may need to shift toward structured, keyword-rich formats that align with how AI agents process and evaluate information, rather than relying on traditional visual design or emotional appeal.

What It All Means

This study investigated how AI agents for three language models (GPT-4o, Claude Sonnet 3.7, and Gemini 2.0 Flash) interact with online advertisements during web-based hotel booking tasks. Each model received the same prompts and completed the same types of booking tasks.

Banner ads received more clicks than sponsored or native ad formats, but the most important factor in ad effectiveness was whether the ad contained relevant keywords in visible text. Ads with text-based content outperformed those with embedded text in images. GPT-4o and Claude were the most responsive to these keyword cues, and Claude was also the most likely among the tested models to quote ad language in its responses.

According to the research paper:

“Another significant finding was the varying degree to which each model incorporated advertisement language. Anthropic’s Claude Sonnet 3.7 when used in ‘Browser Use’ demonstrated the highest advertisement keyword integration, reproducing on average 35.79% of the tracked promotional language elements from the Boutique Hotel L’Amour advertisement in responses where this hotel was recommended.”

In terms of decision-making, GPT-4o was the most decisive, usually selecting a single hotel and completing the booking. Claude was generally clear in its selections but sometimes presented multiple options. Gemini tended to frequently offer several hotel options and completed fewer bookings overall.

The agents showed different behavior in how they used a booking site’s interactive filters. Gemini applied filters heavily. GPT-4o used filters occasionally. Claude’s behavior was between the two, using filters more than GPT-4o but not as consistently as Gemini.

When it came to consistency—how often the same hotel was selected when the same prompt was repeated—GPT-4o and Operator showed the most stable behavior. Claude showed moderate consistency, drawing from a slightly broader pool of hotels, while Gemini produced the most varied results.

The researchers also measured specificity, or how often agents made a single, clear hotel recommendation. GPT-4o was the most specific, with a 95% rate of choosing one option. Claude scored 74%, and Gemini was again the least decisive, with a specificity score of 60%.

What does this all mean? In my opinion, these findings suggest that digital advertising will need to adapt to AI agents. That means keyword-rich formats are more effective than visual or emotional appeals, especially as machines increasingly are the ones interacting with ad content. Lastly, the research paper references structured data, but not in the context of Schema.org structured data. Structured data in the context of the research paper means on-page data like prices and locations and it’s this kind of data that AI agents engage best with.

The most important takeaway from the research paper is:

“Our findings suggest that for optimizing online advertisements targeted at AI agents, textual content should be closely aligned with anticipated user queries and tasks. At the same time, visual elements play a secondary role in effectiveness.”

That may mean that for advertisers, designing for clarity and machine readability may soon become as important as designing for human engagement.

Read the research paper:

Are AI Agents interacting with Online Ads?

Featured Image by Shutterstock/Creativa Images

Ecommerce MGMT 0 Comments
Apr 10 2025
Google Files Patent On Personal History-Based Search via @sejournal, @martinibuster

Google recently filed a patent for a way to provide search results based on a user’s browsing and email history. The patent outlines a new way to search within the context of a search engine, within an email interface, and through a voice-based assistant (referred to in the patent as a voice-based dialog system).

A problem that many people have is that they can remember what they saw but they can’t remember where they saw it or how they found it. The new patent, titled Generating Query Answers From A User’s History, solves that problem by helping people find information they’ve previously seen within a webpage or an email by enabling them to ask for what they’re looking for using everyday language such as “What was that article I read last week about chess?”

The problem the invention solves is that traditional search engines don’t enable users to easily search their own browsing or email history using natural language. The invention works by taking a user’s spoken or typed question, recognizing that the question is asking for previously viewed content, and then retrieving search results from the user’s personal history (such as their browser history or emails). In order to accomplish this it uses filters like date, topic, or device used.

What’s novel about the invention is the system’s ability to understand vague or fuzzy natural language queries and match them to a user’s specific past interactions, including showing the version of a page as it looked when the user originally saw it (a cached version of the web page).

Query Classification (Intent) And Filtering

Query Classification

The system first determines whether the intent of the user’s spoken or typed query is to retrieve previously accessed information. This process is called query classification and involves analyzing the phrasing of the query to detect the intent. The system compares parts of the query to known patterns associated with history-seeking questions and uses techniques like semantic analysis and similarity thresholds to identify if the user’s intent is to seek something they’d seen before, even when the wording is vague or conversational.

The similarity threshold is an interesting part of the invention because it compares what the user is saying or typing to known history-seeking phrases to see if they are similar. It’s not looking for an exact match but rather a close match.

Filtering

The next part is filtering, and it happens after the system has identified the history-seeking intent. It then applies filters such as the topic, time, or device to limit the search to content from the user’s personal history that matches those criteria.

The time filter is a way to constrain the search to within a specific time frame that’s mentioned or implied in the search query. This helps the system narrow down the search results to what the user is trying to find. So if a user speaks phrases like “last week” or “a few days ago” then it knows to restrict the query to those respective time frames.

An interesting quality of the time filter is that it’s applied with a level of fuzziness, which means it’s not exact. So when a person asks the voice assistant to find something from the past week it won’t do a literal search of the past seven days but will expand it to a longer period of time.

The patent describes the fuzzy quality of the time filter:

“For example, the browser history collection… may include a list of web pages that were accessed by the user. The search engine… may obtain documents from the index… based on the filters from the formatted query.

For example, if the formatted query… includes a date filter (e.g., “last week”) and a topic filter (e.g., “chess story”), the search engine… may retrieve only documents from the collection… that satisfy these filters, i.e., documents that the user accessed in the previous week that relate to a “chess story.”

In this example, the search engine… may apply fuzzy time ranges to the “last week” filter to account for inaccuracies in human memory. In particular, while “last week” literally refers to the seven calendar days of the previous week, the search engine… may search for documents over a wider range, e.g., anytime in the past two weeks.”

Once a query is classified as asking for something that was previously seen, the system identifies details in the user’s phrasing that are indicative of topic, date or time, source, device, sender, or location and uses them as filters to search the user’s personal history.

Each filter helps narrow the scope of the search to match what the user is trying to recall: for example, a topic filter (“turkey recipe”) targets the subject of the content; a time filter (“last week”) restricts results to when it was accessed; a source filter (“WhiteHouse.gov”) limits the search to specific websites; a device filter (e.g., “on my phone”) further restricts the search results from a certain device; a sender filter (“from grandma”) helps locate emails or shared content; and a location filter (e.g., “at work”) restricts results to those accessed in a particular physical place.

By combining these context-sensitive filters, the system mimics the way people naturally remember content in order to help users retrieve exactly what they’re looking for, even when their query is vague or incomplete.

Scope of Search: What Is Searched

The next part of the patent is about figuring out the scope of what is going to be searched, which is limited to predefined sources such as browser history, cached versions of web pages, or emails. So, rather than searching the entire web, the system focuses only on the user’s personal history, making the results more relevant to what the user is trying to recall.

Cached Versions of Previously Viewed Content

Another interesting feature described in the patent is web page caching. Caching refers to saving a copy of a web page as it appeared when the user originally viewed it. This enables the system to show the user that specific version of the page in search results, rather than the current version, which may have changed or been removed.

The cached version acts like a snapshot in time, making it easier for the user to recognize or remember the content they are looking for. This is especially useful when the user doesn’t remember precise details like the name of the page or where they found it, but would recognize it if they saw it again. By showing the version that the user actually saw, the system makes the search experience more aligned with how people remember things.

Potential Applications Of The Patent Invention

The system described in the patent can be applied in several real-world contexts where users may want to retrieve content they’ve previously seen:

Search Engines

The patent refers multiple times to the use of this technique in the context of a search engine that retrieves results not from the public web, but from the user’s personal history, such as previously visited web pages and emails. While the system is designed to search only content the user has previously accessed, the patent notes that some implementations may also include additional documents relevant to the query, even if the user hasn’t viewed them before.

Email Clients

The system treats previously accessed emails as part of the searchable history. For example, it can return an old email like “Grandma’s turkey meatballs” based on vague, natural language queries.

Voice Assistants

The patent includes examples of “a voice-based search” where users speak conversational queries like “I’m looking for a turkey recipe I read on my phone.” The system handles speech recognition and interprets intent to retrieve relevant results from personal history.

Read the entire patent here:

Generating query answers from a user’s history

Ecommerce MGMT 0 Comments