Google Patent Aims To Solve Searchers’ Need For Related Media Content

There’s a newly granted Google patent out that entails methods, strategies, and media for presenting content material materials organized by class.

This patent caught my consideration as a result of it talks about related entities and describes how they match collectively on this planet of media.

This advances the idea of performing queryless searches, a trademark of Google Discover, for media content material materials equal to TV displays, movies, and additional.

What may this indicate for searchers — and for search engine optimisation professionals? Let’s take a look.

The Challenges of Meeting Searcher Interest in Related Media

Searchers normally want to find and have a look at media content material materials related to a particular topic.

For occasion, if a person is enthusiastic a few positive television current, they could be captivated with:

• Viewing further or utterly completely different episodes of the television current.
• Video interviews with actors from the television current.
• Episodes of various television displays that embrace the an identical actors.
• Photos of the actors, and so forth.

This patent tells us that it might be troublesome to find out, arrange, and present related content material materials to a searcher. The drawback is more likely to be in getting a computer to affix that knowledge.

The patent makes an try to treatment this by providing new strategies and media for presenting content material materials organized by class.

“Methods, methods, and media for presenting content material organized by class” begins by displaying an informational need that many people have, then describes a solution to that disadvantage in a patent granted on June 15, 2021.

Presenting Content for Related Entities Organized by Categories

The patent displays strategies of presenting content material materials organized by class.

Here’s the best way it really works:

• A search engine receives a request for an internet web page with content material materials a few first entity.
• A main group of related entities is acknowledged based mostly totally on searcher interaction knowledge.
• Quite a number of properties regarding the main entity are acknowledged.
• A second group of entities is acknowledged for each of those properties.
• Each of those properties receives a ranking based partly on the first group of entities (acknowledged based mostly totally on the searcher interaction knowledge) and the second group of related entities for the property.
• The ranking signifies a chance {{that a}} searcher of the searching system could be captivated with content material materials regarding the property.
• The search engine determines a subset of the properties inside the group of properties based mostly totally on the ranking of each one.
• It identifies a gaggle of content material materials objects for each of the properties.
• A subset of content material materials objects from the group of content material materials objects is determined.
• The content material materials objects are confirmed on the computing system.

Here’s what that may seem like in observe, based mostly totally on the above:

This Google Patent Suggests a Media Knowledge Graph

To a stage, this patent is describing a knowledge graph of related entities and attributes with out saying the phrase “information graph.”

But that’s what me in it — seeing the best way it appears to be for related entities, properties and attributes, content material materials objects, and the best way all of the issues is more likely to be linked is on the coronary coronary heart of this technique.

A few technical observations of bear in mind:

• Calculating the ranking for properties may indicate calculating a proportion of entities included inside the second group of entities included inside the first group of entities.
• The subset of content material materials objects can be determined based mostly totally on the popularity of each content material materials merchandise inside the group of content material materials objects.
• The system may also indicate selecting, for each property inside the subset of properties, a content material materials type based mostly totally on the popularity of content material materials objects associated to the property and the content material materials type, the place indications of the subset of content material materials objects are launched in accordance to the content material materials type and the property corresponding to each content material materials merchandise.
• At least one content material materials merchandise included inside the subset of content material materials objects can embrace the aggregation of plenty of films.
• The first entity would be the determine of a character in a media content material materials merchandise.
• The group of properties can embrace a property similar to completely different media content material materials the character has appeared in.

For each property inside the group of properties, determining the second group of entities similar to the property incorporates means for accessing a database that signifies an interconnection of each entity inside the second group of entities to the first entity by the property.

This is a technique of categorizing media content material materials spherical related entities.

This patent can be found at:

Methods, systems, and media for presenting content organized by category
Inventors: Alexander Pak, Felix Raimundo, Sertan Girgin, Raphael Marinier, and Vincent Simonet
US Patent: 11,036,743
Assignee: Google LLC
Granted: June 15, 2021
Filed: May 23, 2016

Abstract:

Methods, strategies, and media for presenting content material materials organized by class are supplied. In some embodiments, the technique incorporates: receiving a request to present an internet web page indicating content material materials related to a main entity; determining a main group of entities related to the first entity based in any case partly on searcher interaction knowledge; determining a gaggle of properties similar to the first entity; determining, for each property, a second group of entities similar to the property; calculating, for each property, a ranking for the property; determining a subset of the properties inside the group of properties based mostly totally on the ranking of each property; determining, for each property inside the subset of properties, a gaggle of content material materials objects similar to the property; determining a subset of content material materials objects from the group of content material materials objects; and inflicting a searcher interface indicating the subset of content material materials objects to be launched.

Mechanisms for Presenting Media Content Organized by Category Are Provided

The objects described inside the patent can decide content material materials related to a particular topic and determine whether or not or not the acknowledged content material materials might be going to be fascinating to a searcher using a combination of plenty of sources of information.

For occasion, the sources of information can level out topics related to a main topic, the place the relatedness of two topics is inferred based mostly totally on searcher habits:

• Search queries of a searcher.
• Content that searchers normally view within the an identical viewing session.
• Any completely different applicable form of searcher habits.

The sources of information may embrace structured knowledge that signifies properties that relate to topics.

In this occasion, the structured knowledge may level out that two movies are related to at least one one other on account of they perform a particular actor (whom Google acknowledges as an entity in every).

We may then see knowledge indicating that two content material materials objects are related, inferred based mostly totally on searcher actions, with structured knowledge indicating a trend by which two content material materials objects are related.

The mechanisms can then set off content material materials acknowledged to be associated to a particular topic to be launched in a searcher interface that can group the acknowledged content material materials based mostly totally on a property that relates the acknowledged content material materials to a particular topic.

A bunch of properties may be associated to a main topic or entity and a gaggle of topics or entities associated to each property.

A property may level out any applicable topic or class associated to a particular topic or entity by way of which that particular topic or entity is expounded to completely different topics, entities, or completely different knowledge.

More Media Examples

For occasion, the place the first entity is a fictional character, the group of properties may embrace:

• Movies by which the fictional character has been portrayed.
• Actors who’ve portrayed the fictional character.
• Other applicable properties.

Continuing with this occasion, a gaggle of related entities similar to the property of flicks by which the fictional character has been portrayed can embrace each movie’s names.

The mechanisms can then decide content material materials objects related to the entities inside the group of related entities. It might set off indications of these content material materials objects to be launched in a searcher interface organized by property.

For occasion, in instances the place the entities inside the group well-known entities embrace names of flicks by which the fictional character was portrayed, the content material materials objects can embrace:

• Clips from the flicks.
• Soundtracks from the flicks.
• Other applicable content material materials objects.

A search system may decide properties associated to the first entity using any applicable strategy or combination of methods.

A candidate group of properties can be acknowledged, as an example. A subset of the candidate group of properties can be acknowledged based mostly totally on how associated each property is to the first entity, which is based on how fascinating each property might be going to be to a searcher.

In one different occasion, the place the first entity is a fictional character, the candidate group of properties may embrace:

• Movies by which the character was portrayed.
• Actors who’ve portrayed the character.
• A fictional birthday of the character. Continuing with this occasion, in some embodiments, the mechanisms described herein can determine that the properties of flicks by which the character was portrayed and actors who’ve portrayed the character are seemingly to be basically essentially the most fascinating to a searcher

Following the an identical occasion, the content material materials corresponding to each property may be acknowledged using any applicable knowledge and methods.

The mechanisms may decide the content material materials objects based mostly totally on the popularity of plenty of content material materials objects associated to each chosen property.

This Patent Points out a Wide Range of Types of Content

Content objects may be any applicable sorts of content material materials, along with:

• Video content material materials.
• Audio content material materials.
• Television functions.
• Movies.
• Live-streaming content material materials.
• Audiobooks.
• Documents.
• Webpages.
• Other applicable sorts of content material materials.

The content material materials objects may also embrace compilations and/or aggregations of plenty of content material materials objects.

The content material materials merchandise may be a playlist of content material materials objects which could be launched in a particular order. They can be a channel of content material materials associated to a particular topic and/or content material materials creator.

The Search Interface for Presenting Content Organized by Category

This is what makes this media-based data graph browseable to find relationships and connections between characters and tales.

The searcher interface can embrace:

• Title.
• Categories.
• Content options all through the category.

The Title can be any applicable title and may level out:

• An entity or a topic, equal to a character (fictional or non-fictional).
• An event.
• A television program.
• A e-book.
• An curiosity or curiosity.
• Any completely different applicable topic.

The title can embrace any applicable textual content material, pictures, graphics, animations, hyperlinks, or completely different applicable content material materials.

The patent tells us to bear in mind that the location of the title all through the searcher interface could be confirmed for instance and that the title can be launched at any applicable location. The title might even be omitted.

The searcher interface may embrace applicable courses related to an entity or topic similar to the searcher interface.

They current an occasion: If the searcher interface is expounded to a particular fictional character (e.g., “Superhero X”). The courses may embrace movies by which the fictional character was portrayed (e.g., “Movie 1”), an actor who has portrayed the fictional character (e.g., “Actor 1”), or another applicable courses.

Note that the patent tells us that any applicable number of courses can be included.

Each class launched inside the searcher interface may embrace indications of media content material materials objects similar to that class.

For occasion, if the category is from a particular movie, then the content material materials recommendation may be a playlist of songs and/or music films included in a soundtrack of the movie.

The content material materials objects might very properly be:

• Videos (e.g., scenes from a movie, interviews with actors from a movie, a trailer, and/or another applicable films).
• A set of media content material materials related to the category (e.g., a playlist of songs, a playlist of flicks related to the category, a channel associated to the category, or another applicable media content material materials).
• Links to web pages associated to the category.
• Images.
• Or another applicable sorts of content material materials.

The patent moreover asks us to bear in mind that each class may embrace any applicable number of content material materials objects.

Categories inside the searcher interface and content material materials objects in each class can be acknowledged using any applicable strategy or combination of methods. Those courses may be based mostly totally on topics or entities associated to a topic.

Or, the content material materials objects launched in each class may be based mostly totally on the popularity of the content material materials objects.

Presenting Content Organized by Categories Overview

I believed this was fascinating on account of it attracts collectively related entities and topics to allow searchers to find these relationships and connections.

Before Google launched the knowledge graph, they’ve been engaged on a Browseable Fact Repository. We are returning to the aim the place they should current us the connections in a topic like media (which can attraction to quite a few consideration).

It covers quite a lot of media and leisure, which a group of Google patents have lined so far. However, they do have a handful of media consumption histories, the place they purchase particulars concerning the media that individuals devour as part of our every day lives.

Google has been putting quite a few effort into fleshing out a media data graph and content material materials objects for utterly completely different related entities to find new episodes, characters who act collectively, assortment of displays, awards, prizes, producers, directories, and others who work on media.

The attainable courses that this patent can cowl can embody an intensive fluctuate. In addition, the media examples it provides can current the best way it might be a part of quite a few related entities.

More Resources: