Relevance-Ordered Categories of Information in Search Results

Relevance Ordered Categories for Search Results on Mobile Devices?

As a reporter for the Daily Planet, Clark Kent most likely searches for news-related information. His Alter Ego and Super Hero, Superman, likely searches for comics-related material such as a kryptonite protection suit.

Imagine a search engine that provides relevance-ordered categories of information to searchers in search results that may use profiles for different searchers. Google has been exploring this concept for at least the last dozen years, having filed an early version of a patent application covering this topic in 2008.

Related Content:

• Technical SEO Agency
• Shopify SEO Services
• B2B SEO Services
• Lawyer SEO Service
• SaaS SEO Agency

A patent about relevance-ordered categories of information in search results was granted to Google in the past week. It describes how Google may provide search results on mobile devices in relevance-ordered categories.

The patent tells us about what the inventors believe are the expectations of searchers on mobile devices:

They expect to have access on the road, in coffee shops, at home, and in the office through mobile devices to information previously available only from a personal computer that was physically connected to an appropriately provisioned network. They want news, stock quotes, maps, and directions, and weather reports from their cell phones; email from their personal digital assistants (PDAs); up-to-date documents from their smartphones; and timely, accurate search results from all their mobile devices.

The phones of 2020 are very different from the mobile devices of 2008. To get an idea of how Google viewed mobile devices back then, I would recommend this whitepaper from the Google Mobile research team: Computers and iPhones and Mobile Phones, oh my!, A logs-based comparison of search users on different devices, by Maryam Kamvar, Melanie Kellar, Rajan Patel, and Ya Xu.

We don’t know if the process described in this just granted patent is one that Google will pursue, but it has value in how it describes how Google attempted to overcome problems with a search that is perceived when it was first filed. Keep in mind that Google has been working on entity-based indexing using the knowledge graph, and mobile-first indexing to make websites better experiences for searchers on mobile devices since 2008.
.

The newly granted patent tells us that there may be some problems with providing the same rich range of results to people on mobile devices as to people using desktop computers. These problems include:

1. Input capabilities may be more limited in a mobile device than in a desktop computer, requiring more effort by a searcher to enter a search query (or other information)
2. Displays in mobile devices tend to be smaller than displays in desktop computers. It may not be possible to display as much information at any given time on a mobile device
3. Data connections between mobile devices and the Internet may be slower than between desktop computers and the Internet

The solution described in the patent to address those problems involves providing relevance-ordered categories of information to a searcher.

We are told that one or more categories of information (e.g., web information, image information, news information, navigational information, etc.) may be provided to a searcher in response to a query. This drawing from the patent gives us a sense of how Google might provide results using relevance-ordered categories of information in search results:

Bringing Relevance-Ordering to Categorized Search Results

The first category of the information shown may be selected based on a prediction of the user’s category of information is likely seeking.

That prediction may be made using rules from a machine learning system, trained using historical search data.

The prediction may be made based on statistical correspondence that has been developed through analysis of other similar queries between the query received and a specific category of information.

The prediction may also be based on a profile corresponding to a searcher profile associated with the query (Clark Kent or Superman). Finally, the prediction could be based on aggregated information across multiple searchers with or without data about a particular user.

The prediction may also be based on a combination of factors, including, for example, the statistical correspondence between the received query and a specific category of information and a user profile associated with the query.

The process behind the patent might work as follows:

1. Receiving from a mobile device a query
2. Generating a number of different category-directed result sets for the query
3. Determining an order for the number of category-directed result sets based on the search
4. Transmitting the number of category-directed result sets to the mobile device in a specific order

It could also involve formatting the number of category-directed results displayed in a tabbed array to decrease the correlation between each category-directed result set and the query.

Determining that order may involve calculating each of the category-directed results sets a likelihood that representing those particular categories of results set is responsive to the received query.

Calculating the likelihood value can include:

1. Retrieving a profile associated with the mobile device, including information about the distribution of previously determined correlations between other queries received from that device and one or more different categories of information.
2. Retrieving data associated with queries received from other mobile devices that are substantially similar to the received search query, including distribution of previously determined correlations between the other substantially similar search queries and one or more different categories of information
3. The distribution can include multiple sub-distributions, each sub-distribution being related to any one or more of (a) classification of a device from which the query was received, (b) a model or model group of a device from which the query was received, (c) a geographic area from which the query was received, and (d) approximate time of day at which the query was received
4. Retrieving a profile associated with the remote device and performing a calculation to obtain the first result based on a portion of the retrieved profile, retrieving data that is associated with the query, and performing a second calculation to obtain a second result based on a portion of the retrieved data, and performing a third calculation based on a weighted contribution of the first result and the second result

The search query can be received from a mobile device.

The different categories can include three or more categories selected from the group consisting of:

• Location-based results
• Web results
• Images
• Video
• Shopping
• Blogs
• Maps
• Books

In some aspects, the results ranker is configured to order categories based on

a) a profile associated with the remote device and
b) relevance data correlating other search queries received from other remote devices and one or more categories of information, the other search queries are substantially similar to the search request.

This relevance-ordered search results patent can be found at.

Providing relevance-ordered categories of information
Inventors: Yael Shacham, Leland Rechis, Scott Jenson, and Gabriel Wolosin
Assignee: Google LLC
US Patent: 10,783,177
Granted: September 22, 2020
Filed: June 20, 2011
Abstract

A computer-implemented method is disclosed. The method includes receiving from a remote device a search query, generating a plurality of different category-directed result sets for the search query, determining an order for the plurality of category-directed result sets based on the search query, and transmitting the plurality of category-directed result sets to the remote device, in a manner that the result sets are to be displayed in the remote device in the determined order.

Take Aways From this Relevance-Ordered Search Results Patent

When I came across this patent, I was excited to share it because it paints a picture of what Google could have given us. It reminds me of a lot of the Universal search that we used to have on desktop computers that showed us a mix of local results, news results, image results, and video results. I wrote about the many updates to Universal Search in the post Google’s New Universal Search Results.

Google has provided more and more knowledge-based results to even mobile search results, with knowledge cards, featured snippets, PAA questions, entity carousels. The 10 blue links of the early 2000s are very different from the search results of today. So the problems this patent was intended to address on mobile devices may be no longer quite the problems they were back in 2008. Because of initiatives such as mobile-first indexing and entity-based search, we may not benefit from relevance-ordered information in search results that the inventors behind this patent may have thought we would. But there is value in considering other approaches that Google could have taken.

Ideally, your website should be easily indexable by both Googlebot Desktop and Googlebot Smartphone, and it displays well on mobile devices in a mobile friendly manner. And on top of that, if it is responsive to your audiences’ informational and situational needs, it will stand a good chance of ranking for the searches they perform on their mobile devices.

I do question whether Google will adopt its relevance-ordered categories approach to search results. They could, but they may not at this point. If Google comes up with a newer version of this patent that describes how knowledge-based results might fit into search results, then maybe we will see this used. I will keep an eye out for that.

About Bill Slawski

With more than 26 years of SEO experience and a Juris Doctor Degree, Bill Slawski is the foremost expert on Google’s patents as related to SEO. Patent Exploration is one of the quickest and most detailed ways to find new information about SEO. Bill is the Editor of SEO by the Sea, a prominent search engine optimization blog, where he is the author of over 1,300 posts. Bill’s experience includes Fortune 500 brands and some of the largest websites in the world. Bill is a contributing author for Moz, Search Engine Land, and Search Engine Journal. In 2014-2021, he spoke at industry-leading international conferences about topics including search engine algorithms, universal and blended search, personalization in search, search and social, and duplicate content problems, structured data, and schema