Recent Updates From Google Patents
Many SEO specialists are carefully analysing new Google Search-related patents that were recently updated. These updates help Google SEO specialists develop new strategies that could enhance their optimisation strategy and help them achieve better rankings on the search engine results pages (SERPs).
Google has been a bit quiet when it came to updating search patents these past few weeks, but recently, two search patents were updated.
The first patent is about querying a data graph using natural language queries, which deals with graphs, entities, and semantic elements. It was filed back in 2013, and it is interesting for the SEO agencies of today because back then, even the experts did not know anything about data graphs and the like.
As such, the topic was rarely brought up in most SEO conversations. Some experts – even to this day – do not get how Google deals with semantics.
The patent discusses how graph data and entity relations were assembled together manually in the past, as well as how they seek to automate this with machine learning. It said that the data graph and entities, such as things, places, people, and concepts, may be stored as nodes.
The relationship between these nodes is indicated by the edges in between. For instance, in a data graph, the nodes of “Maryland” and “United States” are linked by the edges of “in country” or “has state”.
The data that can be extracted from the data graph and text is used as input for the machine learning algorithm, training it to predict the tuples for the data graph. This machine then may create multiple weighted features, with each of them representing the relationship between two entities.
There are implementations that answer natural language queries from the data graph. In this situation, the machine is trained to map features to queries to give the user the best answers possible. The training may involve the use of positive examples from question results or search records that come from document-based search engines.
The machine then creates several weighted features, with each of them representing one possible answer that’s represented by a path in the data graph.
The second patent has something to do with search and retrieval of structured information cards. It was explained how information cards, entities, structured data, and knowledge bases work together. This tells Google SEO specialists how much SEO has changed a lot after many years.
It says that methods, apparatus, systems, and computer programs are encoded on computer storage medium. By doing so, they can facilitate the identification of additional trigger terms for structured information cards.
The method involves accessing data with a template to present structured information. This accessed data will reference two things: a label term and a value.
Some actions might involve acquiring a label term. Other actions may identify one or more entities that are associated with the label term or the candidate label term.
What’s more interesting about the patent is that the card trigger-term identification unit adjusts the grammar over time, which is done by evaluating candidate terms. For instance, the grammar for a “Movie” structured information card is relevant to “movie ticket confirmation”, “movie time”, and “ticket confirmation number”.
The card trigger term identification unit will analyse these terms to identify an additional trigger-term for the “Movie” structured information. Other questions that have received the term “movie ticket” or “movie time – or both – will cause the “Movie” structured information card to be shown as an answer to the question.
Studying Google patents is one great way to gain insights into how Google algorithms really work. This allows SEOs to formulate better strategies to make their businesses more successful.
Position1SEO is composed of an experienced team of SEO specialists who are highly knowledgeable about modern SEO. Not only that, but we are highly flexible and constantly adapt our strategies to suit the every-change Google algorithm.
If you want to work with the best professionals in the industry, you can call us via 0141 404 7515 or email email@example.com.