Searching method and device, electronic equipment and storage medium

Abstract

The invention relates to a search method and device, electronic equipment and a storage medium, and belongs to the field of Internet. The method comprises the following steps: in response to a search instruction in a search page, searching based on a first search text input in the search page; in the search page, displaying search extension information determined through a large language model; the search extension information comprises a semantic description text and a recommended search text, the semantic description text is used for describing the semantics of the first search text, and the recommended search text is a text related to the semantics. According to the method, the richness and integrity of the display content are improved, the semantics of the first search text can be accurately described, the object is guided to think whether the expression of the first search text is reasonable and accurate or not, the recommended search text is used for searching, and the search success rate is improved.

Description

Technical Field

[0001] This disclosure relates to the Internet field, and more particularly to a search method, apparatus, electronic device, and storage medium. Background Technology

[0002] Search functionality is a common feature in various applications and an important way for users to obtain content, knowledge, and items. The accuracy of search results is of paramount importance.

[0003] Typically, after a user enters search text into the search box, the system performs a search based on that text and displays the results. However, if the search text is uncommon or contains typos, the search results may be inaccurate, resulting in a lower success rate. Summary of the Invention

[0004] This disclosure provides a search method, apparatus, electronic device, and storage medium that improves the richness and completeness of displayed content, accurately describes the semantics of the first search text, guides the user to consider whether the expression of the first search text is reasonable and accurate, and uses recommended search text for searching, thereby improving the search success rate.

[0005] According to one aspect of the embodiments of this disclosure, a search method is provided, the method comprising: In response to a search command on the search page, a search is performed based on the first search text entered on the search page; The search page displays search extension information determined by the large language model; The search extension information includes semantic description text and recommended search text. The semantic description text is used to describe the semantics of the first search text, and the recommended search text is text related to the semantics.

[0006] In some embodiments, the method further includes: The first search result corresponding to the first search text is displayed on the search page; The search page displays expanded search information determined by the large language model, including: If the first search result meets the target conditions, the search extension information is displayed on the search page. The target conditions refer to the conditions that the search results that do not meet the search requirements satisfy.

[0007] In some embodiments, the first search result includes multiple search results, and displaying the search extension information on the search page when the first search result meets the target conditions includes at least one of the following: If the matching degree between the search result ranked last among the multiple search results and the first search text is less than a first threshold, the search extension information is displayed on the search page. If the matching degree between the search result ranked at the target position and the first search text is less than a second threshold among the multiple search results, the search extension information is displayed on the search page; If the matching degree between the first search result ranked first among the multiple search results and the first search text is less than a third threshold, the search extension information is displayed on the search page. If the maximum predicted click-through rate among the multiple search results is less than a fourth threshold, the search extension information is displayed on the search page. If the type tags of the multiple search results do not match the interest tags of the currently logged-in object, the search extension information is displayed on the search page.

[0008] In some embodiments, displaying search extension information determined by a large language model on the search page includes: Create a suggestion word, the suggestion word including the first search text, and the suggestion word being used to instruct the large language model to generate search extension information for the first search text; The prompt words are input into the large language model to obtain the search extension information.

[0009] In some embodiments, the process of training the large language model includes: Obtain the pre-trained large language model; Obtain training samples, which include a first sample text, a second sample text, and a sample description text. The sample description text is used to describe the semantics of the first sample text, and the second sample text is text related to the semantics of the first sample text. The large language model is trained based on the training samples.

[0010] In some embodiments, the method further includes: In response to the selection operation of the recommended search text, a search is performed based on the recommended search text; Displays the second search result corresponding to the recommended search text.

[0011] In some embodiments, the search page includes a search results area, which displays a first search result corresponding to the first search text; The display of the second search result corresponding to the recommended search text includes: Replace the first search result in the search results area with the second search result.

[0012] In some embodiments, displaying the second search result corresponding to the recommended search text includes: Display a dialog page, which is used to display the dialogue information between the currently logged-in object and the large language model; The dialog page displays the recommended search text sent by the object, and the second search result replied by the large language model.

[0013] In some embodiments, the dialog page further displays an input box, and the method further includes: Get the question text entered in the input box; The question text sent by the object is displayed on the dialog page; The dialogue page displays the answer text of the large language model.

[0014] In some embodiments, the search extension information further includes a summary text corresponding to the recommended search text; the method further includes: In response to a view of the details of the introductory text, the introductory text corresponding to the recommended search text is displayed on the search page. The introductory text includes the introductory text and other text.

[0015] In some embodiments, the first search text is a term name, the semantic description text describes the semantics of the term represented by the first search text, and the recommended search text is the name of a term related to the semantics of the first search text; or, The first search text is an interrogative sentence, and the recommended search text is the alternative answer text to the first search text; or, The first search text is item text, the semantic description text is used to describe the item represented by the first search text, and the recommended search text is text corresponding to the item that matches the description of the first search text; or, The first search text belongs to the category of teaching text, and the semantic description text includes a learning path and at least one knowledge point in the learning path, wherein each knowledge point is a recommended search text.

[0016] According to another aspect of the present disclosure, a search apparatus is provided, the apparatus comprising: The search unit is configured to perform a search based on the first search text entered on the search page in response to a search command on the search page; The display unit is configured to display search extension information determined by the large language model on the search page; The search extension information includes semantic description text and recommended search text. The semantic description text is used to describe the semantics of the first search text, and the recommended search text is text related to the semantics.

[0017] In some embodiments, the display unit is further configured to display a first search result corresponding to the first search text on the search page; and to display the search extension information on the search page when the first search result meets the target conditions, wherein the target conditions refer to the conditions met by search results that do not meet the search requirements.

[0018] In some embodiments, the first search result includes multiple search results, and the display unit is configured to be at least one of the following: If the matching degree between the search result ranked last among the multiple search results and the first search text is less than a first threshold, the search extension information is displayed on the search page. If the matching degree between the search result ranked at the target position and the first search text is less than a second threshold among the multiple search results, the search extension information is displayed on the search page; If the matching degree between the first search result ranked first among the multiple search results and the first search text is less than a third threshold, the search extension information is displayed on the search page. If the maximum predicted click-through rate among the multiple search results is less than a fourth threshold, the search extension information is displayed on the search page. If the type tags of the multiple search results do not match the interest tags of the currently logged-in object, the search extension information is displayed on the search page.

[0019] In some embodiments, the display unit is configured to create a prompt word, the prompt word including the first search text, and the prompt word being used to instruct the large language model to generate search extension information of the first search text; the prompt word is input into the large language model to obtain the search extension information.

[0020] In some embodiments, the apparatus further includes: The training module is used to obtain the pre-trained large language model; obtain training samples, the training samples including a first sample text, a second sample text and a sample description text, the sample description text being used to describe the semantics of the first sample text, and the second sample text being text related to the semantics of the first sample text; and train the large language model based on the training samples.

[0021] In some embodiments, the search unit is further configured to perform a search based on the recommended search text in response to a selection operation on the recommended search text; and display a second search result corresponding to the recommended search text.

[0022] In some embodiments, the search page includes a search results area displaying a first search result corresponding to the first search text; the search unit is configured to replace the first search result in the search results area with the second search result.

[0023] In some embodiments, the search unit is configured to display a dialogue page, which displays dialogue information between the currently logged-in object and the large language model; the dialogue page displays the recommended search text sent by the object, and the second search result replied by the large language model.

[0024] In some embodiments, the dialog page further displays an input box, and the device further includes: The question-and-answer unit is configured to: retrieve the question text entered in the input box; display the question text issued by the object in the dialogue page; and display the answer text replied by the large language model in the dialogue page.

[0025] In some embodiments, the search extension information further includes a summary text corresponding to the recommended search text; the device further includes: The details viewing unit is configured to display the description text corresponding to the recommended search text on the search page in response to a details viewing operation of the description text, the description text including the description text and other text.

[0026] In some embodiments, the first search text is a term name, the semantic description text describes the semantics of the term represented by the first search text, and the recommended search text is the name of a term related to the semantics of the first search text; or, The first search text is an interrogative sentence, and the recommended search text is the alternative answer text to the first search text; or, The first search text is item text, the semantic description text is used to describe the item represented by the first search text, and the recommended search text is text corresponding to the item that matches the description of the first search text; or, The first search text belongs to the category of teaching text, and the semantic description text includes a learning path and at least one knowledge point in the learning path, wherein each knowledge point is a recommended search text.

[0027] According to another aspect of the present disclosure, an electronic device is provided, the electronic device comprising: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the search method as described above.

[0028] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein instructions in the computer-readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the search method as described above.

[0029] According to another aspect of the present disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the search method as described above.

[0030] The method provided in this disclosure, when searching based on the first search text entered on the search page, also determines extended search information through a large language model to supplement the first search text, increasing the amount of information displayed and improving the richness and completeness of the displayed content. Furthermore, by determining and displaying semantic description text of the first search text through the large language model, the semantics of the first search text are accurately described, guiding the user to consider whether the expression of the first search text is reasonable and accurate. Additionally, by determining and displaying recommended search text semantically related to the first search text through the large language model, the user is guided to use the recommended search text for searching. This eliminates the need for the user to repeatedly modify the first search text to obtain more accurate search text, thus improving the search success rate.

[0031] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0032] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0033] Figure 1This is a schematic diagram illustrating an implementation environment according to an exemplary embodiment.

[0034] Figure 2 This is a flowchart illustrating a search method according to an exemplary embodiment.

[0035] Figure 3 This is a schematic diagram of a search page based on relevant technologies.

[0036] Figure 4 This is a schematic diagram illustrating another search page based on relevant technologies.

[0037] Figure 5 This is a flowchart illustrating another search method according to an exemplary embodiment.

[0038] Figure 6 This is a schematic diagram illustrating a search page according to an exemplary embodiment.

[0039] Figure 7 This is a schematic diagram illustrating another search page according to an exemplary embodiment.

[0040] Figure 8 This is a schematic diagram illustrating another search page according to an exemplary embodiment.

[0041] Figure 9 This is a schematic diagram illustrating yet another search page according to an exemplary embodiment.

[0042] Figure 10 This is a schematic diagram illustrating a dialog page according to an exemplary embodiment.

[0043] Figure 11 This is a schematic diagram illustrating another dialog page according to an exemplary embodiment.

[0044] Figure 12 This is a block diagram illustrating a search device according to an exemplary embodiment.

[0045] Figure 13 This is a block diagram illustrating an electronic device according to an exemplary embodiment.

[0046] Figure 14 This is a block diagram illustrating a server according to an exemplary embodiment. Detailed Implementation

[0047] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0048] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0049] It should be noted that the user information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in this disclosure are all authorized by the user or fully authorized by all parties. The collection, use, and processing of related data shall comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0050] The search method provided in this disclosure is executed by an electronic device, which may include at least one of a terminal and a server. The terminal may be at least one of a smartphone, smartwatch, desktop computer, laptop, wireless terminal, and laptop computer. The server may include at least one of a single server, multiple servers, a cloud computing platform, and a virtualization center.

[0051] Figure 1 This is a schematic diagram of an implementation environment provided by an embodiment of this disclosure. The implementation environment includes: a first terminal 110, a server 120, and a second terminal 130.

[0052] The first terminal 110 has an application 111 installed and running. This application 111 can be a video game program, a search program, a live streaming program, a social networking program, a video playback program, or other types of programs. When the first terminal runs the application 111, the interface of the application 111 is displayed on the screen of the first terminal 110. The first terminal 110 is the terminal used by the first object 112.

[0053] The second terminal 130 has an application 131 installed and running. This application 131 can be a video game program, a search program, a live streaming program, a social networking program, a video playback program, or other types of programs. When the second terminal 130 runs the application 131, the interface of the application 131 is displayed on the screen of the second terminal 130. The second terminal 130 is the terminal used by the second object 113.

[0054] In some embodiments, the applications installed on the first terminal 110 and the second terminal 130 are the same, or the applications installed on the two terminals are the same type of application on different operating system platforms (Android or iOS). The first terminal 110 can refer to one of a plurality of terminals, and the second terminal 130 can refer to another of a plurality of terminals; this disclosure embodiment only uses the first terminal 110 and the second terminal 130 as examples. The device types of the first terminal 110 and the second terminal 130 may be the same or different, and the device types include at least one of: smartphones, tablets, game consoles, smart bracelets, virtual reality devices, laptops, and desktop computers.

[0055] Those skilled in the art will understand that the number of terminals described above can be more or less. For example, there may be only one terminal, or there may be 10 or more terminals. This disclosure does not limit the number of terminals or the type of device.

[0056] It should be noted that, Figure 1 Only two terminals are shown in the diagram, but in different embodiments, multiple other terminals 140 can access the server 120. Optionally, one or more terminals 140 may also be terminals corresponding to developers, on which a development and editing platform for applications supporting one or more virtual scenarios is installed. Developers can edit and update applications on terminals 140 and transmit the updated application installation package to the server 120 via wired or wireless network. The first terminal 110 and the second terminal 130 can download the application installation package from the server 120 to update the application.

[0057] The first terminal 110, the second terminal 130, and other terminals 140 are connected to the server 120 via a wireless network or a wired network.

[0058] Server 120 includes at least one of a single server, multiple servers, a cloud computing platform, and a virtualization center. Server 120 is used to provide background services for applications. Optionally, server 120 undertakes the primary computing work, and the terminal undertakes the secondary computing work; or, server 120 undertakes the secondary computing work, and the terminal undertakes the primary computing work; or, server 120 and the terminal use a distributed computing architecture for collaborative computing.

[0059] Taking the first terminal 110 as an example, the first terminal 110 logs into the first object in the application 111. The first object publishes resources, such as video resources, audio resources, and live streaming resources, in the application 111 for other objects to watch, share, comment on, or like. The first terminal 110 can also play resources from other objects through the application 111 and perform operations such as sharing, commenting on, or liking the played resources. The second terminal 130 works similarly, and will not be described in detail here.

[0060] Resources published by each terminal are stored in server 120. Server 120 acts as a resource-sharing platform, providing a space for multiple objects to interact based on resources. Furthermore, the resources collected by server 120 can be used as search materials for various terminals to search. Taking the first terminal 110 as an example, the first terminal 110 can display a search page. An object enters the first search text on the search page, and the first terminal 110 can perform a search based on the first search text, displaying the first search result. It can also determine the search extension information of the first search text through the large language model provided by server 120, displaying the search extension information on the search page as supplementary information to the first search text.

[0061] The above implementation environment is merely an example. Other implementation environments can also be used in this embodiment, which will not be elaborated here.

[0062] Figure 2 This is a flowchart illustrating a search method according to an exemplary embodiment, see [link to flowchart]. Figure 2 The method is performed by an electronic device, which may include the above-mentioned... Figure 1 The first terminal 110, server 120, or second terminal 130 in the implementation environment shown may also be other types of devices. The method includes the following steps: In step 201, in response to the search instruction on the search page, the electronic device performs a search based on the first search text entered on the search page.

[0063] The search page displays a search box and a search control. The search box is used to enter search text, and the search control is used to trigger the search command. When the electronic device enters the first search text in the search box and triggers the search control, the electronic device receives the search command and performs a search based on the first search text to obtain the first search result. The electronic device can then display the first search result on the search page.

[0064] In some embodiments, the electronic device runs an application that displays a search page, enabling it to search the application's database based on the first search text entered on the search page. This database includes resources published by multiple terminals through the application, such as video resources, audio resources, and live streaming resources.

[0065] In step 202, the electronic device displays search extension information determined by the large language model on the search page.

[0066] After the electronic device performs a search, it will also determine the search extension information of the first search text through a large language model. This search extension information can serve as supplementary information to the first search text for the target to view.

[0067] The search extension information includes semantic description text and recommended search text. Semantic description text describes the semantics of the first search text, explaining its meaning. Recommended search text is text related to this semantic meaning; it can be any text possessing that semantic meaning and can be considered a replacement for the first search text. Displaying the recommended search text guides the user to select it for searching.

[0068] For example, if the first search text is a colloquial expression with unclear semantics, the first search result obtained based on the first search text may not be accurate. However, by using a large language model to perform semantic analysis on the first search text, the resulting semantic description text can more accurately interpret the semantics of the first search text. Furthermore, the recommended search text is a text that more accurately expresses the semantics, such as the formal language corresponding to the first search text. Searching based on the recommended search text can yield more accurate search results.

[0069] Figure 3 This is a diagram illustrating a search page provided by related technologies. See [link / reference]. Figure 3 The original search text was "laotop," which was semantically unclear. The electronic device corrected the search text, determining the correct text to be "laptop." Therefore, the search was performed based on "laptop," displaying the search results for "laptop." This method can correct typos in the search text.

[0070] Figure 4 This is an illustration of another search page provided by related technologies; see [link / reference]. Figure 4The original search text entered in the search box is "tomato and egg noodles". The electronic device displays the search results corresponding to the search text and also corrects the search text. If it is determined that the text that the object originally wanted to search for was probably "tomato and egg noodles", then "tomato and egg noodles" is displayed as the recommended search text. After the object clicks on "tomato and egg noodles", the search text in the search box is changed to "tomato and egg noodles" and the search results corresponding to "tomato and egg noodles" are displayed.

[0071] Both of the above methods can correct typos in search text. However, if the search text does not contain typos, but rather is unclear, colloquial, semantically complex, ambiguous, or contains other errors, the technology will be unable to accurately identify the semantics of the search text, and therefore cannot determine accurate recommended search text.

[0072] The method provided in this disclosure, when searching based on the first search text entered on the search page, also determines extended search information through a large language model to supplement the first search text, increasing the amount of information displayed and improving the richness and completeness of the displayed content. Furthermore, by determining and displaying the semantic description text of the first search text through the large language model, the semantics of the first search text can be accurately described, guiding the user to consider whether the expression of the first search text is reasonable and accurate. Additionally, by determining and displaying recommended search text semantically related to the first search text through the large language model, the user is guided to use the recommended search text for searching. This eliminates the need for the user to repeatedly modify the first search text to obtain more accurate search text, thus improving the search success rate.

[0073] A comparison reveals that related technologies can only perform simple corrections to the original search text, such as modifying a single character or letter. The resulting corrected search text is largely the same as the original search text, only correcting the erroneous parts without altering the sentence structure. In contrast, the recommended search text determined in this embodiment is semantically related to the first search text, but it is not a simple correction; therefore, it is more diverse in form. For example, if the first search text is "red-tailed bird" and the recommended search text is "red-bellied tit," the difference is not merely a single character, but a change in expression.

[0074] Figure 5 This is a flowchart illustrating another search method according to an exemplary embodiment, see [link to flowchart]. Figure 5 The method is performed by an electronic device, which may include the above-mentioned... Figure 1 The first terminal 110, server 120, or second terminal 130 in the implementation environment shown may also be other types of devices. The method includes the following steps: In step 501, in response to a search instruction on the search page, the electronic device performs a search based on the first search text entered on the search page.

[0075] Step 501 is the same as step 201 above, and will not be repeated here.

[0076] In step 502, the electronic device displays the first search result corresponding to the first search text on the search page.

[0077] The database stores multiple resources. Based on the first search text, the electronic device can search the database to obtain the first search result. The first search result may include one or more search results. Each search result includes a resource in the database and the resource information corresponding to that resource, such as the resource's title, publication time, publication target, number of likes, etc.

[0078] In some embodiments, the electronic device determines the matching degree between multiple resources in the database and a first search text. The matching degree represents the degree of matching between the first search text and the resources. The higher the matching degree, the more closely the first search text matches the resource. The multiple resources are then arranged in descending order of matching degree to obtain the first search result. For example, a fixed number of resources are arranged in descending order of matching degree to obtain the first search result. Alternatively, at least one resource with a matching degree greater than a target threshold is selected from the multiple resources, and the at least one resource is arranged in descending order of matching degree to obtain the first search result.

[0079] In some embodiments, the search page includes a search trigger area and a search results area. The search trigger area includes a search box and a search control. The search box is used to input search text, and the search control is used to trigger a search command. The search results area is used to display search results; after the electronic device obtains a first search result, it displays the first search result in the search results area.

[0080] In step 503, if the first search result meets the target conditions, the electronic device displays the search extension information determined by the large language model on the search page.

[0081] Here, the target condition refers to the conditions that search results that do not meet the search requirements must satisfy. In other words, if the first search result meets the target condition, it means that the first search result does not meet the user's search needs and is not the search result the user desired. This target condition can be set by the user or by default on the electronic device. When the first search result meets the target condition, the extended search information of the first search text is determined through a large language model. This extended search information is then displayed on the search page so that the user can optimize their search text based on this information, improving search accuracy and ultimately obtaining more accurate search results, thus increasing the search success rate.

[0082] In some embodiments, the first search result includes multiple search results. If the first search result meets the target criteria, extended search information is displayed on the search page, including at least one of the following: 1. If the match between the last search result in multiple search results and the first search text is less than a first threshold, display extended search information on the search page.

[0083] In the first search result, multiple search results are arranged in descending order of their relevance to the first search text. If the relevance of the last search result to the first search text is less than a first threshold, it indicates that the relevance between the last search result and the first search text is poor, and it is not suitable as a search result for the first search text. In other words, the first search result contains one or more unsuitable search results. In this case, search extension information is displayed on the search page.

[0084] The first threshold can be determined based on the degree of matching between the last search result and the search text under normal circumstances.

[0085] 2. If the match between the search result ranked at the target position and the first search text is less than the second threshold, display extended search information on the search page.

[0086] In the first search result, multiple search results are arranged in descending order of their match with the first search text. If the match between the search result ranked at the target position and the first search text is less than a second threshold, it means that the match between the search result ranked at the target position and the first search text is lower than the match between the search result ranked at the target position and the search text under normal circumstances. This indicates that the match of each search result in the first search result is low and not accurate enough. In this case, search extension information is displayed on the search page.

[0087] The target position can be the first, fourth, etc., and the second threshold can be determined based on the degree of match between the search result ranked at the target position and the search text under normal circumstances. For example, the first and second items above can be combined in any form, such as the electronic device having a first threshold and a second threshold, where the second threshold is greater than the first threshold. If the electronic device finds that the degree of match between the last search result ranked among multiple search results and the first search text is less than the first threshold, or that the degree of match between the search result ranked at the target position and the first search text is less than the second threshold, then extended search information is displayed on the search page.

[0088] 3. If the match between the first search result ranked first among multiple search results and the first search text is less than the third threshold, display extended search information on the search page.

[0089] In the first search result, multiple search results are arranged in descending order of their match with the first search text. If the match between the first search result and the first search text is less than a third threshold, it indicates that the match between the first search result and the first search text is low, and the match between all other search results is also low. Therefore, the first search result is not accurate enough. In this case, search extension information is displayed on the search page.

[0090] The third threshold can be determined based on the degree of match between the first search result and the search text under normal circumstances. For example, the first, second, and third items mentioned above can be combined in any form, such as when the electronic device has a first threshold, a second threshold, and a third threshold, where the second threshold is greater than the first threshold and the third threshold is greater than the second threshold. If the electronic device displays extended search information on the search page when the degree of match between the last search result in multiple search results and the first search text is less than the first threshold, or when the degree of match between the search result in the target position in multiple search results and the first search text is less than the second threshold, or when the degree of match between the first search result in multiple search results and the first search text is less than the third threshold.

[0091] 4. If the maximum predicted click-through rate among multiple search results is less than the fourth threshold, display extended search information on the search page.

[0092] When an electronic device performs a search, it can also determine the predicted click-through rate (CTR) for each search result. The predicted CTR refers to the probability that an object will click on a search result after it is displayed. If the maximum predicted CTR among multiple search results is less than a fourth threshold, it means that even the highest predicted CTR is still below the fourth threshold, indicating a low probability that the object will click on the search result, and the search result does not meet the search requirements. In this case, search extension information is displayed on the search page.

[0093] 5. If the type tags of multiple search results do not match the interest tags of the currently logged-in user, display extended search information on the search page.

[0094] Each resource has a type tag, which indicates the type of the resource. The currently logged-in user has an interest tag, which indicates the types of resources the user is interested in. If the type tag of a search result matches the user's interest tag, it means the user is highly likely to be interested in the search result. If the type tag of a search result does not match the user's interest tag, it means the user is less likely to be interested in the search result. If the type tags of multiple search results do not match the user's interest tags, it means the user is less likely to be interested in any of the search results in the first search result. In this case, search extension information is displayed on the search page.

[0095] This disclosure uses the above five objective conditions as an example for illustration. These five objective conditions can be combined in any way to constitute optional solutions for this disclosure. Other objective conditions may also be used in other embodiments.

[0096] In some embodiments, the process of determining search extension information through a large language model includes: creating a cue word, the cue word including a first search text, and the cue word being used to instruct the large language model to generate search extension information for the first search text; inputting the cue word into the large language model to obtain the search extension information.

[0097] For example, an electronic device obtains a prompt word template, which includes instruction text and a search text filling position. The instruction text is used to instruct the large language model to generate search extension information for the text at the search text filling position. The electronic device fills the search text filling position with the first search text to obtain the prompt word.

[0098] The instruction text can describe the search extension information to be generated, for example, the instruction text could be "The search text used this time is..." <xxx>Please explain the meaning of this text and recommend at least four words that match the description of this search text and facilitate searching.

[0099] In this embodiment of the disclosure, by creating prompt words that include a first search text and instructing a large language model to generate search extension information, and inputting the prompt words into the large language model, the ability of the large language model to dynamically generate content can be fully utilized to obtain accurate search extension information. This not only improves the accuracy of search extension information, but also makes the operation process simple and fast, thus improving processing efficiency.

[0100] In this embodiment, the search extension information includes semantic description text and recommended search text. The semantic description text describes the semantics of the first search text, and the recommended search text is text related to the semantics. In some embodiments, if the semantics of the first search text are unclear due to colloquial expressions, complex semantics, ambiguity, or typos, the semantic description text is used to infer the semantics of the first search text, while the recommended search text is the written text corresponding to the semantics of the first search text predicted by a large language model. In some embodiments, the semantic description text and the recommended search text are independent texts; although logically related, they are distinct parts. In other embodiments, the semantic description text and the recommended search text are located in the same sentence, connected by conjunctions to form a complete sentence.

[0101] See Figure 6 The search page displays a search box and search controls at the top. Entering the initial search text "red-tailed bird" in the search box and clicking the search control displays the first search result in the bottom area of ​​the search page. This first search result includes the video cover, title, publisher, publication time, and number of likes for multiple videos. Additionally, extended search information is displayed in the middle area of ​​the search page, located between the top and bottom areas. This extended search information includes the semantic description "red-tailed bird is likely a common name for several red-feathered birds, and based on common usage, it is speculated to refer to the following birds," the initial search text "red-tailed bird," and recommended search texts "red-billed blue magpie," "red-bellied tit," and "marsh tit." The phrase "the following birds" is a conjunction that connects the semantic description text with the initial and recommended search texts, forming a complete sentence.

[0102] See Figure 6 The search extension information also includes the first search text itself, and the first search text is displayed as selected to indicate that the search results in the area below are the first search results corresponding to the first search text. The first search text and recommended search texts are displayed side by side, and users can select the text they want to search for from these texts.

[0103] In some embodiments, the search extension information also includes a summary text corresponding to the recommended search text, which provides a brief introduction to the recommended search text. When the search extension information includes multiple recommended search texts, at least one of them may have a corresponding summary text. Furthermore, the search extension information may directly display the recommended search text and its corresponding summary text, or it may display the recommended search text first, and then display the summary text corresponding to the recommended search text if the recommended search text is selected.

[0104] The method further includes: in response to a detail viewing operation on the summary text, displaying the introductory text corresponding to the recommended search text on the search page. The introductory text includes the summary text and other text. That is, the introductory text is a more detailed description of the recommended search text. Compared to the summary text, the introductory text contains more information and provides a more detailed explanation of the recommended search text, but occupies a larger display area. Therefore, the electronic device does not initially display the introductory text directly, but first displays the summary text to save display space. Only when a detail viewing operation on the summary text is detected will the introductory text be displayed, thus showing more information and improving the richness and completeness of the information. The detail viewing operation can be a double-click on the summary text, or an operation to click an expand control displayed after the summary text, or other operations, which will not be elaborated here.

[0105] In some embodiments, the first search text and the search extension information satisfy any of the following relationships: 1. The first search text is the term name, and the semantic description text is used to describe the semantics of the term represented by the first search text. The recommended search text is the name of a term that is semantically related to the first search text.

[0106] This disclosure can be applied to scenarios involving term searches. The first search text is the term name entered by the user, with the search intent being to find the term represented by the first search text. The semantic description text is the semantics of the term predicted by the large language model based on the first search text, and the recommended search text is the name of a term related to that semantics. If a search is performed solely based on the first search text, the resulting terms may not be accurate enough. However, by predicting the semantics of the term represented by the first search text and determining related term names as recommended search text, the user can be guided to re-search for more accurate terms using the recommended search text.

[0107] 2. The first search text is an interrogative sentence, so the recommended search text is the alternative answer text of the first search text.

[0108] This disclosure can be applied to situations where a question is being asked. The first search text is an interrogative sentence, and its search intent is to ask a question. The semantic description text is the semantics of the question being asked by the object, which is predicted by the large language model based on the first search text. The recommended search text is text related to this semantics, that is, alternative answer texts for the interrogative sentence. This guides the user to search again using the alternative answer texts and obtain search results associated with the alternative answer texts. These search results can all be considered as resources that can answer the interrogative sentence.

[0109] 3. The first search text is the item text. The semantic description text is used to describe the item represented by the first search text. The recommended search text is the text corresponding to the item that matches the description of the first search text.

[0110] This disclosure can be applied to searching for items, such as searching for goods in e-commerce shopping applications. The first search text is item text, that is, text used to describe an item, and its search intent is to find items that match the item text description. The semantic description text is normalized text predicted by a large language model based on the first search text, which can describe the desired item. The recommended search text is the text corresponding to the item that matches the description in the first search text, and can be considered as the text corresponding to items that the object may be interested in.

[0111] 4. The first search text belongs to the teaching text category. The semantic description text includes the learning path and at least one knowledge point in the learning path, and each knowledge point is a recommended search text.

[0112] This disclosure can be applied to educational scenarios. The first search text is educational text, such as a knowledge point, like a theory, formula, concept, or news event. The learning path included in the semantic description text refers to the learning path required to learn the first search text. This learning path includes at least one knowledge point; that is, if an object wants to learn the knowledge point represented by the first search text, it needs to follow this learning path. For example, if the learning path includes multiple knowledge points, these points are arranged sequentially, and the object needs to learn each knowledge point in order to learn the knowledge point represented by the first search text.

[0113] Furthermore, each knowledge point in the learning path can serve as a recommended search text, and each knowledge point is available for selection. By selecting any knowledge point, one can search for related resources based on that knowledge point in order to learn that knowledge point in depth.

[0114] The embodiments disclosed herein are applicable to various types of search text. Users can input any form of first search text according to their needs. The electronic device will adaptively determine the matching semantic description text and recommended search text based on the type of the first search text, rather than being limited to determining semantic description text and recommended search text with a fixed style. This improves flexibility, enhances adaptability, and expands the scope of applicable scenarios.

[0115] In this embodiment, the extended search information of the first search text is not directly displayed. Instead, it first determines whether the first search result meets the target conditions. If the first search result does not meet the target conditions, the extended search information is not displayed. Only if the first search result meets the target conditions will the extended search information be displayed on the search page. Therefore, this embodiment only displays extended search information when the first search result does not meet the search requirements, i.e., the first search result is poorly performed. This avoids occupying the display area of ​​the search page and interfering with the viewer's viewing of the first search result by displaying extended search information when the first search result meets the search requirements. This improves the utilization rate of the display area of ​​the search page and eliminates the need to perform the step of determining extended search information, thereby saving resources and time.

[0116] In addition, this embodiment of the disclosure takes into account the differences between search results that meet the search requirements and those that do not, thereby providing multiple target conditions for determining whether the first search result meets the search requirements, which can improve the accuracy of the judgment and determine the right time to display the recommended extended information.

[0117] In step 504, in response to the selection operation of recommended search text in the search extension information, a search is performed based on the recommended search text, and the second search result corresponding to the recommended search text is displayed.

[0118] When searching based on recommended search text, the recommended search text is displayed as selected on the search page to distinguish it from other recommended search texts. Additionally, when searching based on recommended search text, the search box can either retain the primary search text or automatically replace the primary search text with the recommended search text.

[0119] In this embodiment of the present disclosure, the user only needs to select any recommended search text to directly perform a search based on the recommended search text and display the second search result corresponding to the recommended search text. There is no need to perform the operation of entering the recommended search text in the search box, which saves search time, simplifies the search process, and improves search efficiency.

[0120] In some embodiments, the search page includes a search results area displaying a first search result corresponding to a first search text; then, displaying a second search result corresponding to a recommended search text includes replacing the first search result in the search results area with the second search result.

[0121] See Figure 6 With the primary search text "red-billed pheasant" selected, and the recommended search text "red-billed blue magpie" selected, then see [link to relevant search terms]. Figure 7 The recommended search text "Red-billed Blue Magpie" is displayed as selected, while the first search text "Red-billed Bird" is displayed as unselected. At this point, the first search result in the area below will be replaced with the second search result corresponding to the recommended search text "Red-billed Blue Magpie". See below for further details. Figure 8 If the object selects the recommended search text "red-bellied tit," then "red-bellied tit" will be displayed as selected, while "red-billed blue magpie" will be displayed as unselected. At this point, the second search result corresponding to "red-billed blue magpie" in the lower area will be replaced with the second search result corresponding to "red-bellied tit." Furthermore, since the large language model has already determined the descriptive text corresponding to "red-bellied tit," this descriptive text and an expand control will be displayed below "red-bellied tit." Clicking the expand control will lead to... Figure 9 This displays the descriptive text corresponding to the recommended search term "red-bellied tit".

[0122] See Figure 10 The search extension information includes two recommended search terms, "red-bellied tit" and "red-billed blue magpie," and a brief description of each recommended search term is displayed after it.

[0123] In this embodiment of the disclosure, search results are displayed in the search results area of ​​the search page, and the search results are replaced to avoid the search results occupying too much display area of ​​the search page, and also to avoid confusion between search results corresponding to different search texts.

[0124] In some embodiments, the search page includes a search trigger area and a search results area. The search trigger area includes a search box and a search control. The search box is used to input search text, and the search control is used to trigger a search command. The search results area is used to display search results. Additionally, when it is necessary to display extended search information, the search page also includes a search extension area. The search extension area is used to display extended search information and may be located between the search trigger area and the search results area.

[0125] When the amount of extended search information to be displayed in the search extension area increases, the search results area can be moved downwards to reserve more display space, thereby expanding the size of the search extension area and ensuring that the extended search information can be fully displayed. The search extension area can be displayed in card format or other formats.

[0126] For example, the search extension area displays recommended search text. When this recommended search text is selected, the second search result corresponding to that recommended search text will be displayed in the search results area. Therefore, there is no obvious separating symbol between the search extension area and the search results area, making them appear as a single visual unit. Users can quickly understand that the displayed search result is for the selected recommended search text in the search extension area. Furthermore, separating the search extension area and the search results area ensures that the content in each area is independent and does not affect the layout of the other, avoiding page display clutter.

[0127] In some embodiments, displaying a second search result corresponding to the recommended search text includes: displaying a dialogue page, which displays dialogue information between the currently logged-in object and the large language model; and displaying the recommended search text sent by the object and the second search result replied by the large language model on the dialogue page.

[0128] The dialogue page is used for dialogue between the currently logged-in user and the large language model. After the user selects a recommended search text, the electronic device displays the dialogue page and shows the recommended search text as the user's question to the large language model. The large language model then searches based on the recommended search text to obtain a second search result, which is displayed as a response to the question on the dialogue page, thus achieving the dialogue effect of user inquiry and large language model response.

[0129] Additionally, electronic devices can navigate from the search page to a chat page, and subsequently re-enable the search page upon exiting the chat page. Alternatively, the electronic device can display the chat page while maintaining the search page, with the chat page positioned above the search page. The chat page can be in the form of chat cards or other formats.

[0130] See Figure 11 The search extended information includes two recommended search texts, "red-bellied tit" and "red-billed blue magpie". After the object selects the recommended search text "red-bellied tit", a dialog page pops up. The dialog page sends "red-bellied tit" as the currently logged-in object and displays the response of the large language model to "red-bellied tit", which is the second search result obtained based on the search for "red-bellied tit".

[0131] See Figure 10 The search extension information includes two recommended search texts, "red-bellied tit" and "red-billed blue magpie," as well as a brief description of each recommended search text. When the user selects the recommended search text "red-bellied tit," a dialog page pops up. The dialog page sends the message "red-bellied tit" as the currently logged-in user and displays the response from the large language model to "red-bellied tit," which is the second search result obtained based on the search for "red-bellied tit."

[0132] In other embodiments, the dialogue page also displays an input box, and the method further includes: obtaining the question text entered in the input box; displaying the question text issued by the object on the dialogue page; and displaying the answer text of the large language model's reply on the dialogue page.

[0133] Users can not only select recommended search text to initiate a dialogue with the large language model, but also enter question text in the input box on the dialogue page to initiate a dialogue with the large language model. See also Figure 10 and Figure 11 An input box is displayed at the bottom of the dialog page. The user can enter text in the input box and then click the confirmation control or select the Enter key on the keyboard to start the dialog.

[0134] In this embodiment, selecting recommended search text displays a dialog page, showing the recommended search text and its corresponding second search result in a dialogue format. This simulates a conversation between the currently logged-in user and the large language model, enhancing interactivity. Furthermore, this question-and-answer format allows the recommended search text and the second search result to be displayed in groups, enabling the user to quickly distinguish the search text and its corresponding search result when viewing the dialog page, thus improving the clarity of the display. Additionally, entering question text in the input box on the dialog page allows the user to ask the large language model a question and receive its answer. This allows for deeper communication between the user and the large language model on the dialog page, and this communication is not limited by the format of the search text or the number of characters in the search box, enhancing the flexibility of the interaction.

[0135] In some embodiments, the display style of the second search result can be the same as that of the first search result. For example, for each resource in the second search result, the resource's cover image, title, target audience, number of likes, etc., can be displayed. In other embodiments, when the second search result is displayed in a dialog page, the second search result may include text, and may also include images, such as resource covers, as well as resource links or other resource information. Clicking on a resource cover image or resource link can redirect to a resource playback page to play the corresponding resource.

[0136] This disclosure applies to situations where a large language model has already been trained, and the electronic device uses the trained large language model to determine search extension information. In some embodiments, the large language model is a pre-trained general-purpose large language model, or a large language model that has been pre-trained and then specifically trained. The process of training the large language model includes: obtaining the pre-trained large language model; obtaining training samples, which include a first sample text, a second sample text, and sample description text, wherein the sample description text describes the semantics of the first sample text, and the second sample text is text semantically related to the first sample text; and training the large language model based on the training samples.

[0137] Since the pre-trained large language model has been trained on a large amount of text data, it can perform various functions such as text classification, image segmentation, and intelligent dialogue. However, it has not been specifically trained for search scenarios. The training samples are samples for search scenarios. Therefore, training the large language model based on the training samples can effectively improve the specificity of the large language model, thereby ensuring that the large language model is suitable for search scenarios.

[0138] For example, the first sample text might be colloquial, while the second sample text might be formal, with both representing the same or similar semantics. Alternatively, the first sample text might contain more content and be more complex and difficult to understand, while the second sample text might summarize and organize the semantics of the first sample text, making it simpler and easier to understand. Or, the first sample text might contain incorrect information, potentially causing ambiguity, while the second sample text might be correct and less likely to cause ambiguity. Training a large language model based on training samples allows it to learn the ability to infer the correct semantics from colloquial, complex, or incorrect text, and to convert the original text into semantically accurate, well-defined, and easily understood text. This improves the performance of the large language model, enabling it to accurately predict the semantics of the first search text in search scenarios and provide more accurate recommended search text.

[0139] When the first search text is relatively obscure, ambiguous, or inaccurate, the existing resources cannot explicitly contain the first search text, resulting in a low degree of matching between the resources and the first search text. Consequently, the number of resources contained in the first search result is limited, and the resources contained in the first search result have low relevance to the first search text. This makes it difficult for users to obtain effective search results, leading them to believe that the current application has insufficient resources, thereby reducing their trust in the application and the length of time they stay on the site.

[0140] By using a large language model to determine extended search information for the initial search text, the amount of information displayed is increased, enhancing the richness and completeness of the displayed content. This makes the search page structure more robust, ensuring that users receive complete and valuable information feedback in any search scenario. Furthermore, by determining and displaying semantic descriptions of the initial search text through the large language model, the semantics of the initial search text are accurately described, guiding users to consider the rationality and accuracy of the initial search text's expression. Additionally, by determining and displaying recommended search texts semantically related to the initial search text, users are guided to use these recommended search texts. This eliminates the need for users to repeatedly modify the initial search text to obtain more accurate search results, constructing a more natural content extension path. This enhances the smoothness and enjoyment of the search process, increases the search success rate, improves user satisfaction with the application and click-through rates on search results, thereby increasing trust in the application and dwell time.

[0141] Figure 12 This is a block diagram illustrating a search device according to an exemplary embodiment. See also Figure 12 The device includes: Search unit 1201 is configured to perform a search based on the first search text entered on the search page in response to a search instruction on the search page; Display unit 1202 is configured to display search extension information determined by a large language model on the search page; Search extension information includes semantic description text and recommended search text. Semantic description text describes the semantics of the first search text, while recommended search text is text that is semantically related.

[0142] In some embodiments, the apparatus further includes: The display unit 1202 is also configured to display the first search result corresponding to the first search text on the search page; and to display extended search information on the search page if the first search result meets the target conditions, wherein the target conditions refer to the conditions that the search results that do not meet the search requirements meet.

[0143] In some embodiments, the first search result includes multiple search results, and the display unit 1202 is configured to be at least one of the following: If the match between the last search result in multiple search results and the first search text is less than a first threshold, then search extension information is displayed on the search page. If the match between the search result ranked at the target position and the first search text is less than the second threshold among multiple search results, search extension information is displayed on the search page. If the match between the first search result ranked first among multiple search results and the first search text is less than the third threshold, display extended search information on the search page. If the maximum predicted click-through rate among multiple search results is less than the fourth threshold, display extended search information on the search page; If the type tags of multiple search results do not match the interest tags of the currently logged-in object, display extended search information on the search page.

[0144] In some embodiments, the display unit 1202 is configured to create a prompt word, the prompt word including a first search text, and the prompt word being used to instruct a large language model to generate search extension information of the first search text; the prompt word is input into the large language model to obtain search extension information.

[0145] In some embodiments, the apparatus further includes: The training module is used to obtain a pre-trained large language model; obtain training samples, which include a first sample text, a second sample text, and sample description text. The sample description text is used to describe the semantics of the first sample text, and the second sample text is text that is semantically related to the first sample text; and train the large language model based on the training samples.

[0146] In some embodiments, the search unit 1201 is further configured to perform a search based on the recommended search text in response to a selection operation on the recommended search text; and to display the second search result corresponding to the recommended search text.

[0147] In some embodiments, the search page includes a search results area displaying a first search result corresponding to a first search text; the search unit 1201 is configured to replace the first search result in the search results area with a second search result.

[0148] In some embodiments, the search unit 1201 is configured to display a dialogue page, which displays dialogue information between the currently logged-in object and the large language model; the dialogue page displays the recommended search text sent by the object, as well as the second search result replied by the large language model.

[0149] In some embodiments, the dialog page also displays an input box, and the device further includes: The question-and-answer unit is configured to retrieve the question text entered in the input box; display the question text issued by the object in the dialogue page; and display the answer text of the large language model in the dialogue page.

[0150] In some embodiments, the search extension information further includes descriptive text corresponding to the recommended search text; the apparatus further includes: The details view unit is configured to display the description text corresponding to the recommended search text on the search page in response to the details view operation of the description text. The description text includes the description text and other text.

[0151] In some embodiments, the first search text is a term name, the semantic description text describes the semantics of the term represented by the first search text, and the recommended search text is the name of a term whose semantics are related to the first search text; or, The first search text is a question; therefore, the recommended search text is the alternative answer text to the first search text; or, The first search text is item text. The semantic description text describes the item represented by the first search text. The recommended search text is the text corresponding to the item described in the first search text; or, The first search text belongs to the teaching text category. The semantic description text includes the learning path and at least one knowledge point in the learning path, where each knowledge point is a recommended search text.

[0152] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0153] This disclosure provides an electronic device including a processor and a memory for storing processor-executable instructions. The processor is configured to execute the instructions to implement the search method described above.

[0154] Figure 13 This is a block diagram illustrating an electronic device according to an exemplary embodiment. The electronic device 1300 may be: a smartphone, a tablet computer, an MP3 (Moving Picture Experts Group Audio Layer III) player, an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer, or a desktop computer. The electronic device 1300 may also be referred to as a user device, a portable terminal, a laptop terminal, a desktop terminal, or other names.

[0155] Typically, electronic device 1300 includes a processor 1301 and a memory 1302.

[0156] Processor 1301 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 1301 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 1301 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 1301 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 1301 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0157] The memory 1302 may include one or more computer-readable storage media, which may be non-transitory. The memory 1302 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 1302 are used to store at least one program code, which is executed by the processor 1301 to implement the search method provided in the method embodiments of this disclosure.

[0158] In some embodiments, the electronic device 1300 may optionally include a peripheral device interface 1303 and at least one peripheral device. The processor 1301, memory 1302, and peripheral device interface 1303 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 1303 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of the following: a radio frequency circuit 1304, a display screen 1305, a camera assembly 1306, an audio circuit 1307, and a power supply 1308.

[0159] Peripheral device interface 1303 can be used to connect at least one I / O (Input / Output) related peripheral device to processor 1301 and memory 1302. In some embodiments, processor 1301, memory 1302 and peripheral device interface 1303 are integrated on the same chip or circuit board; in some other embodiments, any one or two of processor 1301, memory 1302 and peripheral device interface 1303 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0160] The radio frequency (RF) circuit 1304 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The RF circuit 1304 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 1304 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals back into electrical signals. In some embodiments, the RF circuit 1304 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, etc. The RF circuit 1304 can communicate with other electronic devices via at least one wireless communication protocol. This wireless communication protocol includes, but is not limited to: metropolitan area networks (MANs), various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks (WLANs), and / or WiFi (Wireless Fidelity) networks. In some embodiments, the RF circuit 1304 may also include circuitry related to NFC (Near Field Communication), which is not limited herein.

[0161] Display screen 1305 is used to display a UI (User Interface). This UI may include graphics, text, icons, videos, and any combination thereof. When display screen 1305 is a touch display screen, it also has the ability to collect touch signals on or above its surface. These touch signals can be input as control signals to processor 1301 for processing. In this case, display screen 1305 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 1305, which serves as the front panel of electronic device 1300; in other embodiments, there may be at least two display screens, respectively disposed on different surfaces of electronic device 1300 or in a folded design; in still other embodiments, display screen 1305 may be a flexible display screen, disposed on a curved or folded surface of electronic device 1300. Furthermore, display screen 1305 may also be configured as a non-rectangular, irregular shape, i.e., a non-rectangular screen. The display screen 1305 can be made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0162] The camera assembly 1306 is used to acquire images or videos. In some embodiments, the camera assembly 1306 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the electronic device, and the rear-facing camera is located on the back of the electronic device. In some embodiments, there are at least two rear-facing cameras, which are any one of a main camera, a depth-sensing camera, a wide-angle camera, and a telephoto camera, to achieve background blurring by fusion of the main camera and the depth-sensing camera, panoramic shooting by fusion of the main camera and the wide-angle camera, VR (Virtual Reality) shooting, or other fusion shooting functions. In some embodiments, the camera assembly 1306 may also include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm-light flash and a cool-light flash, which can be used for light compensation at different color temperatures.

[0163] The audio circuit 1307 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, converting them into electrical signals that are input to the processor 1301 for processing, or to the radio frequency circuit 1304 for voice communication. For stereo sound acquisition or noise reduction purposes, multiple microphones may be used, each located at a different part of the electronic device 1300. The microphone may also be an array microphone or an omnidirectional microphone. The speaker is used to convert electrical signals from the processor 1301 or the radio frequency circuit 1304 into sound waves. The speaker may be a conventional diaphragm speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can convert electrical signals not only into audible sound waves but also into inaudible sound waves for purposes such as distance measurement. In some embodiments, the audio circuit 1307 may also include a headphone jack.

[0164] Power supply 1308 is used to supply power to various components in electronic device 1300. Power supply 1308 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 1308 includes a rechargeable battery, the rechargeable battery can support wired charging or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0165] Those skilled in the art will understand that Figure 13 The structure shown does not constitute a limitation on the electronic device 1300, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0166] Figure 14 This is a schematic diagram of a server structure according to an exemplary embodiment. The server 1400 can vary significantly due to different configurations or performance. It may include one or more Central Processing Units (CPUs) 1401 and one or more memories 1402. The memories 1402 store at least one computer program, which is loaded and executed by the processor 1401 to implement the methods provided in the various method embodiments described above. Of course, the server may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The server may also include other components for implementing device functions, which will not be elaborated upon here.

[0167] In an exemplary embodiment, a computer-readable storage medium is also provided, which, when executed by a processor of an electronic device, enables the electronic device to perform the steps in the search method described above. For example, the computer-readable storage medium may be a ROM (Read Only Memory), RAM (Random Access Memory), CD-ROM (Compact Disc Read-Only Memory), magnetic tape, floppy disk, and optical data storage device, etc.

[0168] In an exemplary embodiment, a computer program product is also provided, including a computer program that, when executed by a processor of an electronic device, implements the steps in the search method described above.

[0169] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0170] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.< / xxx>

Claims

1. A search method, characterized in that, The method includes: In response to a search command on the search page, a search is performed based on the first search text entered on the search page; The search page displays search extension information determined by the large language model; The search extension information includes semantic description text and recommended search text. The semantic description text is used to describe the semantics of the first search text, and the recommended search text is text related to the semantics.

2. The search method according to claim 1, characterized in that, The method further includes: The first search result corresponding to the first search text is displayed on the search page; The search page displays expanded search information determined by the large language model, including: If the first search result meets the target conditions, the search extension information is displayed on the search page. The target conditions refer to the conditions that the search results that do not meet the search requirements satisfy.

3. The search method according to claim 2, characterized in that, The first search result includes multiple search results. When the first search result meets the target conditions, displaying the extended search information on the search page includes at least one of the following: If the matching degree between the search result ranked last among the multiple search results and the first search text is less than a first threshold, the search extension information is displayed on the search page. If the matching degree between the search result ranked at the target position and the first search text is less than a second threshold among the multiple search results, the search extension information is displayed on the search page; If the matching degree between the first search result ranked first among the multiple search results and the first search text is less than a third threshold, the search extension information is displayed on the search page. If the maximum predicted click-through rate among the multiple search results is less than a fourth threshold, the search extension information is displayed on the search page. If the type tags of the multiple search results do not match the interest tags of the currently logged-in object, the search extension information is displayed on the search page.

4. The search method according to any one of claims 1-3, characterized in that, The search page displays expanded search information determined by the large language model, including: Create a suggestion word, the suggestion word including the first search text, and the suggestion word being used to instruct the large language model to generate search extension information for the first search text; The prompt words are input into the large language model to obtain the search extension information.

5. The search method according to any one of claims 1-3, characterized in that, The process of training the large language model includes: Obtain the pre-trained large language model; Obtain training samples, which include a first sample text, a second sample text, and a sample description text. The sample description text is used to describe the semantics of the first sample text, and the second sample text is text related to the semantics of the first sample text. The large language model is trained based on the training samples.

6. The search method according to any one of claims 1-3, characterized in that, The method further includes: In response to the selection operation of the recommended search text, a search is performed based on the recommended search text; Displays the second search result corresponding to the recommended search text.

7. The search method according to claim 6, characterized in that, The search page includes a search results area, which displays the first search result corresponding to the first search text. The display of the second search result corresponding to the recommended search text includes: Replace the first search result in the search results area with the second search result.

8. The search method according to claim 6, characterized in that, The display of the second search result corresponding to the recommended search text includes: Display a dialog page, which is used to display the dialogue information between the currently logged-in object and the large language model; The dialog page displays the recommended search text sent by the object, and the second search result replied by the large language model.

9. The search method according to claim 8, characterized in that, The dialog page also displays an input box, and the method further includes: Get the question text entered in the input box; The question text sent by the object is displayed on the dialog page; The dialogue page displays the answer text of the large language model.

10. The search method according to any one of claims 1-3, characterized in that, The search extension information also includes a summary text corresponding to the recommended search text; the method further includes: In response to a view of the details of the introductory text, the introductory text corresponding to the recommended search text is displayed on the search page. The introductory text includes the introductory text and other text.

11. The search method according to any one of claims 1-3, characterized in that, The first search text is a term name, the semantic description text describes the semantics of the term represented by the first search text, and the recommended search text is the name of a term that is semantically related to the first search text; or, The first search text is an interrogative sentence, and the recommended search text is the alternative answer text to the first search text; or, The first search text is item text, the semantic description text is used to describe the item represented by the first search text, and the recommended search text is text corresponding to the item that matches the description of the first search text; or, The first search text belongs to the category of teaching text, and the semantic description text includes a learning path and at least one knowledge point in the learning path, wherein each knowledge point is a recommended search text.

12. A search device, characterized in that, The device includes: The search unit is configured to perform a search based on the first search text entered on the search page in response to a search command on the search page; The display unit is configured to display search extension information determined by the large language model on the search page; The search extension information includes semantic description text and recommended search text. The semantic description text is used to describe the semantics of the first search text, and the recommended search text is text related to the semantics.

13. An electronic device, characterized in that, The electronic device includes: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the search method as described in any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the search method as described in any one of claims 1 to 11.

15. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the search method according to any one of claims 1 to 11.

Images