Search methods, devices, electronic equipment and storage media

Abstract

This disclosure relates to the field of computer science, specifically to the technical fields of artificial intelligence, big data, and intelligent search, and specifically to a search method, apparatus, electronic device, and storage medium. The specific implementation scheme is as follows: in response to a current search operation, a search request carrying the current search term is sent to the server; the initial search results corresponding to the current search term, obtained by the server using the search engine, are displayed; in response to a browsing operation on the initial search results, user behavior information is generated to facilitate sending a result generation request carrying the user behavior information to the server; the generative search results corresponding to the current search term, obtained by the server based on the user behavior information and using a large language model, are displayed. Using this disclosure can improve the accuracy of search results.

Description

Technical Field

[0001] This disclosure relates to the field of computers, specifically to the technical fields of artificial intelligence, big data, and intelligent search, and specifically to a search method, device, electronic device, and storage medium. Background Technology

[0002] A search engine is a search technology that retrieves specific information from the internet and returns it to the user based on their needs and certain algorithms and strategies. While advancements in computer technology have improved search efficiency, they still struggle to provide users with highly accurate search results. Summary of the Invention

[0003] This disclosure provides a search method, apparatus, electronic device, and storage medium.

[0004] According to one aspect of this disclosure, a search method is provided, comprising:

[0005] In response to the current search operation, a search request carrying the current search term is sent to the server;

[0006] The display server uses the initial search results corresponding to the current search term obtained by the search engine;

[0007] In response to browsing actions in response to initial search results, user behavior information is generated to facilitate sending a result generation request carrying the user behavior information to the server.

[0008] The display server generates search results corresponding to the current search term based on user behavior information and a large language model.

[0009] According to another aspect of this disclosure, a search method is provided, comprising:

[0010] Receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term;

[0011] Use a search engine to obtain initial search results for the current search term, so that the initial search results can be sent to the terminal device;

[0012] The receiving terminal device initiates a result generation request carrying user behavior information after generating user behavior information in response to a browsing operation for the initial search results.

[0013] Based on user behavior information, a large language model is used to obtain generative search results for the current search term, so that the generative search results can be sent to the terminal device.

[0014] According to another aspect of this disclosure, a search apparatus is provided, comprising:

[0015] The first request initiating unit is used to initiate a search request carrying the current search term to the server in response to the current search operation;

[0016] The first display unit is used to display the initial search results corresponding to the current search term obtained by the server using the search engine;

[0017] The second request initiating unit is used to generate user behavior information in response to browsing operations for the initial search results, so as to send a result generation request carrying the user behavior information to the server.

[0018] The second display unit is used to display the generative search results corresponding to the current search term, obtained by the server based on user behavior information and using a large language model.

[0019] According to another aspect of this disclosure, a search apparatus is provided, comprising:

[0020] The first request receiving unit is used to receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term.

[0021] The search unit is used to obtain initial search results for the current search term using a search engine, so as to send the initial search results to the terminal device.

[0022] The second request receiving unit is used to receive a result generation request carrying user behavior information initiated by the terminal device after generating user behavior information in response to the browsing operation for the initial search results.

[0023] The result generation unit is used to obtain generative search results for the current search term based on user behavior information and a large language model, so as to send the generative search results to the terminal device.

[0024] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0025] At least one processor;

[0026] The memory that is communicatively connected to the at least one processor;

[0027] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform any of the methods described in the present disclosure.

[0028] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform any of the methods according to embodiments of this disclosure.

[0029] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the methods according to embodiments of this disclosure.

[0030] Using this disclosure can improve the accuracy of search results.

[0031] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0032] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0033] Figure 1 A flowchart illustrating a search method (terminal device side) provided in an embodiment of this disclosure;

[0034] Figure 2 A schematic diagram illustrating the display effect of a search page provided in an embodiment of this disclosure;

[0035] Figure 3 An illustrative diagram illustrating an initial search result provided for embodiments of this disclosure;

[0036] Figure 4A and Figure 4B An illustration of a folding method for an initial search result provided in an embodiment of this disclosure;

[0037] Figure 5 A flowchart illustrating a search method (server-side) provided in an embodiment of this disclosure;

[0038] Figure 6 A flowchart illustrating the completeness of a search method provided in this disclosure.

[0039] Figure 7 A schematic diagram illustrating a search method provided in an embodiment of this disclosure;

[0040] Figure 8 A schematic structural block diagram of a search device provided in an embodiment of this disclosure;

[0041] Figure 9 A schematic structural block diagram of a search device provided in an embodiment of this disclosure;

[0042] Figure 10 This is a schematic structural block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0043] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0044] Currently, search results are typically obtained by aggregating results from across the web in response to users' online needs. In terms of usage, they usually combine an input box with a list of search results. When a user needs to change their search terms, they enter the new terms in the input box, obtain new search results to replace the original results, and then reload the new search terms and the new search results list. Here, the new search results are unrelated to the previous search results. Technically, web technologies are typically used to build the user interface, allowing users to input search terms. A server index is used to query relevant results, and based on these results, the recommended search results are determined and displayed to the user. When a new search term is entered, a new Uniform Resource Locator (URL) is constructed and the page is refreshed, or new search results are retrieved asynchronously using JavaScript and XML (Asynchronous JavaScript and XML, Ajax). The original search results are then removed, and finally, the new search results are rendered.

[0045] The above methods have two main drawbacks. First, the search results are obtained by integrating results from the entire internet, resulting in a wide search scope. Second, users typically need to input search terms with clear semantics and possess certain search term editing skills to obtain relatively accurate search results. However, users often lack these abilities. Therefore, existing search methods cannot provide users with accurate search results, especially when dealing with complex or long-tail search needs.

[0046] Based on the above research, this disclosure provides a search method that can be applied to electronic devices, such as terminal devices. The following will be combined with... Figure 1 The flowchart shown illustrates a search method provided by an embodiment of this disclosure. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order.

[0047] Step S101: In response to the current search operation, a search request carrying the current search term is sent to the server;

[0048] Step S102: Display the initial search results obtained by the server using the search engine corresponding to the current search term;

[0049] Step S103: In response to the browsing operation for the initial search results, generate user behavior information so as to send a result generation request carrying user behavior information to the server.

[0050] Step S104: Display the generative search results corresponding to the current search term obtained by the server based on user behavior information and using a large language model.

[0051] The current search operation can be triggered by the user entering the current search term in the search interface. In a specific example, the current search operation can be triggered by the user entering the current search term in the search session area loaded in the search interface. The search term can be a word, phrase, sentence, etc. After triggering the current search operation, the terminal device can respond to the current search operation by generating a search request carrying the current search term and sending the search request to the server.

[0052] After receiving a search request, the server can use a search engine to obtain initial search results corresponding to the current search terms carried in the search request, and then send the initial search results to the terminal device for display. The search engine can be a traditional search engine such as a full-text search engine, directory search engine, or meta-search engine, which retrieves search results corresponding to the current search terms from the internet as the initial search results. The initial search results may include multiple search sub-results.

[0053] After the terminal device displays the initial search results, the user can browse any of the multiple sub-results included in the initial search results. In response to browsing the initial search results, the terminal device can generate user behavior information to characterize the user's intent and determine the content the user is interested in. Subsequently, the terminal device can generate a result generation request carrying the user behavior information and send it to the server.

[0054] After receiving the result generation request, the server, based on the user behavior information carried in the request, uses a large language model to obtain generative search results corresponding to the current search term, and sends the generative search results to the terminal device for display. The large language model is pre-trained, possesses general language knowledge, world knowledge, and domain-specific expertise, which are stored internally as parameters, and has strong language interaction capabilities. In a specific example, the large language model can be an autoregressive generative model with a Transformer architecture. The generative search results can be generated as multimodal information such as plain text, tables, images, and videos; this disclosure does not impose specific limitations on this.

[0055] The search method provided in this disclosure can, in response to a current search operation, initiate a search request carrying the current search term to the server; display the initial search results corresponding to the current search term obtained by the server using a search engine; generate user behavior information in response to a browsing operation on the initial search results, so as to initiate a result generation request carrying the user behavior information to the server; and display the generative search results corresponding to the current search term obtained by the server based on the user behavior information using a large language model. The generative search results, as the final search results for the current search term, are not solely obtained from a traditional search engine. Instead, after the server obtains the initial search results corresponding to the current search term using a traditional search engine, the terminal device generates user behavior information in response to a browsing operation on the initial search results, and then the server obtains the generative search results corresponding to the current search term using a large language model based on the user behavior information. The user behavior information is used to characterize the user's intent and determine the content the user is interested in. Therefore, obtaining the generative search results corresponding to the current search term using a large language model based on user behavior information can improve the personalized matching degree between the generative search results and the user, thereby improving the accuracy of the generative search results.

[0056] Furthermore, as mentioned above, in this embodiment of the disclosure, the initial search results corresponding to the current search term may include multiple search sub-results. To facilitate user filtering, users can view more search sub-results corresponding to the current search term by infinitely scrolling down or loading pages.

[0057] In some alternative implementations, "generating user behavior information in response to a browsing action on the initial search results" may include the following steps:

[0058] In response to browsing actions based on the initial search results, determine the current landing page;

[0059] If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms, as well as the historical landing pages corresponding to the historical search results;

[0060] Generate user behavior information based on the current landing page and historical landing pages.

[0061] The current landing page can be used to represent the page corresponding to any of the multiple sub-results included in the initial search results. For example, if a user viewed sub-result A1 in the initial search results, the current landing page includes the page corresponding to sub-result A1; as another example, if a user viewed both sub-results A1 and A2 in the initial search results, the current landing page includes the page corresponding to sub-result A1 and the page corresponding to sub-result A2.

[0062] The current search term can be a search term in the search session. That is, the search session can include one or more search terms. When the search session includes multiple search terms, the multiple search terms can be sent serially, and the sending sequence number is the round number of the search term. For example, when the first search term is issued, it is used as the current search term, and steps S101 to S104 are executed. Subsequently, when the second search term is issued, it is used as the current search term, and steps S101 to S104 are executed. At this time, the first search term is a historical search term in the search session other than the current search term (the second search term), and the initial search result and generated search result corresponding to the first search term are both historical search results corresponding to the historical search term. If a third search term is issued, it is used as the current search term, and steps S101 to S104 are executed. At this time, the first and second search terms are historical search terms in the search session other than the current search term (the third search term), and the initial search result and generated search result corresponding to the first search term, as well as the initial search result and generated search result corresponding to the second search term, are both historical search results corresponding to the historical search term.

[0063] Similarly, it is understood that in this embodiment of the disclosure, the historical landing page can be used to represent the page corresponding to each of the multiple sub-results included in the historical search results. For example, if a user browses search sub-result B1 in the historical search results, the historical landing page includes the page corresponding to search sub-result B1; as another example, if a user browses search sub-results B1 and B2 in the historical search results, the historical landing page includes the page corresponding to search sub-result B1 and the page corresponding to search sub-result B2.

[0064] Through the above steps, in this embodiment of the disclosure, after determining the current landing page and obtaining the historical landing page corresponding to the historical search results, user behavior information can be generated based on the current landing page and the historical landing page, thereby improving the representativeness of user behavior information, further improving the degree of personalized matching between generated search results and users, and thus improving the accuracy of generated search results.

[0065] In some alternative implementations, "generating user behavior information based on the current landing page and historical landing pages" may include the following steps:

[0066] Retrieve page-related information for the target landing page;

[0067] User behavior information is generated based on page-related information.

[0068] The target landing page can be either the current landing page or a historical landing page. In other words, obtaining the page-related information of the target landing page includes: taking the current landing page and the historical landing page as the target landing page respectively, and obtaining the page-related information of the target landing page.

[0069] The relevant page information includes at least one of the following: the target landing page's content title, content information, and page source.

[0070] In one specific example, the content information could be the page content of the target landing page; in another specific example, the content information could be a summary of the page content of the target landing page.

[0071] In a specific example, the page source can be a web address.

[0072] Through the above steps, this embodiment of the disclosure can obtain page-related information of the target landing page and generate user behavior information based on the page-related information. Since the target landing page can be either the current landing page or a historical landing page, and the page-related information includes at least one of the target landing page's content title, content information, and page source, the comprehensiveness of the page-related information can be improved, thereby further improving the degree of personalized matching between the generated search results and the user, and thus improving the accuracy of the generated search results.

[0073] In some alternative implementations, the search method may further include the following steps:

[0074] The display server uses one or more predicted search terms obtained from a large language model based on user behavior information.

[0075] That is, after generating user behavior information, the server can also use the large language model to obtain one or more predicted search terms based on the user behavior information, and send these one or more predicted search terms to the terminal device for display. Afterwards, the user can directly click on any of the one or more predicted search terms to use that predicted search term as the current search term, and execute steps S101 to S104. The predicted search terms can be search terms related to the user behavior information, and are predicted to be the next search term the user might use.

[0076] Through the above steps, in this embodiment of the disclosure, one or more predicted search terms obtained by the server based on user behavior information and using a large language model can be directly displayed on the terminal device. Users can directly click on any of the predicted search terms to use as the current search term and execute steps S101 to S104, thereby simplifying the user's operation process and improving the usability and scalability of the search method.

[0077] Please combine Figure 2 In a specific example, after a user enters their current search term on the search page of their terminal device, the initial search results corresponding to the current search term are first displayed in the overall search results area of ​​the search page. Then, generated search results corresponding to the current search term are displayed in the personalized results area of ​​the search page. Simultaneously, one or more predicted search terms corresponding to the current search term are displayed in the search recommendation area of ​​the search page. The generated search results corresponding to the current search term can be displayed before the initial search results and the one or more predicted search terms corresponding to the current search term; that is, the generated search results corresponding to the current search term will be displayed first to facilitate timely browsing by the user.

[0078] In some alternative implementations, "displaying the initial search results obtained by the server using the search engine corresponding to the current search term" may include the following steps:

[0079] Create a root sandbox corresponding to the current search term;

[0080] Create a sub-sandbox within the root sandbox for each search sub-result included in the initial search results;

[0081] For each sub-result included in the initial search results, the display logic for the sub-result is executed in the corresponding sub-sandbox to display the sub-result.

[0082] Specifically, for the current search term, a sandbox can be created and bound to the round number of the current search term to complete the step of creating a root sandbox corresponding to the current search term.

[0083] Subsequently, sub-sandboxes can be created within the root sandbox for each search sub-result included in the initial search results. During this process, a corresponding Application Programming Interface (API) can be created for each sub-sandbox to facilitate communication between the sub-sandbox and the external environment. The API created for each sub-sandbox can include Window API, Document API, Element API, etc. Finally, for each search sub-result included in the initial search results, the display logic for that search sub-result can be executed in the corresponding sub-sandbox to display the search sub-result.

[0084] Please combine Figure 3 In a specific example, the initial search results corresponding to the current search term include sub-results C1, C2, and C3. After creating the root sandbox D1 corresponding to the current search term, sub-sandboxes D11, D12, and D13 can be created within root sandbox D1 for each of the sub-results C1, C2, and C3, respectively. This allows the display logic for sub-result C1 to be executed in sub-sandbox D11, displaying sub-result C1; the display logic for sub-result C2 to be executed in sub-sandbox D12, and the display logic for sub-result C3 to be executed in sub-sandbox D13, displaying sub-result C3.

[0085] Through the above steps, in this embodiment of the disclosure, a root sandbox corresponding to the current search term can be created to protect and isolate the display logic of the initial search results corresponding to the current search term. At the same time, in the root sandbox corresponding to the current search term, a sub-sandbox can be created for each search sub-result included in the initial search results, which can avoid conflicts in the display logic of the search sub-results included in the initial search results (e.g., conflicts in variables, functions, Document Object Model (DOM) etc.), thereby ensuring that the initial search results are displayed normally.

[0086] In some alternative implementations, the search method may further include the following steps:

[0087] In response to the next round of search operations, the display status of the initial search results is switched from expanded to collapsed, or the display status of the unviewed sub-results in the initial search results is switched from expanded to collapsed.

[0088] It is understood that in this embodiment of the disclosure, the initial search results corresponding to the current search term can be displayed by using a supplementary result folding method. That is, in response to the next round of search operation, the display state of the initial search results can be switched from an expanded state to a folded state.

[0089] Please combine Figure 4A In a specific example, after a user enters the search term "I want to travel to region E" on the search page of their terminal device, the initial search results corresponding to "I want to travel to region E" are first displayed in the overall search results display area, including search sub-results F1, F2, and F3. Then, generated search results corresponding to "I want to travel to region E" are displayed in the personalized results generation area. Simultaneously, one or more predicted search terms corresponding to "I want to travel to region E" are displayed in the search recommendation area. Taking the supplementary results collapse method as an example, after the user enters the next search term "What's the local weather like?" on the search page of their terminal device, the display state of the initial search results can be switched from expanded to collapsed.

[0090] If the search method also includes the step of "displaying one or more predicted search terms obtained by the server based on user behavior information using a large language model", then when displaying the initial search results corresponding to the current search term using the supplementary result folding method, the display status of one or more predicted search terms corresponding to the current search term can also be switched from the expanded state to the collapsed state.

[0091] It is understood that, in this embodiment of the disclosure, the initial search results corresponding to the current search term can also be displayed in a folded manner. That is, in response to the next round of search operation, the display state of the unviewed search sub-results in the initial search results can be switched from an expanded state to a folded state.

[0092] Please combine Figure 4BIn another specific example, after a user enters the search term "I want to travel to region E" on the search page of their terminal device, the initial search results corresponding to "I want to travel to region E" can first be displayed in the overall search results display area, including search sub-results F1, F2, and F13. Then, generated search results corresponding to "I want to travel to region E" can be displayed in the personalized results generation area. Simultaneously, one or more predicted search terms corresponding to "I want to travel to region E" can be displayed in the search recommendation area. Taking a collapsible inset approach as an example, after the user enters the next search term "What's the local weather like?" on the search page of their terminal device, the display status of unviewed search sub-results in the initial search results (e.g., search sub-result F13) can be switched from an expanded state to a collapsed state.

[0093] If the search method also includes the step of "displaying one or more predicted search terms obtained by the server based on user behavior information using a large language model", then when displaying the initial search results corresponding to the current search term using an inset collapse method, the display status of one or more predicted search terms corresponding to the current search term can also be switched from an expanded state to a collapsed state.

[0094] Through the above steps, in this embodiment of the disclosure, in response to the next round of search operations, the display state of the initial search results can be switched from an expanded state to a collapsed state, or the display state of the unviewed search sub-results in the initial search results can be switched from an expanded state to a collapsed state. In this way, while ensuring that the core search results (the generated search results corresponding to the current search term) are displayed, browsing space is saved, making the search interface more concise.

[0095] This disclosure provides a search method that can be applied to electronic devices, such as servers. The following will be combined with... Figure 5 The flowchart shown illustrates a search method provided by an embodiment of this disclosure. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order.

[0096] Step S501: Receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term;

[0097] Step S502: Use a search engine to obtain initial search results for the current search term, so as to send the initial search results to the terminal device.

[0098] Step S503: Receive a result generation request carrying user behavior information initiated by the terminal device after generating user behavior information in response to the browsing operation for the initial search results.

[0099] Step S504: Based on user behavior information, use a large language model to obtain generative search results for the current search term, so as to send the generative search results to the terminal device.

[0100] The current search operation can be triggered by the user entering the current search term in the search interface. In a specific example, the current search operation can be triggered by the user entering the current search term in the search session area loaded in the search interface. The search term can be a word, phrase, sentence, etc. After triggering the current search operation, the terminal device can respond to the current search operation by generating a search request carrying the current search term and sending the search request to the server.

[0101] After receiving a search request, the server can use a search engine to obtain initial search results corresponding to the current search terms carried in the search request, and then send the initial search results to the terminal device for display. The search engine can be a traditional search engine such as a full-text search engine, directory search engine, or meta-search engine, used to retrieve initial search results corresponding to the current search terms from the internet. The initial search results may include multiple search sub-results.

[0102] After the terminal device displays the initial search results, the user can browse any of the multiple sub-results included in the initial search results. In response to browsing the initial search results, the terminal device can generate user behavior information to characterize the user's intent and determine the content the user is interested in. Subsequently, the terminal device can generate a result generation request carrying the user behavior information and send it to the server.

[0103] After receiving the result generation request, the server uses the user behavior information carried in the request to generate generative search results corresponding to the current search term obtained from the large language model. These generative search results are then sent to the terminal device for display. The large language model is pre-trained, possessing general language knowledge, world knowledge, and domain-specific expertise, all stored internally as parameters, and has strong language interaction capabilities. In a specific example, the large language model could be an autoregressive generative model based on the Transformer architecture.

[0104] The search method provided in this disclosure can receive a search request initiated by a terminal device in response to a current search operation, carrying the current search term; obtain initial search results for the current search term using a search engine, so as to send the initial search results to the terminal device; receive a result generation request initiated by the terminal device after generating user behavior information in response to a browsing operation for the initial search results; and obtain generative search results for the current search term using a large language model based on the user behavior information, so as to send the generative search results to the terminal device. The generative search results, as the final search results for the current search term, are not solely obtained from a traditional search engine. Instead, after the server obtains the initial search results corresponding to the current search term using a traditional search engine, the terminal device generates user behavior information in response to a browsing operation for the initial search results, and then the server obtains the generative search results corresponding to the current search term using a large language model based on the user behavior information. The user behavior information is used to characterize user intent and determine the content of interest to the user. Therefore, obtaining generative search results corresponding to the current search term using a large language model based on user behavior information can improve the degree of personalized matching between the generative search results and the user, thereby improving the accuracy of the generative search results.

[0105] In some alternative implementations, "obtaining initial search results for the current search term using a search engine, so as to send the initial search results to the terminal device" may include the following steps:

[0106] If the search session also includes historical search terms other than the current search term, the target search term is obtained based on the historical search terms and the current search term;

[0107] Use a search engine to obtain search results for the target search term; the search results for the target search term are the initial search results.

[0108] Send the initial search results to the terminal device.

[0109] The current search term can be a search term in the search session. That is, the search session can include one or more search terms, and when the search session includes multiple search terms, the multiple search terms can be issued sequentially, and the issuance sequence number is the round number of the search term. For example, when issuing the first search term, the first search term is used as the current search term, and steps S101 to S104 are executed; subsequently, when issuing the second search term, the second search term is used as the current search term, and steps S101 to S104 are executed. At this time, the first search term is a historical search term in the search session other than the current search term (the second search term); if a third search term is issued, the third search term is used as the current search term, and steps S101 to S104 are executed. At this time, the first search term and the second search term are historical search terms in the search session other than the current search term (the third search term).

[0110] Historical search terms can be stored on the server so that they can be retrieved directly from the server when performing the above steps, or they can be sent to the server in the search request. This disclosure does not impose any specific limitations on this.

[0111] In a specific example, "obtaining the target search term based on historical search terms and the current search term" can include: rewriting the current search term based on historical search terms to obtain the target search term. For example, based on historical search terms, using a query rewriting model, the current search term can be rewritten to obtain the target search term. For instance, if the current search term is "What's the local weather like?", and the historical search term is "I want to travel to region E", then based on the historical search term "I want to travel to region E", the current search term "What's the local weather like?" can be rewritten to obtain the target search term "What's the weather like in region E?".

[0112] Through the above steps, in this embodiment of the disclosure, when the search session also includes historical search terms other than the current search term, the target search term can be obtained based on the historical search terms and the current search term. A search engine is then used to obtain search results for the target search term, which serve as the initial search result corresponding to the current search term. This initial search result is then sent to the terminal device. Since the target search term is obtained based on historical search terms and the current search term—that is, obtained in conjunction with the context—it has clear semantics. Even if the user lacks search term editing skills, the reliability of the initial search result can be improved, thereby increasing the accuracy of the generated search results.

[0113] In some alternative implementations, "obtaining generative search results for the current search term using a large language model based on user behavior information, so as to send the generative search results to the terminal device" may include the following steps:

[0114] Generate prompts based on user behavior information;

[0115] Input the prompt information into the large language model to obtain the generative search results for the current search term output by the large language model;

[0116] Generative search results are sent to the terminal device.

[0117] Further explanation of user behavior information can be found in the relevant descriptions in the embodiments of the search method applied to terminal devices, and will not be repeated here.

[0118] Based on the above steps, it can be understood that in this embodiment of the disclosure, after obtaining user behavior information, prompt information (also known as the prompt information of the large language model) can be generated based on the user behavior information, and the prompt information can be input into the large language model to obtain the generative search results output by the large language model for the current search term.

[0119] Through the above steps, in this embodiment of the disclosure, prompt information can be generated based on user behavior information to guide the large language model to output generative search results for the current search term according to the prompt information. This not only improves the degree of personalized matching between generative search results and users, thereby improving the accuracy of generative search results, but also further improves search efficiency based on a simplified search process.

[0120] In some alternative implementations, "generating prompt information based on user behavior information" may include the following steps:

[0121] If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms;

[0122] Based on search sessions, historical search results, initial search results, and user behavior information, prompt messages are generated.

[0123] As mentioned earlier, the current search term can be a search term in the search session. That is, the search session can include one or more search terms, and when the search session includes multiple search terms, the multiple search terms can be sent serially, and the sending sequence number is the round number of the search term. For example, when the first search term is issued, it is used as the current search term, and steps S101 to S104 are executed. Subsequently, when the second search term is issued, it is used as the current search term, and steps S101 to S104 are executed. At this time, the first search term is a historical search term in the search session other than the current search term (the second search term), and the initial search result and generated search result corresponding to the first search term are both historical search results corresponding to the historical search term. If a third search term is issued, it is used as the current search term, and steps S101 to S104 are executed. At this time, the first and second search terms are historical search terms in the search session other than the current search term (the third search term), and the initial search result and generated search result corresponding to the first search term, as well as the initial search result and generated search result corresponding to the second search term, are both historical search results corresponding to the historical search term.

[0124] In this embodiment of the disclosure, the historical search results and the initial search results used to generate the prompt information may specifically include at least one of the content title, content information and internal source of any search sub-result in the historical search results, and at least one of the content title, content information and internal source of any search sub-result in the initial search results.

[0125] Furthermore, in this embodiment of the disclosure, historical search terms, historical search results, and initial search results other than the current search term in the search session can be stored on the server so that they can be directly retrieved from the server when performing the above steps, or they can be sent to the server in the result generation request. This embodiment of the disclosure does not impose any specific limitations on this.

[0126] Through the above steps, in this embodiment of the disclosure, when the search session also includes historical search terms other than the current search term, historical search results corresponding to the historical search terms can be obtained, and prompt information can be generated based on the search session, historical search results, initial search results, and user behavior information. Since the prompt information is generated based on the search session, historical search results, initial search results, and user behavior information, it has a certain degree of comprehensiveness, thus further improving the accuracy of the generated search results.

[0127] In some alternative implementations, the search method may further include the following steps:

[0128] One or more predicted search terms are obtained using a large language model based on user behavior information;

[0129] Send one or more predicted search terms to the terminal device.

[0130] That is, after generating user behavior information, the server can also use the large language model to obtain one or more predicted search terms based on the user behavior information, and send these one or more predicted search terms to the terminal device for display. Afterwards, the user can directly click on any of the one or more predicted search terms to use that predicted search term as the current search term, and execute steps S101 to S104. The predicted search terms can be search terms related to the user behavior information, and are predicted to be the next search term the user might use.

[0131] Through the above steps, in this embodiment of the disclosure, one or more predicted search terms obtained by the server based on user behavior information and using a large language model can be directly displayed on the terminal device. Users can directly click on any of the predicted search terms to use it as the current search term, and execute steps S101 to S104, thereby simplifying user operations and improving the usability of the search method.

[0132] It should be noted that, in the embodiments of this disclosure, when the search session also includes historical search terms other than the current search term, one or more predicted search terms may be generated during the process of generating prompt information based on the search session, historical search results corresponding to historical search terms, initial search results corresponding to the current search term, and user behavior information, and then inputting the prompt information into the large language model to obtain the generative search results for the current search term output by the large language model.

[0133] The following will combine Figure 6 The complete process of a search method provided in the embodiments of this disclosure will be described.

[0134] (1) After the user enters the search application installed on the terminal device, the search page is displayed on the terminal device.

[0135] (2) When a user enters the current search term on the search page and triggers the current search operation, the user sends a search request to the server carrying the current search term. If the search session also includes historical search terms other than the current search term, the search request also carries the historical search terms.

[0136] (3) Based on historical search terms, the server rewrites the current search term using the Query rewriting model to obtain the target search term, and uses the search engine to obtain the search results for the target search term as the initial search results corresponding to the current search term, and sends the initial search results to the terminal device.

[0137] (4) Display the initial search results corresponding to the current search term in the full search results display area on the search page; wherein, “display the initial search results corresponding to the current search term” includes: creating a root sandbox corresponding to the current search term; creating a sub-sandbox in the root sandbox for each search sub-result included in the initial search results; and executing the display logic of the search sub-result in the sub-sandbox corresponding to each search sub-result included in the initial search results to display the search sub-result.

[0138] (5) In response to the browsing operation for the initial search results, the terminal device generates user behavior information so as to send a result generation request carrying the user behavior information to the server; wherein, if the search session also includes historical search terms other than the current search term, the result generation request also carries historical search terms other than the current search term in the search session, historical search results corresponding to the historical search terms, and the initial search results corresponding to the current search term.

[0139] (6) Based on the search session, historical search results corresponding to historical search terms, initial search results corresponding to the current search term, and user behavior information, generate prompt information, input the prompt information into the big language model, obtain the generative search results for the current search term output by the big language model, and use server-send event (SSE) technology to send the generative search results to the terminal device; in this process, the big language model can also be given risk control intervention capabilities to effectively avoid the generation of unreasonable content. At the same time, the big language model can also output one or more predicted search terms corresponding to the current search term and send them to the terminal device.

[0140] (7) The terminal device displays the generated search results for the current search term, as well as one or more predicted search terms corresponding to the current search term, and records the landing page corresponding to the generated search results.

[0141] (8) In response to the next round of search operation, switch the display state of the initial search results from expanded to collapsed, or switch the display state of the unviewed search sub-results in the initial search results from expanded to collapsed, and execute the next round of search operation.

[0142] Please see Figure 7 This is a schematic diagram of a search method provided in an embodiment of the present disclosure.

[0143] As previously described, the search method provided in this disclosure can be applied to electronic devices. Electronic devices are intended to represent various forms of digital computers, such as servers, or terminal devices like desktop computers, laptops, and mobile terminals.

[0144] Terminal devices can be used for:

[0145] In response to the current search operation, a search request carrying the current search term is sent to the server;

[0146] The display server uses the initial search results corresponding to the current search term obtained by the search engine;

[0147] In response to browsing actions in response to initial search results, user behavior information is generated to facilitate sending a result generation request carrying the user behavior information to the server.

[0148] The display server generates search results corresponding to the current search term based on user behavior information and a large language model.

[0149] The server can be used for:

[0150] Receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term;

[0151] Use a search engine to obtain initial search results for the current search term, so that the initial search results can be sent to the terminal device;

[0152] The receiving terminal device initiates a result generation request carrying user behavior information after generating user behavior information in response to a browsing operation for the initial search results.

[0153] Based on user behavior information, a large language model is used to obtain generative search results for the current search term, so that the generative search results can be sent to the terminal device.

[0154] The current search operation can be the search operation triggered when the user enters the current search term in the search interface.

[0155] It should be noted that, in the embodiments disclosed herein, Figure 7 The schematic diagrams shown are for illustrative purposes only and are not restrictive. Those skilled in the art can use them as a basis for their own interpretation. Figure 7 The examples may be modified in various obvious ways and / or substitutions, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of this disclosure.

[0156] To better implement the search method applicable to terminal devices described above, embodiments of this disclosure also provide a search device that can be integrated into an electronic device, such as a terminal device. The following will be combined with... Figure 8 The schematic diagram shown illustrates a search device 800 provided in a public embodiment.

[0157] Search device 800 includes:

[0158] The first request initiating unit 801 is used to initiate a search request carrying the current search term to the server in response to the current search operation.

[0159] The first display unit 802 is used to display the initial search results corresponding to the current search term obtained by the server using the search engine;

[0160] The second request initiating unit 803 is used to generate user behavior information in response to the browsing operation for the initial search results, so as to send a result generation request carrying the user behavior information to the server.

[0161] The second display unit 804 is used to display the generative search results corresponding to the current search term obtained by the server based on user behavior information and using a large language model.

[0162] In some optional implementations, the second request initiating unit 803 is used to:

[0163] In response to browsing actions based on the initial search results, determine the current landing page;

[0164] If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms, as well as the historical landing pages corresponding to the historical search results;

[0165] Generate user behavior information based on the current landing page and historical landing pages.

[0166] In some optional implementations, the second request initiating unit 803 is used to:

[0167] Obtain page-related information for the target landing page; wherein, the target landing page is either the current landing page or a historical landing page, and the page-related information includes at least one of the target landing page's content title, content information, and page source;

[0168] User behavior information is generated based on page-related information.

[0169] In some alternative implementations, the search device 800 further includes a third display unit for:

[0170] The display server uses one or more predicted search terms obtained from a large language model based on user behavior information.

[0171] In some alternative implementations, the first display unit 802 is used for:

[0172] Create a root sandbox corresponding to the current search term;

[0173] Create a sub-sandbox within the root sandbox for each search sub-result included in the initial search results;

[0174] For each sub-result included in the initial search results, the display logic for the sub-result is executed in the corresponding sub-sandbox to display the sub-result.

[0175] In some alternative implementations, the search device 800 further includes a display status control unit for:

[0176] In response to the next round of search operations, the display status of the initial search results is switched from expanded to collapsed, or the display status of the unviewed sub-results in the initial search results is switched from expanded to collapsed.

[0177] For a description of the specific functions and examples of each unit of the search device 800 in this embodiment, please refer to the relevant descriptions of the corresponding steps in the foregoing search method embodiments applicable to terminal devices, which will not be repeated here.

[0178] To better implement the above-described search method applicable to servers, embodiments of this disclosure also provide a search device that can be integrated into an electronic device, such as a server. The following will be combined with... Figure 9 The schematic diagram shown illustrates a search device 900 provided in a disclosed embodiment.

[0179] Search device 900 includes:

[0180] The first request receiving unit 901 is used to receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term.

[0181] Search unit 902 is used to obtain initial search results for the current search term using a search engine, so as to send the initial search results to the terminal device.

[0182] The first request receiving unit 903 is used to receive a result generation request carrying user behavior information initiated by the terminal device after generating user behavior information in response to the browsing operation for the initial search results.

[0183] The result generation unit 904 is used to obtain generative search results for the current search term based on user behavior information and using a large language model, so as to send the generative search results to the terminal device.

[0184] In some alternative implementations, the search unit 902 is used for:

[0185] If the search session also includes historical search terms other than the current search term, the target search term is obtained based on the historical search terms and the current search term;

[0186] Use a search engine to obtain search results for the target search term, as the initial search results;

[0187] Send the initial search results to the terminal device.

[0188] In some alternative implementations, the result generation unit 904 is used for:

[0189] Generate prompts based on user behavior information;

[0190] Input the prompt information into the large language model to obtain the generative search results for the current search term output by the large language model;

[0191] Generative search results are sent to the terminal device.

[0192] In some alternative implementations, the result generation unit 904 is used for:

[0193] If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms;

[0194] Based on search sessions, historical search results, initial search results, and user behavior information, prompt messages are generated.

[0195] In some optional implementations, the search unit is also used for:

[0196] One or more predicted search terms are obtained using a large language model based on user behavior information;

[0197] Send one or more predicted search terms to the terminal device.

[0198] For a description of the specific functions and examples of each unit of the apparatus 900 in this embodiment, please refer to the relevant descriptions of the corresponding steps in the foregoing search method embodiments applicable to servers, which will not be repeated here.

[0199] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0200] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0201] Figure 10A schematic block diagram of an example electronic device 1000 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0202] like Figure 10 As shown, device 1000 includes a computing unit 1001, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 1002 or a computer program loaded from storage unit 1008 into random access memory (RAM) 1003. The RAM 1003 may also store various programs and data required for the operation of device 1000. The computing unit 1001, ROM 1002, and RAM 1003 are interconnected via bus 1004. An input / output (I / O) interface 1005 is also connected to bus 1004.

[0203] Multiple components in device 1000 are connected to I / O interface 1005, including: input unit 1006, such as keyboard, mouse, etc.; output unit 1007, such as various types of displays, speakers, etc.; storage unit 1008, such as disk, optical disk, etc.; and communication unit 1009, such as network card, modem, wireless transceiver, etc. Communication unit 1009 allows device 1000 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0204] The computing unit 1001 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as search methods. For example, in some embodiments, the search method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on device 1000 via ROM 1002 and / or communication unit 1009. When the computer program is loaded into RAM 1003 and executed by the computing unit 1001, one or more steps of the search method described above may be performed. Alternatively, in other embodiments, the computing unit 1001 may be configured to perform the search method by any other suitable means (e.g., by means of firmware).

[0205] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0206] Program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0207] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM) or flash memory, optical fibers, compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0208] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a cathode ray tube (CRT) monitor or a liquid crystal display (LCD)) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0209] The systems and technologies described herein can be implemented in computing systems that include back-end components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected via digital data communication (e.g., a communication network) of any form or medium. Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0210] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0211] This disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to perform a search method.

[0212] This disclosure also provides a computer program product, including a computer program that implements a search method when executed by a processor.

[0213] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure is achieved, and this is not limited herein. Furthermore, in this disclosure, relational terms such as "first," "second," and "third" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Additionally, "multiple" in this disclosure can be understood as at least two.

[0214] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A search method applied to a terminal device; the method comprising: In response to the current search operation, a search request carrying the current search term is sent to the server; wherein, the server is used to use a search engine to obtain an initial search result corresponding to the current search term carried in the search request, and return the initial search result to the terminal device; The server displays the initial search results corresponding to the current search term, obtained using the search engine. In response to a browsing operation on the initial search results, user behavior information is generated to initiate a result generation request carrying the user behavior information to the server. The server displays generative search results corresponding to the current search term, obtained based on the user behavior information and using a large language model. The user behavior information is used to characterize user intent and determine the content the user is interested in; the generation of user behavior information in response to a browsing operation on the initial search result includes: In response to a browsing operation on the initial search result, a current landing page is determined; wherein, the current landing page is used to represent the page corresponding to any of the multiple sub-results included in the initial search result; If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms and the historical landing pages corresponding to the historical search results; wherein, the historical landing pages are used to represent the page corresponding to any of the multiple search sub-results included in the historical search results; The user behavior information is generated based on the current landing page and the historical landing pages.

2. The method according to claim 1, wherein, Based on the current landing page and the historical landing pages, user behavior information is generated, including: Obtain page-related information of the target landing page; wherein, the target landing page is either the current landing page or the historical landing page, and the page-related information includes at least one of the content title, content information and page source of the target landing page; The user behavior information is generated based on the page-related information.

3. The method according to claim 1 or 2, further comprising: The server displays one or more predicted search terms obtained using a large language model based on the user behavior information.

4. The method according to claim 1 or 2, wherein, The display of the initial search results obtained by the server using the search engine corresponding to the current search term includes: Create a root sandbox corresponding to the current search term; Create a sub-sandbox within the root sandbox for each search sub-result included in the initial search results; For each search sub-result included in the initial search results, the display logic of the search sub-result is executed in the sub-sandbox corresponding to the search sub-result to display the search sub-result.

5. The method according to claim 1 or 2, further comprising: In response to the next round of search operations, the display state of the initial search results is switched from expanded to collapsed, or the display state of the unviewed sub-results in the initial search results is switched from expanded to collapsed.

6. A search method applied to a server; the method comprising: Receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term; An initial search result for the current search term is obtained using a search engine, so as to send the initial search result to the terminal device; wherein the terminal device is used to display the initial search result; The terminal device receives a result generation request carrying the user behavior information after generating user behavior information in response to a browsing operation for the initial search results. Based on the user behavior information, a generative search result for the current search term is obtained using a large language model, so as to send the generative search result to the terminal device. The user behavior information is used to characterize user intent and determine the content the user is interested in; the generation of user behavior information in response to a browsing operation on the initial search result includes: In response to a browsing operation on the initial search result, a current landing page is determined; wherein, the current landing page is used to represent the page corresponding to any of the multiple sub-results included in the initial search result; If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms and the historical landing pages corresponding to the historical search results; wherein, the historical landing pages are used to represent the page corresponding to any of the multiple search sub-results included in the historical search results; The user behavior information is generated based on the current landing page and the historical landing pages.

7. The method according to claim 6, wherein, The step of obtaining initial search results for the current search term using a search engine, and then sending the initial search results to the terminal device, includes: If the search session also includes historical search terms other than the current search term, the target search term is obtained based on the historical search terms and the current search term; The search results for the target search term are obtained using a search engine and used as the initial search results; The initial search results are sent to the terminal device.

8. The method according to claim 6, wherein, The step of obtaining generative search results for the current search term using a large language model based on the user behavior information, so as to send the generative search results to the terminal device, includes: Based on the user behavior information, a prompt message is generated; The prompt information is input into the large language model to obtain the generative search results for the current search term output by the large language model; The generative search results are sent to the terminal device.

9. The method according to claim 8, wherein, The step of generating a prompt message based on the user behavior information includes: If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms; The prompt message is generated based on the search session, the historical search results, the initial search results, and the user behavior information.

10. The method of claim 6, further comprising: Based on the user behavior information, one or more predicted search terms are obtained using a large language model; Send the one or more predicted search terms to the terminal device.

11. A search device applied to a terminal device; The device includes: The first request initiating unit is used to initiate a search request carrying the current search term to the server in response to the current search operation; wherein the server is used to use a search engine to obtain an initial search result corresponding to the current search term carried in the search request, and return the initial search result to the terminal device; The first display unit is used to display the initial search results corresponding to the current search term obtained by the server using the search engine; The second request initiating unit is used to generate user behavior information in response to the browsing operation for the initial search results, so as to initiate a result generation request carrying the user behavior information to the server. The second display unit is used to display the generative search results corresponding to the current search term obtained by the server based on the user behavior information and using a large language model. The user behavior information is used to characterize user intent and determine the content the user is interested in; the second request initiating unit is used for: In response to a browsing operation on the initial search result, a current landing page is determined; wherein, the current landing page is used to represent the page corresponding to any of the multiple sub-results included in the initial search result; If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms and the historical landing pages corresponding to the historical search results; wherein, the historical landing pages are used to represent the page corresponding to any of the multiple search sub-results included in the historical search results; The user behavior information is generated based on the current landing page and the historical landing pages.

12. The apparatus according to claim 11, wherein, The second request initiating unit is used for: Obtain page-related information of the target landing page; wherein, the target landing page is either the current landing page or the historical landing page, and the page-related information includes at least one of the content title, content information and page source of the target landing page; The user behavior information is generated based on the page-related information.

13. The apparatus according to claim 11 or 12, further comprising a third display unit for: The server displays one or more predicted search terms obtained using a large language model based on the user behavior information.

14. The apparatus according to claim 11 or 12, wherein, The first display unit is used for: Create a root sandbox corresponding to the current search term; Create a sub-sandbox within the root sandbox for each search sub-result included in the initial search results; For each search sub-result included in the initial search results, the display logic of the search sub-result is executed in the sub-sandbox corresponding to the search sub-result to display the search sub-result.

15. The apparatus according to claim 11 or 12, further comprising a display status control unit, configured to: In response to the next round of search operations, the display state of the initial search results is switched from expanded to collapsed, or the display state of the unviewed sub-results in the initial search results is switched from expanded to collapsed.

16. A search device applied to a server; the device comprising: The first request receiving unit is used to receive a search request initiated by the terminal device in response to the current search operation, carrying the current search term. The search unit is used to obtain initial search results for the current search term using a search engine, so as to send the initial search results to the terminal device; wherein the terminal device is used to display the initial search results; The second request receiving unit is configured to receive a result generation request carrying the user behavior information initiated by the terminal device after generating user behavior information in response to the browsing operation for the initial search results. The result generation unit is used to obtain generative search results for the current search term based on the user behavior information using a large language model, so as to send the generative search results to the terminal device. The user behavior information is used to characterize user intent and determine the content the user is interested in; the generation of user behavior information in response to a browsing operation on the initial search result includes: In response to a browsing operation on the initial search result, a current landing page is determined; wherein, the current landing page is used to represent the page corresponding to any of the multiple sub-results included in the initial search result; If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms and the historical landing pages corresponding to the historical search results; wherein, the historical landing pages are used to represent the page corresponding to any of the multiple search sub-results included in the historical search results; The user behavior information is generated based on the current landing page and the historical landing pages.

17. The apparatus according to claim 16, wherein, The search unit is used for: If the search session also includes historical search terms other than the current search term, the target search term is obtained based on the historical search terms and the current search term; The search results for the target search term are obtained using a search engine and used as the initial search results; The initial search results are sent to the terminal device.

18. The apparatus according to claim 16, wherein, The result generation unit is used for: Based on the user behavior information, a prompt message is generated; The prompt information is input into the large language model to obtain the generative search results for the current search term output by the large language model; The generative search results are sent to the terminal device.

19. The apparatus according to claim 18, wherein, The result generation unit is used for: If the search session also includes historical search terms other than the current search term, retrieve the historical search results corresponding to the historical search terms; The prompt message is generated based on the search session, the historical search results, the initial search results, and the user behavior information.

20. The apparatus of claim 16, further comprising a search term prediction unit for: Based on the user behavior information, one or more predicted search terms are obtained using a large language model; Send the one or more predicted search terms to the terminal device.

21. An electronic device, comprising: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 10.

22. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1 to 10.

23. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 10.

Images