Thematic and fan-shaped query output from Google: understanding the nuances of technology

Google is one of the largest and most popular search engines today. In particular, statistics show that this corporation controls almost 90% of all Internet search. The company has been operating in the market for a long time and all this time has put a lot of effort into achieving the result that is currently observed. But today all this is under serious threat. And the main reason for this is the emergence of ChatGPT and other language models based on artificial intelligence. This is what has really changed the search principle itself today: users can receive generalized answers to their queries directly in the search results, without clicking on links, that is, without visiting sites, without generating organic traffic. For people who just want to get their bearings in the matter, this is very convenient, because they will spend much less time searching for the necessary data, but the search engine itself is losing relevance.

All these transformations that have begun to be observed on the market in the last few years have forced the Google search engine to radically reconsider its general policy and go looking for ways to solve the problems that have arisen. Thus, in May of this year, the company launched an innovative search function - AI Mode. It is capable of responding to user queries in a conversational format, that is, without providing a classic list of links. Such a launch was announced at a conference for developers as a tool that would allow competing with more popular and previously existing AI alternatives.

The situation really turned out to be threatening, since many platforms and companies began to implement not Google solutions, but an alternative. In particular, Apple representatives announced that they plan to add the ChatGPT or Perplexity tool to Safari this year, literally putting an end to the use of classic search. And the situation is not saved even by the fact that the AI Overviews Google function launched last year, which covers over one and a half billion users monthly, did not allow Gemini to bypass its more famous competitor ChatGPT.

Standing out in this entire technology race is not as easy as it may seem at first glance. Realizing all these nuances, Google decided to launch radically new technologies. So, along with the AI Mode intelligent search, a patent for thematic search is also under development. In essence, both of these technologies overlap quite seriously, and at the output we get an innovative system for studying broad, deep and complex queries. This patent was submitted for consideration to the relevant authorities back in December 2024 and in essence is also close to the fan query technique that the corporation uses in its solutions based on artificial intelligence.

Within the framework of the patent “Thematic Search” all the nuances of how the search engine will generate answers in AI Mode are disclosed, and several options for launching a new content strategy as a whole are considered. Today, it is not yet possible to say that all the developments have come into force and the audience can use them, but there is a possibility that these high-tech solutions will be launched in the foreseeable future. Therefore, it is important to navigate these technologies, to understand what exactly is being discussed. This is what will allow, so to speak, to keep up with the times.

In today's review, we will talk about what AI Mode is in principle and what is the essence of fan input of queries. We will talk about the features of the patent "Thematic search". We will provide a brief comparison of technologies that may be adopted by the search engine in the foreseeable future. The information presented will allow you to understand what these technologies are and what impact they can have on search in the foreseeable future.

What is Google's AI Mode technology?

Before we get directly acquainted with the features of thematic search, let's talk a little about the technology, which in many ways can be called its basis. In particular, we are talking about AI Mode, which is already available to users in many countries around the world. In the spring of this year, it was launched in the United States of America, and then it began to spread to other countries of the world. With its help, it became possible to conduct a more detailed search in order to get an accurate answer to your own request. Here, users can ask clarifying questions, telling the search engine what kind of answer they expect from it. How to enable AI Mode Google? You don't need to make any settings here. Just pay attention to the search bar itself in the Google application. You should have a new tab "AI Mode". Here you can do direct work.

The developers launched this solution on the market in order to improve the quality of processing complex user queries, including those related to data analysis, comparison of several product items. There are also a number of additional functions that many of you will be able to appreciate and start using in practice:

• Deep search. It is designed to perform hundreds of search queries in conjunction with the analysis of large volumes of data. As a result, it generates an expert-class report, which, among other things, will also include links to primary sources for a more detailed acquaintance. The convenient thing is that you will receive such a report literally a few minutes after entering a request. Thanks to this, time is significantly saved on finding a detailed answer, manual processing of hundreds, or even thousands of pages is excluded.
• Agentic capabilities. This is a functional Project Mariner AI agent that allows you to automatically book tables in restaurants, buy tickets for various events, and also find current events in a particular country in the world. All that is required of you in this case is to simply select one of the proposed options and confirm it.
• Shopping partner. It is clear from the name that this AI tool is directly related to shopping. In particular, here you can upload your photo in order to subsequently try on various clothes virtually, while minimizing the likelihood of buying something that you don’t like. At the same time, this tool can track prices for selected items and help you make a purchase in literally one click.
• Live capabilities. If you have previously encountered a tool such as Google Lens, then here we get its improved version. Now you can literally point your smartphone camera at any object in real time and get the most detailed information about it from artificial intelligence, as well as links to official pages with relevant information for detailed acquaintance, if necessary.
• Personal context. A tool that provides recommendations in response to certain user requests. It will take information for them from previously made searches. Alternatively, it can offer interesting locations, recreation areas, restaurants that this person has visited before. That is, if you like Italian cuisine, you can get a selection of such restaurants in your immediate vicinity.

Along with this, the developer claims that some of the AI Mode functionality will also be directed to AI Overview. Thanks to this, responses to user queries will become more expanded and voluminous. Among other things, the platform will generate tables and graphs, which will significantly simplify the perception and processing of complex multi-level information.

Many of these AI Mode features are also implemented in thematic search. Now we will get acquainted with the patent that is being registered in more detail.

Features of the patent "Thematic search" from Google

Patent "Thematic search" Google describes in detail the thematic search system, within which all related queries related to the main one will be processed simultaneously, and the results will be issued by categories, which are also called topics. Within each such topic, there will be a short summary, which will allow users to find out the answers to their own queries even without visiting sites. That is, there is no need to follow links, spend additional time on this work.

The patent also includes information on the specifics of the system's operation for deep research related to more complex and advanced queries. The innovativeness of this solution is that it will already implement automatic identification of topics by analyzing classic search results using artificial intelligence technology. As a result, informative and convenient summaries on each topic will be created, literally at the level of visual perception. They will include the content and description of the results.

We also suggest studying such a question as «The Impact of Artificial Intelligence on Search Engines», in order to identify certain patterns in the work of search in general, relevant in the era of digitalization.

What is Google's Thematic Search

Artificial intelligence is developing at an accelerated pace today, finding active application in more and more new niches. And in order to survive in a highly competitive environment, Google needed a serious transformation. This is how the concept of launching a thematic search engine appeared. In many ways, there is a return to the earlier stages of organic search. In particular, it is based on large language models, as well as a resume generator.

If you analyze this patent, which is generally publicly available, you can see that it also mentions thematic search engines that will receive a request from a user and pass it on for processing directly to the search engine. As a response, it will receive search results and transform them into separate groups - topics. At the same time, there is active interaction with the classic search engine, but large language models are already involved at the stage of generating a resume for grouped search results. Another distinctive feature is that for one user request, the thematic search engine can send several additional requests to the main search engine, already taking into account the created “subtopics”.

That is, thematic search involves selecting answers based on interests. Here, user requests are processed and analyzed in order to determine their subject matter and ultimately provide the most relevant answers. The system compares the entered query with indexed data, identifying patterns between this query and the context posted on the site pages. Various methods can be used to implement this idea: keyword analysis, machine learning, contextual analysis. It is by combining these tools that the topic of the query is determined and the results are ranked.

If we describe step by step the principle of thematic search, it will be as follows:

1. Analysis of the user query. Initially, Google analyzes the user query, identifying key phrases in it, evaluating the context itself.
2. Content indexing. The search engine scans and indexes Internet pages, extracting titles, metadata, text content and related information from them.
3. Comparison of information. Here, a comparison of the user query and indexed content is performed, which ultimately allows you to find exactly those pages that will be most relevant to the entered phrase.
4. Result ranking. It is performed based on a fairly large number of factors, including the relevance of the content, its quality, the authority of the site, etc.
5. Result output. The user receives a structured list that will contain links to online pages, video content, pictures, tables and everything that is relevant to his request.

We would like to draw your attention to the fact that thematic search involves the possibility of using various criteria. As an option, this could be:

1. Topic. Let's say that you need information about a holiday in Montenegro. In this case, Google will analyze the request and find all those pages that contain information about this country directly in the context of the tourist's interests and will display you the results that are relevant to it.
2. Interests. Perhaps you may be interested in growing hydrangeas in Sochi. Here, in addition to analyzing key queries, your location will already be taken into account, because the specifics of caring for plants will differ significantly depending on the region.
3. Personality. Relevant for queries related to the life, career, and interests of famous people. Let's say you want to find out information about the life and work of Madonna. The search engine will define such a query as biographical and will provide you with detailed information about the biography of this singer. That is, there will already be dates, key events from her life, and corresponding photos.

Thanks to such an organized structure, it is possible to get the most detailed, structured, and substantiated answers to user queries, even without going directly to the sites themselves. But what is also very important is that they will fully answer your question. That is, if you are interested in information about the life and work of Madonna, then you will not receive an article in the selection dedicated to some particular event from her life. If you are interested in a vacation in Montenegro, you will not be provided with information about the country's economy or political life.

Thematic search also improves user experience, saves time and effort in obtaining the necessary information. This technology also opens up new opportunities for launching more effective advertising campaigns that will be aimed exclusively at people interested in a given product, service, or information. That is, thematic search can truly be called innovative.

What is fan search

The system described in the patent is only a display of the official documentation of the Google search engine regarding the Query Fan-Out technology, which is also called fan search. In short, this is the so-called "query branching method". It turns out that several search queries related to each other depending on the data sources or themes will be sent for processing at the same time. But at the output, the results for each such subtopic will be combined with each other, forming a simple answer that will be understandable to absolutely any user, regardless of how deeply he is immersed in this issue. Thanks to this approach, it is possible to provide the most expanded and deep access to information in comparison with the work of a classic search engine.

Here we can highlight a number of key advantages:

1. Completeness of the answer. The user receives truly comprehensive information on his request, including a very complex one, literally in a matter of seconds. To do this, he will not have to independently enter different requests, go to dozens of sites, make additional clarifications, etc.
2. High efficiency in processing fuzzy requests. Surely, when working with complex topics, you have more than once encountered in practice how difficult it can be to set the most correct search request. That is, difficulties arise already at the stage of its formulation. And here Query Fan-Out technology will become a reliable assistant. It will find the most correct answers to those queries that involve a comprehensive study of the topic.
3. Significant time savings. All the work associated with preparing additional queries, analyzing and processing the results obtained will now be replenished by the machine. You get rid of such a need, save your time and effort.

That is, the technique of correct queries involves simultaneously sending several search phrases related to each other by data sources, subtopics with subsequent unification. That is, despite their obvious similarity, Query Fan-Out technology and thematic search have a number of differences in implementation. And now you understand what exactly we are talking about. But still, they are both aimed at providing the user with the most relevant and competent answers to his queries.

Getting to know the summary generator

A summary generator, or as it is often called a resume generator, is one of the key elements of a thematic search system. It is designed to form a kind of generalization on each of the topics that the search engine will provide after processing the initial request. In practice, this is implemented as follows:

• A summary generator is a large language model that has been previously trained to generate unique text materials.
• The tool can use one or more excerpts of source materials in its work, which the system will pre-group by each topic. In addition to this, information from metadata, headings that are directly related to the selected text elements can also be used, which ultimately has a positive effect on the quality of the finished summary and its information content.
• A summary generator can be launched simultaneously with the user sending a request. But it can also be connected to work directly by the command of the thematic search engine directly at the moment of waiting for the request to be processed.

Unlike the classic ranking, which we are used to earlier, here the results of the requests will be grouped by separate topics, on the basis of which the summary will be created. That is, thematic search has a direct impact on what content will be displayed to the audience. Any user request will now no longer be the final one, but on the contrary, a starting point for the subsequent creation of a chain of smaller requests that allow you to fully reveal the topic.

As a result, the user literally goes through the entire path of discovering subtopics related to the original request. As a result, it may happen that the site that ultimately receives a click may turn out to be completely different from the one that was at the top of the classic search results for the main request. With a high degree of probability, this will be a completely different page, corresponding to additional queries, but, nevertheless, answering the user's question as much as possible.

Options for displaying answers in thematic search

Thematic search engine can work in different ways:

1. Topics will be displayed directly on the search results pages, and in some cases they can be organized by pre-setting filters in accordance with the topics. The display of the topic itself will be marked based on its main query.
2. In some cases, the thematic search system can take as a basis the significance or relevance of the content to the search query and rank the topics based on this.
3. The display of the answer on the search results page itself can be done in the form of lists, vertical or horizontal carousels, cards, UI objects (user interface), etc.
4. All those search results that will ultimately be structured by topics are often called thematic search results. They are often displayed together with the results of the classic search engine results. Sometimes the results page can be divided into 2 parts, due to which a visual separation of the regular results and thematic ones is organized.

As you can see, thematic search results have quite a lot of nuances, and it is still extremely difficult to navigate their specifics. It is really difficult to predict what solution artificial intelligence will choose in each specific case. But as soon as this product is launched, everything will become simpler and more understandable. Now we can only talk about how technologically advanced such a search will be, how seriously it will simplify the work of a modern user.

How materials from different sources are combined

To make it easier to navigate the features of thematic search, let's pay attention to how content obtained from different sites is combined in this case. So, the summaries that artificial intelligence will ultimately generate will be based on a fairly large number of sites that the system previously selects and groups according to relevant topics. That is, it is impossible to predict in advance how links will be distributed, what the visibility of a particular site will be, and how much traffic it will be able to attract.

There is a possibility that thematic search will use unstructured data as a basis, having previously analyzed the content of the pages, and then will ensure their structuring in accordance with the main topic. Let's return to our example with a vacation in Montenegro. So, as soon as you enter such a request, the thematic search engine will analyze a certain number of sites located at the top of the regular search results for this request. As a result, it will generate separate topics, such as hotels in Montenegro, apartments in Montenegro, the best regions for holidays in Montenegro, food in Montenegro, restaurants in Montenegro, beaches in Montenegro, pros and cons of holidays in Montenegro, cost of holidays in Montenegro and many others, based on the content that will be presented on the selected sites.

For greater ease of identification, the topic name itself will include a phrase generated by the neural network that describes the topic itself. In some cases, these may be key queries taken from the text content presented on relevant pages. Thanks to this, the user receives a structured search result containing information on key issues in the topic of interest to him.

Nuances of interaction with the user audience

From the information that we provided above, it is already clear that a thematic search engine based on artificial intelligence has a fundamental difference from the classic ranking of a list of sites that match a user request. Here we are already talking about collecting information on various topics related to the main phrase entered in the search bar. And directly, how a specific user will interact with the generated summary will have a direct impact on the list of sites that will ultimately receive traffic.

All those subtopics that the system will generate automatically are a kind of alternative directions corresponding to the user's path, starting from his initial search query and up to all those that artificial intelligence considers suitable and helping to reveal the main question. We have already mentioned that when creating a summary, the site's metadata is always used, including document titles, as well as all the information that surrounds the text content. That is, we can conclude that initially well-structured material can ensure the competent formation of the final summary.

Thematic search and SEO

The last point to which we want to draw your attention is this is that the results of thematic search will also contain URLs of those pages that were used by the platform to select the initial information. In most cases, the selection will also display information about the title of the corresponding page and related metadata. Thematic search also pays special attention to attribution, links, which in principle has a positive effect on the results of website promotion. But still, there are a number of nuances that must be taken into account.

Thus, thematic search and AI Mode, in particular, suggest the possibility of 2 scenarios:

1. Users can get answers directly from drop-down windows, thematic links. And this means that the percentage of results with the so-called "zero click" ultimately increases. This means that traffic to the site will worsen. This is especially true for cases when users will have enough information that will be presented in the results of thematic search.
2. The transition to the site will be carried out, that is, the site will receive a click. This is possible in cases when the user goes to the page on which his information journey ends.

That is, all this allows us to conclude that we are talking about technologies that will ensure maximum convenience in finding the necessary information for sites, and will also generate good traffic to sites offering high-quality content relevant to user queries. But in any case, the entire process of search engines will have to be reviewed. The paradigm of ranking by key queries, which was relevant for a very long period of time, has simply lost its relevance today. Here, the emphasis must now be on integrating your site into the information journey that the user will go through for a specific search query. And it is desirable to make your site, so to speak, the destination.

Summing up

As you can see, Google is making a lot of efforts to withstand the realities in which this niche has found itself as a result of the active introduction of artificial intelligence into search, and also not to lose all that mass audience and primacy in the market that was won over decades. Now all these technologies are gradually being implemented into work, and many users have already been able to evaluate their effectiveness in practice. But it is still, of course, too early to talk about impressive results and unrivaled convenience, since many works and technologies are still at the stage of improvements and testing. But the situation here is changing every day. This means that gradually users will become more and more immersed in the world of advanced technologies and extraordinary technical solutions.

But in any case, it is important not to forget about the security and stability of work on the Internet. You should remember that hackers are also actively introducing artificial intelligence into their fraudulent schemes, making them even more sophisticated and technological. And if you do not think about how to protect yourself, then, alas, no one else will do it for you. But here everything is much simpler and more reliable, because there are mobile proxies from the MobileProxy.Space service. Here you can find out the most detailed information about what this product is, evaluate its functionality, a huge variety of geolocations, available tariffs, and also undergo free testing. So you can be convinced of the high technology of this product and its indispensability for everyone who would like to ensure the most functional, confidential and safe work on the Internet without risks, restrictions. You will also be able to bypass various access blocking, connect to sites that are blocked in your country at the regional level and not only.

If difficulties arise in the work process and additional consultations are required, the technical support service is available 24/7. For those who are looking for options for earning additional income, we recommend evaluating the possibility of earning money on mobile proxies. All that is required of you in this case is to regularly top up SIM cards, providing stable access to the Internet.

Makaleyi paylaşın: