Google Search Patents 2020: The Mega-Post Roundup

Hello and welcome to our great, big, mega-post round-up of Google Search patents from 2020.

Throughout the year, I spend time collecting and writing about Google Search patents that are of interest to the SEO community. I also write about the interesting ones every now and again here on Search Engine Journal.

Now that we’ve made it to 2021 (thank goodness), I wanted to share the entire 2020 collection with you.

They’re grouped to make things easier and so that if you’d like to do some research in a particular area, you can dig in.

And that’s the thing, really — nobody reads these for fun! Well, I do, but I’m strange (so the missus tells me).

It’s all about trying to learn more about various aspects of information retrieval and Google’s algorithm. I hope you find it useful.

Google Assistant

• Implementations for voice assistant on devices
• Proactive virtual assistant
• Invoking an automated assistant to perform multiple tasks through an individual command
• Tailoring an interactive dialog application based on creator provided content
• Automated assistants that accommodate multiple age groups and/or vocabulary levels
• State-dependent query response
• Detecting and correcting potential errors in user behavior
• Systems, methods, and apparatus for providing image shortcuts for an assistant application
• Virtual assistant generation of group recommendations
• Determining state of automated assistant dialog

Authorship

• Generating author vectors

Conversational Search / Voice

• Biasing voice correction suggestions
• Speech recognition with attention-based recurrent neural networks
• Speech endpointing based on word comparisons
• End of query detection
• Search-based dynamic voice activation
• Word-level correction of speech input
• Systems and methods for recognizing user speech
• Customizing actions based on contextual data and voice-based inputs
• Search-based dynamic voice activation
• Word-level correction of speech input
• Systems and methods for recognizing user speech
• Speech recognition and summarization
• Speech recognition with parallel recognition tasks
• Detecting and suppressing voice queries
• Automated speech recognition using language models that selectively use domain-specific model components
• Recognizing accented speech

Discover

• Selecting, ranking, and/or presenting microsite content

Entities

• Entity-based searching with content selection
• Structured entity information page
• Entity display priority in a distributed geographic information system
• Surfacing live events in search results
• Identifying a level of relevancy of a keyword cluster related to an event category for a given time period relative to the event
• Entity disambiguation in a mobile environment
• Selecting textual representations for entity attribute values
• Generating ranked lists of entities
• Modeling personal entities on a mobile device using embedding
• Querying a data graph using natural language queries

Image Search

• Architecture for responding to visual query
• Digital supplement association and retrieval for visual search
• Systems and methods for locating image data for selected regions of interest
• Efficient image analysis
• Image search query predictions by a keyboard
• Images for query answers

Indexing

• Template-based structured document classification and extraction

Knowledge Graph

• Computerized systems and methods for enriching a knowledge base for search queries

Local

• Three-dimensional annotations for street view data
• Attaching visible networks for synchronous local search results
• Suggestion refinement
• Image retrieval with deep local feature descriptors and attention-based keypoint descriptors
• Attaching visible networks for synchronous local search results
• Automatic continued search
• Generation of enhanced search results

Links

• Personalization of web search results using term, category, and link-based user profiles
• Generating textual summary about physical location based on computing interactions pertaining to physical location
• Geocoding personal information

Machine Learning

• Multi-task learning using knowledge distillation
• Training distilled machine learning models
• Regularization of machine learning models
• Machine learning system and method of classifying an application link as broken or working
• Personal search result identifying a physical location previously interacted with by a user
• Selecting content items using reinforcement learning
• Searchable index
• Machine learning to identify opinions in documents
• Searchable index
• Robotic grasping prediction using neural networks and geometry aware object representation

Mobile

• Navigation queries
• Search engine

NLP

• Generation of a topic index with natural language processing

Neural Networks

• Training a document classification neural network
• Complex evolution recurrent neural networks
• End-to-end text-to-speech conversion
• Category learning neural networks
• Processing images using deep neural networks
• Skimming text using recurrent neural networks
• Neural network processor
• Generating larger neural networks
• Determining orders of execution of a neural network
• Robotic grasping prediction using neural networks and geometry aware object representation

Personalization

• Detecting and correcting potential errors in user behavior
• Personal search result identifying a physical location previously interacted with by a user
• Discovery of unique entities across multiple devices
• Dynamic and personalized filtering of media content
• Personal search result identifying a physical location previously interacted with by a user
• Personalized search results
• Personalization of web search results using term, category, and link-based user profiles

Predictive

• Predicting user needs for a particular context

Query Classification

• Query modification based on non-textual resource context
• Generating elements of answer-seeking queries and elements of answers
• Classification of offensive words
• Query suggestion templates
• Sentiment-based classification of media content

Ranking / Scoring

• Contextual ranking of photos and text in search
• Selectively modifying scores of youth-oriented content search results
• Scoring content based on social interaction
• Contextual ranking of photos and text in search
• Selectively modifying scores of youth-oriented content search results
• Information ranking based on properties of a computing device
• Detection of popular content with narrow appeal

Recommendation Engine

• Query refinements using search data
• Query completions

Searching

• Identifying teachable moments for contextual search
• Navigation system and methods for generating enhanced search results
• Determining search queries for obtaining information during a user experience of an event
• Performing multiple related searches
• System and method for improving access to search results
• Methods, systems, and media for presenting related media content items
• Synthesis of webpage snippets using sub-pages of the webpage
• Determining a set of steps responsive to a how-to query
• Enhancing sitelinks with creative content
• Television related searching
• Search engine
• Answer facts from structured content
• Dynamic determination of filters for flight search results
• Generation of enhanced search results

Semantic

• Semantic model for tagging of word lattices
• Identifying information using referenced text

Social Networking

• Determining a quality score for a content item

Spam

• Detection of malware features in a content item

Temporal

• Temporal content selection

Vectors

• Generating vector representations of documents
• Learning highlights using event detection

Videos

• Providing an indication of highlights in a video content item
• Automatic detection of zones of interest in a video
• Generating theme-based videos
• Displaying information related to content playing on a device
• Methods, systems, and media for providing video content suitable for audio-only playback
• Detecting similar live streams ingested ahead of the reference content
• Conditional display of hyperlinks in a video
• Detecting multiple parts of a screen to fingerprint to detect abusive uploading videos
• Relevance-based image selection
• Methods, systems, and media for providing video content suitable for audio-only playback
• Methods and systems for person detection in a video feed
• Detecting actions to discourage recognition
• Systems and methods of person recognition in video streams
• Learning highlights using event detection
• Detection of popular content with narrow appeal

And there we have it.

Whew, what a workout!

Next up, I’ll be sharing some of the odd, non-search related patents Google was awarded in 2020. Just for fun.

Then we’re back on with the new Google Search Patents for 2021.

Stay geeky out there!

More Resources:

• 3 Latest Google Patents of Interest – October 13, 2020
• Google Search Patent Update – December 18, 2020
• A Rundown of Google Patents from the First Half of 2020

Image Credits

Featured image by author, January 2021