- What is this new search intent patent about?
- Predicting intent in context-aware searches
- Expand intent in search results
For many years I have looked for sources of information that discuss the intent of search queries. I pay attention to patents from search engines and what search engineers have said about patents as well.
So if a newly granted patent from Google is called "Predicting the intent of a search for a particular contextThe authors of the patent caught my attention. This is a patent that was granted in February 2021.
My interest in these search intent-based patents has led me to post such as "A Well-Formed Query Helps Search Engines Understand User Intent In The Query And Why Do Google Users Google?" To compose. Understanding user data for measuring viewfinder intent.
This has also been the subject of presentations I have seen at search conferences. The moderators of these sessions insisted that an SEO professional look at the SERPs that the desired keywords lead to
Read on below
These speakers recommend that you look at the intent behind the pages for the keywords you selected to get an idea of how other people have optimized the page ranking for specific keywords.
What is this new search intent patent about?
The predictive intent of a search for a particular contextual patent begins with a searcher being able to use a computing device to obtain information and facts that will assist the searcher in completing a particular task.
For a search engine, every search begins with a searcher guessing query terms (or keywords) that are being used to attempt to find information.
This patent tells us that a searcher must provide sufficient information (e.g., search query terms) to guide a computing device to locate the information a searcher is looking for.
We caution you that if a search query is not tailored closely, or if the finder does not provide much additional information beyond the query, a computer may return too much information.
Read on below
This can mean that some of the most interesting or relevant information is difficult for a searcher to find.
As with many patents, the beginning of the description is made to indicate problems that a seeker may have.
These are the problems a seeker may face when trying to find information:
The searcher may be under stress and / or waste valuable time and resources entering very detailed queries into a computing device, causing the computing device to perform multiple searches or to search through large amounts of search results to obtain information necessary to accomplish a particular task are .
Most patents provide algorithms that are designed to provide methods of solving problems that can be identified in the patent, and this gets to the point:
Techniques from this patent can help a computer system predict the search intent of a query for a particular context of a computing device.
A search engine can look at context information to get an idea of the intent behind a query.
The patent gives us a few examples:
- User interest.
- Times of day.
It then tells us that the computer system can then define a relevant context for a search query and, based on that relevant context, intention, or purpose of a search using the search query, predict it in the relevant context.
The patented process can then customize the search results so that the intent information is highlighted over other information returned by the search.
The patent then provides examples to illustrate what they will try:
After a user of a computing device purchases tickets for a particular movie that is shown in theaters, a user can cause the computer system to perform a search using the name of the particular movie as part of a query.
The patented system can receive contextual information, including an indication that tickets have already been purchased for a future showing of the particular film.
In response, the system may conclude (e.g., based on historical data indicating user-initiated actions taken by other computing devices) that the search, including the name of the particular movie, is for a purpose other than purchasing additional tickets .
The system may therefore adjust the search results returned from the search so that movie times are ranked lower than other information (e.g., reviews, memorabilia, trivia, etc.) about the particular movie.
Read on below
The inventors of the patent explain why they used this example of a film to tell us about the context.
They say that by automatically adjusting search results to highlight information that a searcher is more likely to look for in a current context (after purchasing movie tickets), the system can enable searchers to experience less stress and / or none of value Wasting time and resources.
The time it takes to search for information in and between search results (by including the movie name in the query it performs) is reduced.
The patent tells us that using contextual information in the query to get better results based on the intent behind the search ensures that there are no results that provide personally identifiable information.
In simple terms: give people what they want!
And they also take privacy seriously:
For example, the identity of a user can be treated in such a way that no personally identifiable information about the user can be ascertained, or the geographic location of a user can be generalized when location information is obtained (e.g. at the city, zip code, or state level) that a specific location of a user cannot be determined.
Read on below
This patent can be found at:
Predicting the intent of a search for a particular context
Inventors: Eibe Jin Lim, Josep Linn, Yuling Liang, Carsten Steinebach, Wei Lwun Lu, Dong Hyun Kim, James Kunz, Lauren Koepnick and Min Yang
Agent: Google LLC
U.S. Patent: 10,909,124
Granted: February 2, 2021
Saved: May 18, 2017
A computer system is described which, based on user-initiated actions performed by a group of computing devices, determines the intent of a search using a particular search query received from a computing device.
The computer system adjusts at least a certain portion of the search results obtained from the search using the search query based on the intention by highlighting information that satisfies the intention.
The computer system sends a display of the customized search results to the computing device.
Predicting intent in context-aware searches
Google tells us the types of contextual information that can be taken into account when executing queries, and information such as:
Read on below
- Topics of interest to users (e.g., a user's preferred "things", usually managed as a user interest chart or some other type of data structure).
- Contact information associated with users (e.g., a user's personal contact information, as well as information about a user's friends, co-workers, social media connections, family, and so on).
- Browse histories.
- Long and short term tasks.
- Calendar information.
- Application histories.
- Purchase histories.
- Other information that computing devices and information server systems may collect about a user of computing devices.
- Information about the health of a computing device.
- The physical and / or virtual environment of the user and / or the computing device in different places and at different times.
How can a search engine use this context to predict a query that will provide useful and helpful information to a searcher?
This is what the patent indicates:
The prediction module can execute a machine learning model (e.g., a deep learning model) that receives as inputs: a search query (or part of a search query) and a current context that is received by the context module.
The machine learning model may produce as output an indication (such as a label or other identifier) of search intent using the search query for the current context.
The intentions determined by the prediction module can be selected from a group of predefined intentions.
Some examples of pre-defined intentions include transportation or travel-related intentions (e.g., ridesharing, flight status, ticket purchase, flight schedules, and other transportation-related intentions) and entertainment-related intentions (e.g., movie review, show times, ticket purchase, cast), member bio, album, or song -Review, artist bio, artist tour dates and other entertainment related intentions).
Read on below
An intentional score can also be set.
This is what the patent tells us about this score:
The score of intent may indicate a level of certainty (e.g., a likelihood or some other level of likelihood) that the purpose of performing a search using the search query in a current context is to obtain information that includes the intent fulfill.
Of course, we've seen a language about certain thresholds hit with that score.
If the score isn't high enough, the search engine may not adjust search results based on intent.
Conversely, if it is above this threshold, the effect of intent on the query can be much stronger.
The patent contains much more detailed information on how intentions can be used in conjunction with the intent to provide results that meet the situational and informational needs of seekers.
Expand intent in search results
Often times, when search engineers have written about intent and search, they talk about intent in terms such as:
Read on below
An information intent receives the response of a result that returns information about a query, e.g. B. the course of the pizza or a pizza recipe.
A commercial intent receives the response to an outcome that returned an opportunity to purchase something related to an inquiry, e.g. For example, a local pizzeria, a shop selling pizza crust and sauce, and ovens.
A transaction intent receives the response to an outcome that enables a visitor to take an action, e.g. B. Ordering a pizza for home delivery.
A navigational intent receives the response of a result that enables a searcher to visit a page that he may have visited before or that has an idea that it exists. An example of this could be a page that states where Dominos Pizza was founded and who is on the company's board of directors.
Read on below
In conclusion, this patent looks at the context of a search and tries to guess the intent behind that search based on that context.
This may include rewriting the query to include information about that context.
Screenshot by the author, March 2021