How is a Knowledge Panel for an Entity Triggered?

Sharing is caring!

This patent is about presenting data with search results and deciding when to show that data in a knowledge panel with those search results.

The Web provides access to a wide variety of resources, such as image files, audio files, video files, and web pages.

A search system may identify resources in response to queries submitted by users and provide information about the resources in a manner that is useful to the users.

The users can then navigate through, e.g., click on, the search results to acquire information of interest to the users.

Users of search systems are often searching for information regarding a specific entity.

For example, users may want to learn about a singer that they just heard on the radio.

Conventionally, the user would submit a search query identifying the singer and select from a list of search results determined to be relevant to the search query.

One innovative aspect of the subject matter described in this patent can be shown in methods that include:

  • Receiving a query
  • Obtaining search results responsive to a received query
  • Identifying the first set of factual entities referenced by the received query
  • Selecting from among the first set of factual entities a particular factual entity for which a knowledge panel is eligible to be provided with the search results, the selection is based on a measure of topicality between each identified factual entity in the first set and one or more of the obtained search results
  • Determining that the knowledge panel for the particular factual entity is to be provided with the search results, the determination is based, at least in part, on the content of the knowledge panel and characteristics of the search results, the knowledge panel including at least one content item, for the selected factual entity, received from a first resource and at least one content item, for the selected factual entity, received from a second resource different than the first resource
  • Providing a knowledge panel

Additional features may include:

  • Selecting, from among the first set of factual entities, a particular factual entity for which a knowledge panel is eligible to be provided with the search results can include identifying a second set of factual entities, each factual entity of the second set being a factual entity determined to be topical to the obtained search results
  • Identifying one or more factual entities included in the set of factual entities and the second set of factual entities
  • Selecting one of the one or more factual entities as the identified factual entity

Identifying the second set of factual entities, each factual entity of the second set being a factual entity determined to be topical to the obtained search results.

These can include:

  • Identifying the third set of factual entities, each factual entity of the third set being a factual entity referenced in at least one of the obtained search results
  • Identifying a proper subset of the obtained search results, the proper subset including a particular number of higher-ranked obtained search results; for each factual entity of the third set of factual entities
  • Determining, relative to the factual entity, a partial topicality score for each search result in the proper subset of obtained search results, the partial topicality score for each search result is a measure of topical relatedness between the search result and the factual entity
  • Selecting from the third set of factual entities the second set of factual entities based on the partial topicality scores

Selecting the second set of factual entities based on the partial topicality scores can include, for each factual entity of the third set:

  • Determining an overall topicality score for the factual entity based on the partial topicality scores for the factual entity
  • Selecting the second set of factual entities based on the overall topicality scores
  • Determining the overall topicality score for the factual entity based on the partial topicality scores for the factual entity can include determining a weighted sum of the partial topicality scores for the factual entity

Selecting a Particular Factual Entity

Factual Information about entities may be chosen for a knowledge panel based on:

  • Determining a topicality score for the search result concerning each factual entity of the first set of factual entities, with the topicality score for a search result concerning a factual entity being a measure of topical relatedness of the search result for the factual entity
  • Comparing each topicality score to a topicality threshold to identify a proper subset of the topicality scores that exceed or satisfy the topicality threshold
  • Determining which one of the proper subsets of topicality scores corresponds to the highest-ranking search result for the proper subset of topicality scores
  • Selecting the factual entity that corresponds to the one partial topicality score as the particular factual entity.

Determining That Information for a Factual Entity Should Be Provided With Search Results

This Decision may include:

  • Identifying a click-through rate for each search result with respect to a received query
  • Determining that the identified click-through rates do not satisfy a threshold click-through rate
  • Determining to provide the knowledge panel in response to determining that the identified click-through rates do not satisfy the threshold click-through rate

Determining That Information for a Factual Entity is to be Provided with Search Results

This can include:

  • Identifying a top ranked search result for the received query
  • Identifying a click-through rate for the top ranked search result with respect to a received query
  • Determining that the identified click-through rate does not exceed satisfy a click-through rate threshold
  • Determining to provide the knowledge panel in response to determining that the identified click-through rate does not exceed satisfy the click-through rate threshold

Determining That Information for a Factual Entity will be shown with Search Results

These may include:

  • Identifying content for inclusion in the knowledge panel
  • Determining that the identified content includes at least a threshold amount of content
  • Determining to provide the knowledge panel in response to determining that the identified content includes at least a threshold amount of content

Determining That Information for a Factual Entity is to be shown with Search Results

These can include:

  • Determining whether the knowledge panel includes links to query refinements for the received query
  • Determining to provide the knowledge panel in response to determining that the knowledge panel includes links to query refinements for the received query

Advantages of Following the Process in the Triggering Knowledge Panels Patent

  • Knowledge panels can improve users’ search experiences, in particular for queries directed to learning, browsing, or discovery. For example, the knowledge panel supplies users with basic factual information from multiple different sources or a summary of information about a particular entity referenced in a search query
  • Knowledge panels can assist users in navigating to multiple different network locations that each includes related content seamlessly and naturally
  • Knowledge panels can supply new content that may not otherwise be encountered by a user without the user selecting several different search results and visiting multiple different network locations. Knowledge panels can also help users obtain information faster than they would if the users were required to click through multiple search results to obtain the information from multiple different network locations

This Triggering knowledge Panels patent can be found at:

Triggering knowledge panels
Inventors: John R. Provine
Assignee: Google LLC
US Patent: 10,922,326
Granted: February 16, 2021
Filed: March 14, 2013

Abstract

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for providing knowledge panels with search results.

In one aspect, a method includes receiving a query.

Search results that are responsive to the received query are obtained.

The first set of factual entities referenced by the received query are identified.

A particular factual entity is selected from the first set.

It is determined that a knowledge panel for the particular factual entity is to be provided with the search results based, at least in part, on the content of the knowledge panel and characteristics of the search results.

The knowledge panel can include at least one content item, for the selected factual entity, received from a first resource and at least one content item, for the selected factual entity, received from a second resource different than the first resource.

The knowledge panel is provided.

What is a Knowledge Panel?

A search may return one or more knowledge panels in a response to a search query.

A knowledge panel is a user interface element that provides a collection of information or other content related to an entity that is referenced in a search query.

An entity could be a person, place, country, landmark, animal, historical event, organization, business, sports team, sporting event, movie, song, album, game, work of art, or any other entity.

An entity may also be a concept, subject, or topic.

When it appears in a search result, a knowledge panel shows a summary of information for a specific entity.

A knowledge panel for a singer may include:

  • The name of the singer
  • An image of the singer
  • A description of the singer
  • One or more facts about the singer
  • Content that identifies songs and albums recorded by the singer
  • Links to searches related to the singer
  • Other types of information and content

A knowledge panel for an entity typically will include content that is obtained from multiple disparate sources or multiple different domains. A knowledge panel for a person could include an image from an official website for the person as well as facts from an online encyclopedia not affiliated with the official website for the person.

When a query is received, the search system can determine whether to provide a knowledge panel with search results for the search query. This may be based on whether the query is determined to reference a known entity.

When the system decides whether the query refers to a known entity based on comparing the received query, or terms of the received query, to an entity index that identifies known entities and their aliases.

If there is a match based on the comparison, the received query is determined to refer to the known entity, and the system may provide a knowledge panel for the matching entity.

A query may refer to multiple entities or one or more terms that can be considered an alias of multiple entities. For Instance, the term “phoenix” may refer to a city in Arizona or a mythical bird. For a query that matches, or is otherwise relevant to, multiple known entities, the system can determine whether to show a knowledge panel for one of the matching entities, for example, if one entity is the most likely entity for the query or to provide a knowledge panel that includes content for multiple entities.

If a query includes “Phoenix, Ariz.” or the query was received from a device located in Arizona, the system may determine that the city in Arizona is the most likely entity being referenced by the search query and provide a knowledge panel for the city of Phoenix. If the query includes only “phoenix” the system may provide a knowledge panel that includes content for the city of Phoenix and the mythical bird.

A Disambiguation Knowledge Panel

Another example of a knowledge panel that may include content for multiple entities is a disambiguation knowledge panel.

A disambiguation knowledge panel is a knowledge panel that includes content for two or more different known entities and/or user interface elements that enable a searcher to specify which of the known entities is of interest to the user.

A “phoenix” disambiguation knowledge panel could include content items including an artistic rendering of the mythical bird and a map representing Phoenix, Ariz. Each of these content items can be linked to more information about the respective entities and/or include instructions that upon user interaction with one of the content items request a knowledge panel for the entity represented by the content item.

Very often, a disambiguation knowledge panel includes the same or a similar amount of content for each entity referenced by the disambiguation knowledge panel.

A disambiguation knowledge panel may include titles for the entities, a brief description of the entities, and links that request resources with additional information or that initiate searches for each entity. The area of the knowledge panel that is designated for each entity may also be the same or similar in size.

A Dominant Entity Knowledge Panel

Another type of knowledge panel including content for multiple entities is a dominant entity knowledge panel.

A dominant entity knowledge panel includes content for multiple entities but has a greater amount of content for one or more of the matching entities.

Though more than one entity may match a query, a particular entity may more closely match the query and maybe more topical to results for the query. It could also be more relevant to the query than the other entities.

And Google may show a dominant entity knowledge panel with more content for the particular query, with also less content for each of the other matching entities.

A dominant entity knowledge panel may show a title, image, description, set of facts, links to resources related to the particular entity, and additional content for the particular entity.

That same dominant entity knowledge panel may also show, for each other matching entity, only a title, image, and link to additional content for the entity.

The system may also consider how well an entity matches the query or search results identified for the query to decide which entity to show a knowledge panel for. The system may select an entity closely matching the query and one that is well represented by the search results for the query.

Deciding Between Entitities

The search system may identify the first set of entities that each matches the query. It may also identify a second set of entities that are considered topical to the search results.

The system may then select between entities that are included in both the first and second sets.
This can help to ensure that the selected entity for which to provide a knowledge panel both matches the search query and is also on-topic for the query.

Other Factors may be used to decide whether to show a knowledge panel for a query, such as click-through rates or other performance measures of search results for the received query and/or content available to include in a knowledge panel for an entity referenced by the received query.

The system may think about whether search results for the query have:

  • A low engagement rate, e.g., a low click-through rate
  • Whether the query is non-navigational, e.g., the click-through rate of a high ranked search result for the query is high, whether there is sufficient content, for the entity referenced by the query, to be included in a knowledge panel
  • Whether the knowledge panel for the entity includes links to common search refinements that users would otherwise have to enter manually

Deciding What Data a Knowledge Panel Contains?

A search service may include an apparatus attached to the search system.

The search system can decide whether to provide a knowledge panel for a search query. If it determines a knowledge panel should be shown, the knowledge panel apparatus can generate the knowledge panel and provide the generated knowledge panel.

The knowledge panel apparatus may be provided to the searcher’s device, bypassing the search system.

A knowledge panel is usually shown for queries determined to refer to a particular factual entity, such as a person, place, country, landmark, animal, historical event, organization, business, sports team, sporting event, movie, song, album, game, work of art, or any other entity. A factual entity can also be a concept, subject, or topic.

A knowledge panel for an entity may include content from multiple disparate resources related to the particular entity. A knowledge panel may include factual information relevant to the entity.

A knowledge panel for an actor may include biographical information for the actor, as well as content associated with shows that the actor has appeared in. The knowledge panel may include a summary of information related to the entity.

A knowledge panel for a nation may include:

  • A map of the nation
  • The flag of the nation
  • The official language of the nation
  • Other facts and content related to the nation

Each of these units of information may be from different resources than the other discrete units of information.

A knowledge panel can be much larger and consume more area of a search interface than standard search results. The knowledge panel may span the length or height of three or more standard search results to accommodate the content items in the knowledge panel and to draw attention to the knowledge panel.

A knowledge panel may be displayed in a knowledge panel area. That knowledge panel area may be presented with or alongside a search results area that presents search results. The knowledge panel area may consume a larger area than the area consumed by two or more of the search results presented in the search results area.

The knowledge panel apparatus includes or is communicably coupled to one or more data storage units that include:

  • A content item store
  • A knowledge panel store
  • An entity index

The content item store may store content items that can be inserted. These can be content items are discrete units of content, such as data files, and can be in the form of text, images, videos, graphics, audio, tables, or other types of content.

The knowledge panel store may store knowledge panels for entities and/or knowledge panel templates that can be populated with content to generate a knowledge panel. In general, a knowledge panel template specifies types of content items to include in the knowledge panel and includes placeholders for content items of the specified type. For example, a knowledge panel template may include placeholders for a title, one or more images, a description, a set of facts, and/or other types of content items.

The entity index can include an index of known entities. For each of those entities, this can include data identifying aliases of the entity (a celebrity may have several nicknames or stage names for which a user may enter as a query for the celebrity.) Each of the nicknames or stage names can be included as aliases for the entity in the entity index.

That entity index can also include an index of content items for each entity. The knowledge panel apparatus may identify content items to include in knowledge panels for each entity and may store the content items in the content item store. The knowledge panel apparatus can also index the content items in the entity index for use in subsequent retrieval of the content items.

Knowledge Panel Templates

Depending on the type of entity that might be identified in an entity found in a query, factual information associated with that entity could be shown. The patent provides examples to make that clearer.

One example shown is for a knowledge patent for a famous actor. That panel may show the Famous Actor being credited with movie appearances and music.

The knowledge panel may include a set of images of movie posters for movies in which a Famous Actor appeared. Those images may be taken from an online movie database and/or include a link to the online movie database.

The knowledge panel could also include a set of images for albums released by a Famous Actor. Those images may be obtained from an online music database or an online music service and include a link to the source of the images. A table may be used to provide information about movies, albums, or songs related to a Famous Actor. The knowledge panel could also include a set of images for related search queries. Those images could each include a link that initiates a search for the subject of the image.

Deciding what to show for an Entity?

Google may look at Topicality Scores to decide whether to show a knowledge panel If an entity in a query is most likely to be the entity described in the query, there is more reason to show a knowledge panel.

Google may also consider performance measures, such as click-through rates to decide whether to show more information for an identified entity. This would be especially true if the knowledge panel would make it more likely to help a searcher find the information for which the user is searching for.

The apparatus may also consider the click-through rate for a highly ranked identified search result (such as a top-ranked search result, or one of the top ten search results.)

If the click-through rate for a high-ranked search result is relatively high (such as greater than a threshold or at least a threshold amount greater than each of the other search results,.)

A knowledge panel may not be shown if the content item store does not contain sufficient content to populate a knowledge panel for the entity, e.g., fewer than a threshold number of content items, the knowledge panel apparatus may determine that a knowledge panel for the entity will not be provided in response to the search query.

The apparatus determines to show a knowledge panel, at least in part, on the type of content that is available for the entity. So the apparatus may have templates that include placeholders for certain types of content to include in the knowledge panels. If that particular type of content that is specified by the template is not available for the entity, then the apparatus may decide to not provide a knowledge panel for the entity.

If a determination is made to not provide a panel for an entity, then it may show search results without a panel.

Why Show a Knowledge Panel When An Entity Appears in a Query

If Google can recognize that an entity shows in a query, it can provide more information about that entity that a searcher may not have been aware of including information that fits into a specific query template that may help a searcher understand better what they are looking for.

Google can show disambiguation panels when there is more than one entity that may have appeared in a query, and the possible entities may have been searched for more or less. Google may also decide that one of the entities possibly in the query is much more dominant than another of the entities.

As a searcher, it can be possible to learn more about what you may be searching for by learning from the search results that a search engine might show you. When a search engine decides to show you a knowledge panel you can learn a lot more about the entity in your query. That can make it possible to refine your query or to search for something else than what you may have thought was possibly a helpful query term or phrase. This can be true when the highest-ranking results for a query have low click-through rates, and the KP may show query revisions or suggestions that you can click through to potentially find answers to your query, which would make a KP very useful.

Last Updated 2/18/2021

Sharing is caring!

17 thoughts on “How is a Knowledge Panel for an Entity Triggered?”

  1. Thank you Bill 🙂 I feel a teeny bit smarter having read this – I’ll need to read it a couple of more times, as I’m sure I’ll pick up even more 🙂

  2. Hi Andy,

    Happy that you like this post. I’ve been following along with knowledge panels at Google since they started appearing, and trying to learn more about them. This one adds information about disambiguation panels and dominant entity panels. and tells us that information in a panel isn’t limited to places like Wikipedia. It’s interesting seeing them appearing the way that they are.

  3. this post. I’ve been following along with knowledge panels at Google since they started appearing, and trying to learn more about them. This one adds information Thank you

  4. trying to learn more about them. This one adds information about disambiguation panels and dominant entity panels. and tells us that information feel a teeny bit smarter having read

  5. Thanks Bill. Summarising, (in my simplified world): Google has a patent that triggers Knowledge Panels when they are relevant to the query!?
    Are you sometimes amazed some of these patents get past the US Patent Office’s receptionist, let alone peer scrutiny? 🙂

  6. Hi Dixon,

    There is a way to look at the dockets for filed patents in the PAIR database on the USPTO website where you can read final and non-final rejections of patents and amendments that are made to filings, and I will sometimes look up some patents to see how they may have been updated and changed and what patents they may have been infringing when they were first filed. IOt’s good to see what a prosecuting attorney at the USPTO might have said about a patent, and is worth taking a look in some cases.

  7. Hello Bill. Nice post. The post you have written on the topic is very informative. Something which, I was not familiar about this feature and thanx to you to let me know about this.
    keep writing this type of post.

  8. I’ve been following along with Google Knowledge panels since they started appearing and trying to learn more about them.

  9. Hi Bill, I am curious to know what determines the click-through rate threshold(number) that satisfies the trigger of KP. Thank you for sharing as always.

  10. Hi Amine,

    The patent does not tell us this information, so unless we somehow gain inside information from Google, it is unlikely that we will find out.

  11. Hi Bill,
    Thank you so much for sharing such informative article. I’m amazed because this “Knowledge Panel” was new to me. I was then wondering how search engine process the information, it seems pretty complicated. my level of knowledge is like a baby, I don’t know that much about this industry but am willing to learn.

    Thank you so much Bill,
    Chris

  12. Hello Bill,
    I have been following you for a long time, everything is going well after following your articles. keep it up. Thanks

  13. Thank you so much for sharing such informative article. I’m amazed because this I was then wondering how search engine process the information it seems pretty complicated. my level of knowledge is like This

  14. Thank you so much Bill for sharing this post! I didn’t quite understand how knowledge panels work, but I really appreciate how you broke things down here.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.