Google patents provide essential information about the company’s most recent developments and top goals for search technology advancement.

This article examines nine intriguing Google patents from 2023 and considers how they might affect SEO going forward.

 

Are Google’s actual practices reflected in the patents?

 

The fact that Google submits and publishes a patent application does not ensure that the techniques described will be included in Google Search.

You can verify whether a methodology or technology is compelling enough for practical application by Google by looking up its patent status in the United States and other nations.

It is necessary to file a claim for patent priority in other nations within a year of the original filing.

Analyzing Google’s patent portfolio is beneficial even in cases where a patent has no practical application. It offers perceptions into the topics and difficulties Google’s product developers concentrate on.

 

1. Filters for search results from resource content

 

Google Search is becoming even more intelligent by adding more filters to focus results. This new patent might serve as the basis for filter techniques.

The patent describes a system that dynamically creates search query filters based on the content of resources (like webpages) pertinent to a user’s query to improve search experiences. This strategy aims to increase the variety and relevancy of search results.

 

2. Assessing a search query’s interpretation

 

For search engines, understanding the meaning and intent of a query is essential. This patent may be incorporated into the process.

One noteworthy mention of BERT (Bidirectional Encoder Representations from Transformers) in the patent raises the possibility that this methodology has a bearing on using BERT in search algorithms.

 

3. Offering search outcomes predicated on a compositional query

 

Google is gradually transforming Search into an entity-oriented search platform. Delivering results that are pertinent to entities is, therefore, essential.

This patent might be a crucial component in our understanding of entities and their interactions.
A method for providing search results based on compositional queries is described in the patent. This approach consists of:

Identifying the various entity types and how the query relates to them.
Identifying nodes in a knowledge graph by pinpoints.
Evaluating attribute values to ascertain the entity references that are produced.

 

4. Putting knowledge panels in context

 

The knowledge panel is the window into the Google knowledge graph and the stored entities.

It is essential to provide accurate and pertinent information based on entities. These panels offer a comprehensive information source because they are integrated with standard search results. Techniques for completing this task are covered in this patent.
The methods, systems, and equipment covered by the patent are centered on improving search engine results by integrating knowledge panels that offer contextual information relevant to search queries.

These knowledge panels are generated Based on identifying entities (such as singers, actors, and musicians) and context terms found in user search queries.

 

5. Document activity log systems and techniques to train machine learning models for document relevance

 

When fine-tuning its machine learning algorithms, Google uses user logs and engagement as critical sources of information, which determine search result rankings. Methods for completing this task are described in this patent.

The systems and procedures for using document activity logs to train a machine-learned semantic matching model to assess document relevance are described in the patent.

This method works exceptionally well in environments with restricted access to content or user interaction data, such as cloud or private document storage.

This approach is helpful when conventional data sources, such as user interaction data or complete document content, are not available.

 

6. The system for composing queries

 

Contextualized search results are becoming more common. Search results and user experience improve when the user’s context and query are recognized. One piece of the puzzle to solve this problem is this patent.

The patent’s primary focus is the methods, systems, and equipment for producing data describing context clusters and context cluster probabilities. These clusters are created based on the query inputs and the context connected to each query input.

The system that uses context clusters to streamline the search query process is described in the patent. The context and content of earlier queries are used to create these clusters.

Upon starting a search, the user can choose a query without typing because the system displays pertinent context clusters.

 

7. Adding search query parameters together that are related to the same topic of study

 

This patent, which focuses on improving search query processing, demonstrates how significant a user’s unique context is to Google.

The method it presents allows for creating a combined search query from the parameters of an existing query and one or more prior queries from the same user, as long as the queries have a similar line of inquiry.
The method described in the patent combines several related search queries into a single, more efficient query, thus streamlining online search experiences.

This method uses user interaction and semantic analysis, which may lessen the duplication of search results and increase the relevance of the retrieved information.

 

8. Displaying details about search results

 

This patent initially looks confusing because it discusses using the user’s device’s content, markups, and annotations. Above all, though, it demonstrates how highly customized search results could be offered by search engines like Google.
The technique for displaying computer-generated search results is the main focus of the patent. It includes:

Obtaining a request for a search.
Recognizing several search results.
Using information from one or more web notebooks, rank these results.
Giving these results in order of presentation.
The patent outlines a process for using web notebook content to improve search result relevancy and accuracy.

 

9. Extracting multiple sources and rating brief query responses

 

The number of direct answers that appear in the SERPs is steadily rising. A few examples are the data output directly from the Knowledge Graph highlighted snippets and the reactions in the SGE Snapshot AI boxes. This patent describes how to generate and choose these kinds of straightforward responses.

The goal of the patent is to raise the caliber of the quick responses that search engines offer. Instead of depending on one top-ranked search result, it presents a method for creating and rating these succinct responses based on various sources.

A technique to improve the dependability and precision of brief responses in search engine results is described in the patent.

 

The role that patents play in SEO

 

Many people improve their websites based on “hacks” they discover on YouTube, blogs, social media, and other online platforms. Still, they need to grasp the underlying concepts of search engine optimization genuinely.

I advise anyone interested in SEO to become familiar with information retrieval, crawling, and indexing fundamentals.

The following action is to comprehend the foundations of:

The technology used by modern search engines.
Entities and Semantic Search.
Natural language interpretation.
Insertions.
With knowledge of scientific principles and technology comprehension, pursuing practical experience will frequently lead us to evaluate things subjectively.
Studying Google patents makes sense regardless of whether a patent is implemented because you can gain insight into the problems and difficulties that Google’s product developers face.

If you still need help and clarification, look at our monthly SEO packages and get professional assistance.