Table of Contents
A New Kind of Search Suggestions at Google Based on Tasks
Google shows a set of drop down search suggestions from their search box that complete queries using auto completion. I have written about search suggestions at Google in the past:
- June 10, 2016 – How Google May Map a Query to Entity Search Suggestions (Updated)
- April 20, 2017 – Local Query Suggestions Based Upon Where People Search
- August 23, 2017 – How Google Predicts Autocomplete Query Suggestions is Updated
- May 11, 2020 – Search Suggestions from Previously Submitted Searcher Queries
A newly granted Google patent is about search suggestions that provide a searcher with task search suggestions for a query.
Related Content:
A searcher may try to find information for a task that needs to be completed. For instance, searching for information about booking a flight.
The patent tells us it is about identifying a query and returning a task search suggestion in response. The search engine may associate the searcher with the task when the searcher selects the task suggestion from a suggestion dropdown. For example, tasks could be identified based on terms in a query. Tasks may also be identified from data that is associated with a searcher. The patent also tells us that additional information related to completing a task may be associated with task search suggestions.
The process behind the patent includes:
- Identifying a query of a searcher
- Identifying a number of suggestions for the query with at least one task suggestion associated with at least one task
- Providing one or more of the suggestions to the user, including the task suggestion associated with the at least one task
- Receiving a searcher selection of the task suggestion
- Associating the task with the searcher in response to receiving the selection of the task suggestion
The method may further include the step of indicating at least one task and providing the task suggestion to the user. It may also allow a searcher to perform a task, such as adding a reminder to a calendar application, or sending an email to a person on a contact list, or opening a browser to a place where a bill could be paid.
The task suggested may be determined based on additional user data that is associated with the searcher. That information can include searcher contact information, searcher webpage navigation history, and searcher e-mails.
Additional Information About Tasks, and the Ability to Complete Tasks
The process from the patent may also include:
- Determining additional information related to completion of the task
- Associating the additional information with the task
- Identifying a task type for the task
- Determining additional information related to completing a task may be based on the identified task type
That additional information may be from the searcher’s contacts, webpage navigation history, or e-mails.
The patent shows an example of a task search suggestion and a way to complete a task identified in the suggestion (in this case, making an appointment):
This task search suggestions patent can be found at:
Associating a task with a user based on user selection of a query suggestion
Inventors: Andrew Tomkins, Tristan Harris, Can Sar, Angelo DiNardi
Assignee: GOOGLE LLC
US Patent: 10,776,407
Granted: September 15, 2020
Filed: November 28, 2017
Abstract
Methods and apparatus related to associating a task with a user based on the user selecting a task suggestion provided to the user in response to a user query. In some implementations, the task may be identified based on similarities between the words and/or phrases of the user query and a task suggestion associated with a task. In some implementations, the task may be identified based on user data associated with the user. In some implementations, the task may be associated with additional information related to completing the task.
More About Task Search Suggestions
A searcher may submit the query “pay my cell phone bill” to pay a cellular phone bill and/or store information related to the process as a future reminder.
A searcher may also submit the query “flight information” to locate information about flights.
The searcher may be shown one or more task search suggestions sharing one or more aspects with the submitted query of the searcher.
A task search suggestion may be an autocomplete suggestion that builds on a partial query of a searcher.
Task Suggestion Types
In some implementations, a task may be categorized with a task type.
A task type may be descriptive of the requirements to perform a task and/or the purpose of the task.
For example, a task may be related to paying a cellular phone bill.
The task may be categorized as a reminder task because the task may be to remind a user to pay a bill, and/or the task may be categorized as a purchase task because the task may be to purchase a product and/or service.
Task types can include:
- Reminder tasks
- Payment tasks
- Event tasks
- Purchasing tasks
- Appointment tasks
- Search query tasks
Additional Information Associated with Tasks
A task may be associated with:
- Flight times
- Destinations
- Airline information
- The purpose of the trip
A task such as “remind me to to to the Grocery Store” can add an entry to a calendar application.
Task Search Queries May Be Marked Differently
The patent tells us that Google could mark a task suggestion with an icon or words to indicated that it is a task suggestion. However, I cannot see any icons or words on my desktop or my phone for a task search suggestion. Words that could be used for such a task might be something such as “Create task” to designate a task suggestion. Perhaps we will see that in the future for such suggestions. In addition, choosing a task Suggestion may offer the option to fulfill that task – this doesn’t seem to be possible in Ordinary Google Search, though Google Assistant can implement tasks.
Some task suggestions may be remembered as past queries used previously by a searcher.
Ranking Task Search Suggestions
Task suggestions may be ranked (for when a suggestion may be possibly associated with more than one task.)
The patent tells us about the factors that might be used to rank task search suggestions:
- Number of terms in the query suggestions
- Length of term(s) in the query suggestions
- Display parameters of the computing device
Terms in a query that show an association with a task may be identified as a task and associated with one when they frequently appear in query suggestions related to tasks, such as one that starts with “pay my…” That is a sign that someone is interested in a payment-type task.
A query that shares words with task suggestions, such as “cellular phone bill,” which shares terms with a task suggestion such as “pay my cellular phone bill,” may also be boosted ats a ranked task suggestion because of the similarity of words with a task in a list of task suggestions.
Some task suggestions may be boosted because they are specific to a searcher based on one or more documents associated with a searcher, such as:
- Emails
- Recently displayed webpages
- Search history
- Calendar entries
- Social media notifications
- Text notifications
- Contacts lists
A task recognition engine might identify a task based on some latent attribute of a searcher. These could include things such as:
- Geographic location of the searcher
- Age of the searcher
- Occupation of the searcher from one or more sources
This is true if the formulated task suggestion is specific to one or more latent attributes.
An example is when a searcher’s calendar says to “Buy Bob a birthday present” and based on the identification of the calendar entry and the similarity of the text of the calendar entry and a query.
Takeaways Regarding Task Search Suggestions
Some of the features described in this patent can be used now, such as setting reminder tasks using the Google Assistant. However, it seems like it would make sense to build some of the other types of tasks into a Google Assistant. For example, we may never see the ability within the search to see task search suggestions and select those who can perform the tasks, such as making a payment or purchasing something or making an appointment, but instead that could become part of Google Assistant.
The patent does mention knowledge graphs and entities associated with tasks, which reminds me of the active ontology that powers Siri. A device used to query task suggestions from may contain a database that recognizes entities
In some implementations, task recognition engine 115 may identify a task and/or additional information related to the completion of a task based on mappings between multiple entities and/or mappings between entities and one or more entity properties. The task recognition engine 115 may identify the entities and/or entity properties by accessing one or more databases such as an entity database.
For example, an entity database may include one or more knowledge graphs mapping entities and entity properties and/or entities and other entities. In some implementations, task recognition engine 115 may determine a likely task and/or property of a task based on mappings between multiple entities and/or entities and entity properties, combined with artificial intelligence having predictive capabilities.
For example, in some implementations, one or more mappings between two or more entities may define relationships between the entities that may be utilized to determine task suggestions. For example, a calendar entry may include the terms “Buy present for Bob’s birthday.” Based on the terms, an entity associated with “birthday party” and an entity associated with the action of “buying a present” may be identified.
That similarity to Siri’s Active Ontology shows that this task search suggestion approach, tied to ways to complete those tasks, enables a device to complete such tasks suggested and be an intelligent assistant.
That is why I believe that these task search suggestions will likely become part of Google Assistant instead of part of all searches on Google.
Search News Straight To Your Inbox
*Required
About Bill Slawski
With more than 26 years of SEO experience and a Juris Doctor Degree, Bill Slawski is the foremost expert on Google’s patents as related to SEO. Patent Exploration is one of the quickest and most detailed ways to find new information about SEO. Bill is the Editor of SEO by the Sea, a prominent search engine optimization blog, where he is the author of over 1,300 posts. Bill’s experience includes Fortune 500 brands and some of the largest websites in the world. Bill is a contributing author for Moz, Search Engine Land, and Search Engine Journal. In 2014-2021, he spoke at industry-leading international conferences about topics including search engine algorithms, universal and blended search, personalization in search, search and social, and duplicate content problems, structured data, and schemaJoin thousands of marketers to get the best search news in under 5 minutes. Get resources, tips and more with The Splash newsletter:
