Content processing method and apparatus, and computer device, storage medium and program product

Abstract

Provided in the embodiments of the present application are a content processing method and apparatus, and a computer device, a storage medium and a program product. The content processing method comprises: displaying a service interface of a first client; receiving a content input operation in the service interface; in response to query symbols that are input by means of the content input operation, displaying an input method panel; and displaying in the input method panel a first query result corresponding to content that is input by means of the content input operation. By means of the embodiments of the present application, a query result corresponding to input content can be quickly acquired during a content input process.

Description

Content processing methods, apparatus and computer equipment, storage media, and program products

[0001] This application claims priority to Chinese Patent Application No. 202410813182.8, filed on June 21, 2024, entitled "Content Processing Method, Apparatus and Computer Equipment, Storage Medium and Program Product", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of computer technology, specifically to a content processing method, a content processing apparatus, a computer device, a computer-readable storage medium, and a computer program product. Background Technology

[0003] Currently, content input is widely used in various business scenarios, such as instant messaging and document editing. During content input, there is a need to query the input content. Therefore, how to quickly obtain the corresponding query results during the content input process has become a current research hotspot. Summary of the Invention

[0004] This application provides a content processing method, apparatus, computer device, storage medium, and program product that can quickly obtain query results corresponding to the input content during the content input process.

[0005] On one hand, embodiments of this application provide a content processing method, which includes:

[0006] Display the service interface of the first client;

[0007] Receive content input operations from the service interface;

[0008] In response to the query symbol entered in the content input operation, the input method panel is displayed; and the first query result corresponding to the content entered in the content input operation is displayed in the input method panel.

[0009] Accordingly, embodiments of this application provide a content processing apparatus, which includes:

[0010] The display unit is used to display the service interface of the first client.

[0011] The display unit is also used to receive content input operations in the service interface;

[0012] The display unit is also configured to display an input method panel in response to a query symbol entered in the content input operation; and to display a first query result in the input method panel corresponding to the content entered in the content input operation.

[0013] Accordingly, embodiments of this application provide a computer device, which includes:

[0014] A processor is a tool for implementing computer programs.

[0015] A computer-readable storage medium storing a computer program adapted to be loaded by a processor and executed by the above-described content processing method.

[0016] Accordingly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when read and executed by a processor of a computer device, causes the computer device to perform the aforementioned content processing method.

[0017] Accordingly, this application provides a computer program product comprising a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the aforementioned content processing method.

[0018] In this embodiment, a service interface of the first client can be displayed, and content input operations can be received in the service interface. Therefore, this embodiment can achieve human-computer interaction through the service interface, facilitating visual content input operations and providing corresponding operation feedback. In response to the query symbol entered during the content input operation, an input method panel can be displayed, and a first query result corresponding to the content entered can be displayed in the input method panel. Thus, this embodiment can quickly invoke the input method panel using the query symbol and display the query result corresponding to the input content. In other words, a query can be triggered by entering the query symbol to obtain and output the query result corresponding to the input content. The input of the query symbol is simple and fast. Therefore, this embodiment can quickly obtain the query result corresponding to the input content during the content input process by entering the query symbol, improving query efficiency and convenience. Attached Figure Description

[0019] Figure 1 is a schematic diagram of the architecture of a content processing system provided in an embodiment of this application;

[0020] Figure 2 is a flowchart illustrating a content processing method provided in an embodiment of this application;

[0021] Figure 3 is a schematic diagram of receiving content input in a service interface according to an embodiment of this application;

[0022] Figure 4a is a schematic diagram of the relationship between the input position of a query symbol and the content being queried, provided in an embodiment of this application.

[0023] Figure 4b is a schematic diagram of the relationship between the input position of a query symbol and the content being queried, provided in another embodiment of this application.

[0024] Figure 4c is a schematic diagram of the relationship between the input position of a query symbol and the content being queried, according to another embodiment of this application.

[0025] Figure 5 is a schematic diagram of the relationship between the input position of a query symbol and the content being queried, provided in another embodiment of this application.

[0026] Figure 6 is a schematic diagram of a query result with a clear intent provided in an embodiment of this application;

[0027] Figure 7 is a schematic diagram of a query result for a general intent provided in an embodiment of this application;

[0028] Figure 8 is a schematic diagram of an interface for further exploration of referenced information provided in an embodiment of this application;

[0029] Figure 9 is a schematic diagram of an interface for further exploration of relevant query content provided in an embodiment of this application;

[0030] Figure 10 is a schematic diagram of an interface for querying images provided in an embodiment of this application;

[0031] Figure 11 is a schematic diagram of a query failure result provided in an embodiment of this application;

[0032] Figure 12 is a schematic diagram of an interface for querying new content in the input method panel according to an embodiment of this application;

[0033] Figure 13 is a flowchart illustrating another content processing method provided in an embodiment of this application;

[0034] Figure 14a is a schematic diagram of an interface for activating a query function according to an embodiment of this application;

[0035] Figure 14b is a schematic diagram of another way to invoke the query function provided in an embodiment of this application;

[0036] Figure 15 is a schematic diagram of a streaming output interface for query results provided in an embodiment of this application;

[0037] Figure 16a is a schematic diagram of a text-based sharing interface for query results provided in an embodiment of this application;

[0038] Figure 16b is a schematic diagram of a query result sharing interface in image form provided in an embodiment of this application;

[0039] Figure 16c is a schematic diagram of a voice-based sharing interface for query results provided in an embodiment of this application;

[0040] Figure 16d is a schematic diagram of a link-based sharing interface for query results provided in an embodiment of this application;

[0041] Figure 16e is a schematic diagram of an interface for selecting the sharing type of query results provided in an embodiment of this application;

[0042] Figure 17 is a schematic diagram of an interface for sharing selected query results according to an embodiment of this application;

[0043] Figure 18 is a schematic diagram of the interaction process between a terminal and a backend server provided in an embodiment of this application;

[0044] Figure 19 is a flowchart illustrating another content processing method provided in an embodiment of this application;

[0045] Figure 20 is a schematic diagram of a query result generation process provided in an embodiment of this application;

[0046] Figure 21 is a schematic diagram of the structure of a content processing device provided in an embodiment of this application;

[0047] Figure 22 is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0048] This application relates to input methods, which are software used for content input. This application provides a content processing method that includes an innovative input method. This input method allows users to quickly obtain query results corresponding to the input content by entering query symbols during the content input process. More specifically, this input method can quickly invoke an AI (Artificial Intelligence) model by entering query symbols during the content input process. The AI ​​model then retrieves and presents the query results corresponding to the input content. The query operation triggered by entering query symbols allows content input users to conveniently resolve their current query needs using a simplified method of invoking the AI ​​model.

[0049] AI large-scale models refer to machine learning models with extremely large-scale parameters (usually more than one billion) and super-powerful computing resources. They are capable of processing massive amounts of data and completing various complex tasks, such as natural language processing and image recognition processing.

[0050] In this embodiment, the content processing method can be executed by a computer device, which runs the input method provided in this embodiment. The computer device can be, for example, a terminal of a content input user, which may run the input method provided in this embodiment. Alternatively, the computer device can be, for example, a content processing system composed of a terminal and a backend server, where the terminal may run a frontend of the input method provided in this embodiment, and the backend server may be a backend of the input method provided in this embodiment, used to provide services to the frontend.

[0051] In one implementation, the computer device can be the terminal of the content input user, and the AI ​​big model can be deployed on the content input user's terminal. When the content input user is using the input method running on the terminal to input content, they can quickly invoke the AI ​​(Artificial Intelligence) big model deployed on the terminal by inputting query symbols, and call the AI ​​big model to obtain and present the query results corresponding to the input content.

[0052] In another implementation, the computer device can be a content processing system consisting of a terminal used by the content input user and a backend server. The AI ​​big data model can be deployed in the backend server of the input method. As shown in Figure 1, the content processing system can include a terminal 101 used by the content input user and a backend server 102 for the input method. In the content processing system, when the content input user is inputting content using the input method running on terminal 101, if terminal 101 detects that the user has entered a query symbol using the input method, it can send a query request to the backend server 102. The backend server 102 can then call the deployed AI big data model to obtain the query results corresponding to the input content and send these results back to terminal 101. Terminal 101 can then display the query results corresponding to the input content to the content input user.

[0053] The terminal 101 mentioned in the embodiments of this application may be a smartphone, tablet computer, laptop computer, desktop computer, smart voice interaction device, smartwatch, vehicle terminal, smart home appliance, aircraft, etc., but is not limited to these; that is, the input method provided in the embodiments of this application can be applied to PC (Personal Computer) terminals (e.g., laptop computers, desktop computers, etc.) and also to mobile terminals (e.g., smartphones, tablet computers, smartwatches, vehicle terminals, etc.).

[0054] The backend server 102 mentioned in this application embodiment can be a single physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms, as described herein.

[0055] It is understood that the content processing system described in the embodiments of this application is for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and does not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0056] This application provides a content processing method, which includes: the content contained in the query results, further exploration of the query results, and different cases of the queried content when the query symbol is input to different positions in the input content. As shown in Figure 2, the content processing method may include, but is not limited to, the following steps S201-S204:

[0057] S201 displays the service interface of the first client.

[0058] The first client can be a client that requires content input, such as a document editing client or an instant messaging client, etc.; the input method used by the first client is the input method provided in the embodiments of this application. The embodiments of this application do not limit the type of client; the client can be, but is not limited to, any of the following: applications, applets, software, and web (World Wide Web) applications, etc. The service interface can be an interface in the first client that requires content input; for example, the service interface can be the document editing interface in a document editing client, where content input is required and the input content can be used as document content; or, for example, the service interface can be the instant messaging interface in an instant messaging client, where content input is required and the input content can be used as the message content of an instant messaging message.

[0059] S202, receive content input operations in the service interface.

[0060] Receiving content input operations in the service interface can include receiving content input operations in the input boxes of the service interface. For example, in Figure 3(a), the text editing interface 30 can include an input box 301, which can receive content input operations in the input box 301 of the text editing interface 30, and the input content is "climate type"; and in Figure 3(b), the instant messaging interface 31 can include an input box 301, which can receive content input operations in the input box 301 of the instant messaging interface 31, and the input content is "climate type".

[0061] S203, in response to the query symbol entered during the content input operation, the input method panel is displayed.

[0062] S204, Display the first query result corresponding to the content entered in the content input operation in the input method panel.

[0063] In steps S203-S204, the input method panel can be displayed in response to the query symbol entered during the content input operation. The input method panel can display the first query result corresponding to the content entered during the content input operation. Specifically, displaying the input method panel in response to the query symbol entered during the content input operation can include: displaying the input method panel in response to the query symbol entered after the target content is entered during the content input operation. The first client uses the input method provided in this application embodiment, and the input method panel is the interface provided by the input method used by the first client. The query symbol refers to a symbol that is used less frequently in content input scenarios. This application embodiment uses "=" as an example for explanation. "=" is used less frequently in content input scenarios. The scenario in which the content inputter enters "=" is usually a calculation scenario or a code editing scenario. Therefore, this method of evoking a query symbol strikes a very good balance between "usefulness" and "distraction." Besides "=", the query symbol can also be other symbols that are used less frequently in content input scenarios; this application embodiment does not limit this.

[0064] It's worth noting that the content retrieved can differ depending on where the query symbol is entered within the target content. Specifically:

[0065] In one implementation, when a query symbol is entered at any position within the target content, the content being queried is the target content. In this implementation, "any position" can be, for example, the end of the target content, the beginning of the target content, or within the target content. Based on this, displaying the first query result corresponding to the content entered in the input method panel can include: displaying the first query result corresponding to the target content in the input method panel.

[0066] For example, in Figure 4a, the target content 3011 entered in the input box 301 is "climate type". In response to the query symbol 3012 entered at the end of the target content 3011, the input method panel 32 can be displayed, and the first query result 321 corresponding to the target content 3011 can be displayed in the input method panel 32. As another example, in Figure 4b, the target content 3011 entered in the input box 301 is "climate type". In response to the query symbol 3012 entered at the beginning of the target content 3011, the input method panel 32 can be displayed, and the first query result 321 corresponding to the target content 3011 can be displayed in the input method panel 32. For example, in Figure 4c, the target content 3011 entered in the input box 301 is "climate type". In response to the query symbol 3012 entered in the target content 3011 (for example, in Figure 4c, the query symbol 3012 is entered between "climate" and "type"), the input method panel 32 can be displayed, and the first query result 321 corresponding to the target content 3011 can be displayed in the input method panel 32.

[0067] In another implementation, when the query symbol is at the beginning or end of the target content, the first query result corresponding to the target content can be displayed in the input method panel, as shown in Figures 4a and 4b. When the query symbol is within the target content, the first query result corresponding to the first content can be displayed in the input method panel. The first content is the content within the target content preceding the query symbol. For example, in Figure 5, the target content 3011 entered in the input box 301 is "climate type". In response to the query symbol 3012 entered in the target content 3011 (e.g., in Figure 5, the query symbol 3012 is entered between "climate" and "type"), the input method panel 32 can be displayed, and the first query result 321 corresponding to the content "climate" preceding the query symbol 3012 can be displayed in the input method panel 32.

[0068] The first query result may include at least one of the following: answer content corresponding to the content entered in the content input operation, citation information of the answer content, and query content semantically related to the content entered in the content input operation. Here, answer content refers to the content that answers a question, using the content entered in the content input operation as the question. Citation information of the answer content refers to the descriptive information of the cited content; for example, the cited content could be books, online diaries (e.g., blogs, posts, and articles), etc.; descriptive information of the cited content refers to information that introduces the cited content. For example, when the cited content is a book, the descriptive information could be the book title, author, and cited chapter, etc.; similarly, when the cited content is an online diary, the descriptive information could be the diary title and author, etc. Query content semantically related to the content entered in the content input operation can be understood as content that semantically expands upon the content entered in the content input operation. That is, semantically related means that the semantics of the query content includes the semantics of the input content and is expanded based on the semantics of the input content. For example, if the input content is "climate type," the query content could be "what are the differences between various climate types."

[0069] Optionally, the types of result content included in the query results are related to the query intent of the content entered in the content input operation. Specifically, the input method to which the input method panel belongs has set multiple preset intents. These preset intents can include explicit intents and general intents. Explicit intents refer to the intents used to obtain specific types of query results, while general intents refer to the intents used to obtain general query results. When the query intent of the content entered in the content input operation is an explicit intent, the first query result can include the answer content corresponding to the content entered in the content input operation. When the query intent of the content entered in the content input operation is a general intent, the first query result can include the answer content corresponding to the content entered in the content input operation, the reference information of the answer content, and query content that is semantically related to the content entered in the content input operation.

[0070] For example, as shown in Figure 6, the target content 3011 entered in the input box 301 is "Beijing weather". The query intent of this content is to obtain the query result of weather type. The query intent of this content is explicit. In response to the query symbol 3012 entered at the end of the target content 3011, the input method panel 32 can be displayed. The first query result 321 displayed in the input method panel 32 includes the answer content 3211. As shown in Figure 7, the target content 3011 entered in the input box 301 is "climate type". The query intent of this content is more general. The query intent of this content is general. In response to the query symbol 3012 entered at the end of the target content 3011, the input method panel 32 can be displayed. The first query result 321 displayed in the input method panel 32 can include the answer content 3211, the reference information 3212 of the answer content 3211, and the query content 3213 that is semantically related to the target content 3011.

[0071] In this way, when the query intent of the input content is to obtain specific information, such as the weather, the query results only include the answer content. The answer content is sufficient to provide the specific information that the user wants, thus quickly providing the desired query results. If the query results also include citations and query content, the results become redundant, as this information may not be what the user wants, reducing the accuracy and efficiency of obtaining the query results. When the query intent of the input content is a general intent, the query results can include the answer content, citations to the answer content, and related query content. The citations to the answer content are more reliable and comprehensive introductions to books, etc., providing the user with more reliable and comprehensive query results, thus improving the accuracy of the query results.

[0072] It should be noted that, due to the size limitations of the input method panel, the query results may not be fully displayed in the input method panel. In this case, you can view the complete query results by sliding your finger, scrolling with the mouse wheel, or sliding the scroll bar in the input method panel (for example, the input method panel 32 shown in Figure 7 may include a scroll bar 322). Figure 7 shows the complete query results.

[0073] Furthermore, the query results can support further exploration. The content supporting further exploration within the query results can include citation information and related query content. Specifically:

[0074] (1) Further exploration of the cited information:

[0075] In response to a triggering operation on the citation information of the answer content, the original text of the citation content can be displayed. Specifically, the answer content may include one or more sub-contents. Each target sub-content containing citation content in the answer content is marked with a citation tag. The citation information may include the citation information of each target sub-content, which refers to the descriptive information of the citation content referenced by the corresponding target sub-content. The citation tag of the citation information of each target sub-content corresponds to the citation tag added to it. In response to a triggering operation on the citation information of any target sub-content, the original text of the citation content referenced by that target sub-content can be displayed.

[0076] Furthermore, the source of the quoted content in the answer can be a second client. This second client is different from the first client and is not provided by the input method. In response to a trigger operation on the quoted information, the original text of the quoted content can be displayed in the second client. It can be seen that knowledge sharing between different clients is achieved through the input method. The input method can integrate knowledge from different clients to generate answer content, thus improving the accuracy of the answer.

[0077] For example, as shown in Figure 7, the first query result 321 displayed in the input method panel 32 corresponding to the target content 3011 may include the answer content 3211, the reference information 3212 of the answer content 3211, and the query content 3213 semantically related to the target content 3011. The five target sub-contents in the answer content 3211 have reference content, and the five target sub-contents are marked with reference markers "①-⑤" in the order of their appearance. The reference information includes the reference information of the five target sub-contents, and the reference markers of the reference information of the target sub-contents correspond to the reference markers added to the target sub-contents. For example, the reference marker "①" added to "content 1" corresponds to the reference marker "①" of the reference information of "content 1".

[0078] As shown in Figure 8, the second client can be, for example, a reading client. In response to a triggering operation on the citation information 801 corresponding to the citation mark "①" (e.g., the citation information in Figure 8 stating "cited from Chapter 1 of Book 1 by author 1"), the original text corresponding to the sub-content marked by citation mark "①" (e.g., "Content 1" in Figure 8) can be displayed on the original text details page 802 of the second client. Furthermore, the sub-content marked by citation mark "①" can be highlighted in the original text (e.g., "Content 1" in Figure 8 is highlighted by adding a dashed box).

[0079] In this way, content input users can continue to explore the cited information and view the original text of the cited text. Taking books as an example, the search results display the content of the books cited in the answer, which can be used for further reference by the content input user, or to view the original text of the cited books. This helps the content input user to quickly obtain more information related to the content entered in the content input operation.

[0080] (2) Further exploration of related query content:

[0081] In response to a query operation, a second query result corresponding to the query content can be displayed in the input method panel. As shown in Figure 9, the input target content 3011 is "climate type", and there are four query contents in the query results corresponding to the input target content 3011. If the first query content 901 (for example, "What are the differences between various climate types?" in Figure 9) is triggered, the query question in the query box 325 in the input method panel 32 is replaced by the first query content 901 instead of the target content 3011, and a second query result 902 corresponding to the first query content 901 can be displayed in the input method panel 32.

[0082] In this way, content input users can continue to explore the query content related to the input content and explore new knowledge, which helps them to quickly obtain more information related to the content input operation.

[0083] The triggering operation in the embodiments of this application may include, but is not limited to, any of the following: single-click operation, double-click operation, and long-press operation, as described herein. The being triggered in the embodiments of this application may include, but is not limited to, any of the following: being clicked, being double-clicked, and being long-pressed, as described herein.

[0084] The above steps S201-S204 use text content as an example to introduce the content processing method provided by the embodiments of this application. Optionally, the content processing method provided by the embodiments of this application is also applicable when the content input operation is other types of content (e.g., images, voice, and video, etc.). As shown in Figure 10, taking the content input in the input box 301 as an example, if the content input is an image 1001 (e.g., an image containing a dog), in response to the query symbol 3012 entered at the end of the image 1001, the input method panel 32 can be displayed, and the query result 1002 corresponding to the image 1001 can be displayed in the input method panel 32 (e.g., the query result about dogs in Figure 10). In this way, it is possible to support querying different types of content, and query results corresponding to different types of content can be quickly obtained by entering query symbols. The cases where the content input operation is video or voice are similar to the cases where the content input operation is an image, and will not be described again here.

[0085] The above steps S201-S204, taking the successful query of the content entered in the content input operation as an example, introduce the content processing method provided by the embodiments of this application. Optionally, if the query of the content entered in the content input operation fails, a query failure prompt can be output in the input method panel. Specifically, after the input method panel is displayed in response to the query symbol entered in the content input operation, if the content entered in the content input operation meets the query conditions, the first query result corresponding to the content entered in the content input operation can be triggered and displayed in the input method panel; if the content entered in the content input operation does not meet the query conditions, a query failure prompt can be displayed in the input method panel. The following are some of the reasons why the content input operation does not meet the query conditions: the number of queries performed by the content input operator within the statistical time period exceeds the query threshold (e.g., triggering queries 20 times by entering query symbols within 1 minute, exceeding the query threshold of 15 times); the content input operator is an illegal object (i.e., a blacklisted object); the content input operation contains illegal words; the content input operator does not have query permissions (e.g., the content input operator has not enabled query permissions); and the content input operation is meaningless and cannot match the preset intent of the input method.

[0086] As shown in Figure 11, when the search criteria are not met, a search failure message 1101 can be output in the input method panel 32 (for example, the message "No relevant answer" in Figure 11). In this way, the search function of the input method can be controlled to avoid wasting search resources by frequently searching for entertainment, searching for illegal content, searching for illegal objects, or making meaningless searches.

[0087] The content processing method provided in this application embodiment can also support using the input method panel as a query panel after it is invoked, allowing users to continue querying new content. Specifically, as shown in Figure 12, the input method panel 32 displays a first query result 321 corresponding to the input target content 3011 (the target content 3011 is "climate type"). The input method panel 32 may also include a query box 325 and a first query control 326. The query box 325 displays the target content 3011. The user can delete the target content 3011 in the query box 325, re-enter new content 1201 (for example, the re-entered new content is "Beijing weather"), and query the new content 1201 by triggering the first query control 326. The input method panel 32 then displays a query result 1202 corresponding to the new content 1201. In this way, the user can perform multiple queries during the content input process without having to end the content input process and jump to a dedicated content query terminal, thus improving the efficiency of content querying during the content input process.

[0088] Furthermore, this application embodiment can also block clients with frequent needs for inputting query symbols by triggering the query through inputting query symbols. In detail, in response to the query symbol input during the content input operation, if the business type of the first client is a first business type, the input method panel can be displayed. The first business type refers to a business type with low-frequency query symbol input. Here, low frequency means that the cumulative number of query symbols input in the first client is less than or equal to a threshold. Specifically, it means that within a preset time period before responding to the query symbol input during the content input operation, the cumulative number of query symbols input in the first client is less than or equal to the threshold. For example, the first business type may include document editing business, instant messaging business, etc., which have a low frequency of query symbol input during business execution. If the business type of the first client is a second business type, the input method panel is not displayed. The second business type refers to a business type with high-frequency query symbol input. Here, high frequency means that the cumulative number of query symbols input in the first client is greater than a threshold. Specifically, it means that within a preset time period before responding to the query symbol input during the content input operation, the cumulative number of query symbols input in the first client is greater than the threshold. For example, the second business type may include code editing business, formula editing business, calculator business, etc., which have a high frequency of query symbol input during business execution. In this way, it is possible to effectively avoid situations where clients with frequent needs to input query symbols trigger the query function due to incorrect input of query symbols.

[0089] Optionally, embodiments of this application may also set different query symbols for clients of the first service type and clients of the second service type. The first query symbol set for the client of the first service type is used less frequently in the client of the first service type, and the second query symbol set for the client of the second service type is used less frequently in the client of the second service type. This allows both the client of the first service type and the client of the second service type to quickly invoke the query function and obtain query results by entering the query symbol.

[0090] In this embodiment, query results can be obtained by inputting query symbols. Inputting query symbols is quick and easy, allowing for rapid retrieval of results during content input. This embodiment utilizes query symbols to obtain results during content input. The advantage of this method is that when a user has a query need during input, they do not need to end the input process and switch to a dedicated query tool. The user can quickly and conveniently access the query function and obtain results during the input process, thus improving query efficiency.

[0091] This application provides a content processing method, which includes sharing query results. As shown in Figure 13, the content processing method may include, but is not limited to, the following steps S1301-S1305:

[0092] S1301 displays the service interface of the first client.

[0093] In this embodiment, the execution process of step S1301 is the same as that of step S201 in the embodiment shown in FIG2 above. For details, please refer to the execution process of step S201 in the embodiment shown in FIG2 above, which will not be repeated here.

[0094] S1302, receives content input operations in the service interface.

[0095] In this embodiment, the execution process of step S1301 is the same as that of step S201 in the embodiment shown in FIG2 above. For details, please refer to the execution process of step S201 in the embodiment shown in FIG2 above, which will not be repeated here.

[0096] S1303, in response to the query symbol entered during the content input operation, displays the input method panel.

[0097] In step S1303, the input method panel can be displayed in response to the query symbol entered during the content input operation. That is, the query function can be invoked by entering the query symbol. Besides entering the query symbol, other methods can also be used to invoke the query function: for example, the query function can be invoked in response to a trigger operation on the query control (here, the query control refers to the second query control); as shown in Figure 14a, the input box 301 may include the second query control 3013. If the target content 3011 entered in the input box 301 during the content input operation is "climate type", the input method panel 32 can be displayed in response to a trigger operation on the second query control 3013. Alternatively, the query function can be invoked in response to an input query gesture; as shown in Figure 14b, if the target content 3011 entered in the input box 301 during the content input operation is "climate type", the input method panel 32 can be displayed in response to a query gesture entered in the input box 301 (for example, an "S" shaped gesture in Figure 14b). This approach allows for the use of various methods to query the input content, all of which are convenient and fast, enabling the retrieval of query results corresponding to the input content in a variety of ways.

[0098] S1304, Display the first query result corresponding to the content entered during the content input operation in the input method panel.

[0099] The first query result corresponding to the content entered in the content input operation can be displayed segment by segment in the input method panel. As shown in Figure 15, the first query result is displayed segment by segment in the input method panel 32, that is, the embodiment of this application supports the "whole segment output" streaming output of query results. Compared with most word-by-word output methods, the "whole segment output" method makes the query results more readable.

[0100] The input method to which the input method panel belongs (i.e., the input method used by the first client) has multiple preset intents. The first query result is the query result under the preset intent with the highest relevance to the query intent of the content input operation. Specifically, displaying the first query result corresponding to the content input operation in the input method panel can include: displaying the query result corresponding to the content input operation under the first preset intent in the input method panel; the first preset intent is the preset intent with the highest relevance to the query intent of the content input operation among multiple preset intents. Here, relevance refers to the similarity between the query intent and the preset intent; higher similarity means higher relevance. For example, if the input method has preset intents such as stocks, weather, exchange rate, translation, and book search (i.e., searching in books), and the content input in the content input operation is "Beijing weather," then the output will be the query result under the "weather" intent; as another example, if the content input operation is "Hello English," then the output will be the query result under the "translation" intent; and as yet another example, if the content input operation is "climate type," then the output will be the query result under the "book search" intent. In this way, the search results with the highest relevance to the search intent of the input content are output, allowing the content input user to quickly obtain the expected search results.

[0101] S1305, share the first query result.

[0102] In this embodiment, sharing of query results is supported. The first query result in the input method panel can be shared, where sharing refers to using the first query result. For example, in a document editing scenario, sharing means using the first query result for document editing; similarly, in an instant messaging scenario, sharing means using the first query result for instant messaging.

[0103] The first query result can be displayed in text format. Sharing the first query result can include any of the following methods: First, sharing the first query result in text format; sharing in text format means using the first query result presented in text format as text. Second, sharing the first query result in image format; sharing in image format means converting the first query result presented in text format into image format before use. Third, sharing the first query result in voice format; sharing in voice format means converting the first query result presented in text format into voice format before use. Fourth, sharing the first query result in link format; sharing in link format means converting the first query result presented in text format into link format before use.

[0104] As shown in Figure 16a, after selecting the first query result in the input method panel 32 (for example, the selected first query result is highlighted with a dashed box in Figure 16a), the first query result can be copied (for example, by entering a copy command on the keyboard (for example, the copy command can be Ctrl+C)). After copying, the first query result can be pasted as text into the input box of the instant messaging interface for sharing as an instant messaging message, or the first query result can be pasted as text into the input box of the document editing interface for document editing content.

[0105] As shown in Figure 16b, the input method panel 32 may include an image copy control 327 (e.g., a control named "Copy as Image" in Figure 16b). In response to a trigger operation on the image copy control 327, the first query result can be converted into a query result image 1601. The query result image 1601 can be pasted into the input box of the instant messaging interface for sharing as an instant messaging message, or the query result image can be pasted into the input box of the document editing interface for use as document editing content.

[0106] As shown in Figure 16c, the input method panel 32 may include a voice copy control 328 (for example, a control named "Copy as Voice" in Figure 16c). In response to the triggering operation of the voice copy control 328, the first query result can be converted into query result voice 1602, and the query result voice 1602 can be pasted into the input box of the instant messaging interface for sharing as an instant messaging message.

[0107] As shown in Figure 16d, the input method panel 32 may include a link copy control 329 (e.g., a control named "Copy as Link" in Figure 16d). In response to a trigger operation on the link copy control 329, the first query result can be converted into a query result link 1603. The query result link 1603 can be pasted into the input box of the instant messaging interface for sharing as an instant messaging message. Alternatively, the query result image can be pasted into the input box of the document editing interface as document editing content.

[0108] As shown in Figure 16e, the input method panel 32 may include a copy control 330 (e.g., the control named "Copy" in Figure 16e). In response to a triggering operation on the copy control 330, a sharing type list 331 can be displayed, allowing selection of the sharing format for the first query result (e.g., text, image, voice, and link, etc.). For example, when the selected sharing format is an image, the first query result can be converted into a query result image, which can then be pasted into an input box in the instant messaging interface for sharing as an instant messaging message. Alternatively, the query result image can be pasted into an input box in the document editing interface for use as document editing content.

[0109] By sharing query results, the results can not only be viewed but also used, improving their usability and meeting the needs of content users. Furthermore, query results can be used in different formats (e.g., text, images, voice, and links), enriching their usage and further enhancing their usability.

[0110] The above sharing of query results refers to the sharing of the entire query results. This application embodiment also supports partial sharing of query content. Specifically, the shared result content can be selected from the first query results. As shown in Figure 17, after selecting a portion of the result content from the first query results in the input method panel 32 (for example, the selected result content in Figure 17 is highlighted with a dashed box), in response to the triggering operation of the copy control 330 (for example, the control named "Copy" in Figure 17), a sharing type list 331 can be displayed. The sharing type list 331 allows selection of the sharing format of the selected result content (e.g., text, image, voice, and link, etc.).

[0111] It should be noted that Figure 17 illustrates the process of partially sharing query results by taking the example of selecting a sharing method after selecting the results. Besides this method, similar to sharing all query results, the selected results can be shared as text by entering a copy command, as an image by using a separate image copy control, as an audio copy control, or as a link by using a separate link copy control. It's easy to see that the content inputter may only need to use a portion of the query results. Therefore, sharing the selected results allows the content inputter to use the query results on demand, making the use of the query results more flexible.

[0112] In this embodiment, query results can be obtained by inputting query symbols. Inputting query symbols is quick and easy, allowing for rapid retrieval of results during content input. Sharing the query results not only allows them to be viewed but also further utilized, thus improving their usability.

[0113] Before detailing the content processing method provided in this application from a technical perspective, the interaction flow between the content input user's terminal and the backend server when executing the content processing method is first described. The interaction flow between the content input user's terminal and the backend server when executing the content processing method is shown in Figure 18, and may include:

[0114] (1) The terminal receives the content input operation from the content input user. If it detects that the content input operation has entered a query symbol, it can send a query request to the backend server and simultaneously display the input method panel (the input method panel can also be called the AI ​​panel). The query request can carry the content entered by the content input operation (hereinafter referred to as the input content).

[0115] (2) After receiving a query request, the backend server can perform a first validity check on the query request to confirm whether the query request is valid. If the query request is valid, (3) can be executed. If the request is invalid, the query service can be refused, and a query failure message can be displayed in the input method panel.

[0116] The first legitimacy check here refers to verifying the identity and permissions of the query request. This verification determines whether the query request was sent by the input method running on the terminal. If so, the query request is considered legitimate; otherwise, it is considered illegitimate.

[0117] (3) The backend server can perform a second validity check on the input content carried in the query request to confirm whether the input content is valid. If the input content is valid, (4) can be executed. If the input content is invalid, the query service can be refused, and a query failure message can be displayed in the input method panel.

[0118] The second legality check here refers to checking whether there are illegal words in the input content (for example, illegal words refer to sensitive words). If there are illegal words in the input content, it can be determined that the input content is illegal. If there are no illegal words in the input content, it can be determined that the input content is legal. In detail, the backend server can maintain a cache of a list of illegal words (for example, a list of sensitive words). If it finds that the list has expired, it can pull the latest list of illegal words from the CMS (Content Management System). The backend server can use the AC (Aho-Corasick) algorithm to check the input content to determine whether it contains illegal words. If the input content contains illegal words, it can be determined that the input content is illegal, and the query service can be refused. The query failure prompt can be output in the input method panel. If the input content does not contain illegal words, it can be determined that the input content is legal, and (4) can be executed.

[0119] (4) The backend server can perform intent recognition on the input content.

[0120] In detail, the backend server can call the intent recognition model to perform intent recognition on the input content. If one or more matching intents of the input content are determined among multiple preset intents, (5) can be executed; if there is no matching intent of the input content among multiple preset intents, the query service can be refused and a query failure prompt can be output in the input method panel. Among them, multiple preset intents are set by the input method, and each preset intent corresponds to its own query model. The query model refers to the AI ​​big model; multiple preset intents are, for example, preset intent 1-preset intent n in Figure 18, where n is an integer greater than 1.

[0121] (5) The backend server can call the query model corresponding to each matching intent through the interface corresponding to each matching intent to perform result query on the input content and obtain the query results of the input content under each matching intent.

[0122] (6) The backend server can sort the query results for each matching intent based on the input content.

[0123] In detail, the backend server can obtain the relevance between the input content and the query results under each matching intent; the backend server can also obtain the completeness of the query results under each matching intent; the backend server can determine the ranking weight of the query results under each matching intent based on the relevance between the input content and the query results under each matching intent, as well as the completeness of the query results under each matching intent, and sort the query results under each matching intent according to the ranking weight of the query results under each matching intent.

[0124] (7) The backend server can determine the first query result corresponding to the input content based on the sorting result, and return the first query result to the terminal, which will then display the first query result to the content input user.

[0125] The following, with reference to Figure 19, details the content processing method provided in this application from a technical perspective. This content processing method may include, but is not limited to, the following steps S1901-S1908:

[0126] S1901 displays the service interface of the first client.

[0127] S1902, receive content input operations in the service interface.

[0128] S1903, in response to the query symbol entered during the content input operation, the input method panel is displayed.

[0129] S1904, perform intent recognition on the content input by the content input operation, and determine one or more matching intents of the content input by the content input operation from multiple preset intents.

[0130] S1905, call the query model corresponding to each matching intent, perform result query on the content input operation, and obtain the query results of the content input operation under each matching intent.

[0131] Any one of one or more matching intents can be represented as the target matching intent.

[0132] When the target matching intent is explicit, the process of calling the query model corresponding to the target matching intent to query the results of the content input operation (hereinafter referred to as input content) and obtaining the query results of the content input operation under the target matching intent may include: calling the query model corresponding to the target matching intent, obtaining the answer content corresponding to the input content, and determining the answer content corresponding to the input content as the query result of the input content under the target matching intent.

[0133] When the target matching intent is a general intent, the process of calling the query model corresponding to the target matching intent to query the content entered in the content input operation (hereinafter referred to as the input content) and obtaining the query results of the content entered in the content input operation under the target matching intent may include:

[0134] (1) Call the query model corresponding to the target matching intent to generate the answer content corresponding to the content input operation.

[0135] In detail, the input content of the content input operation can be retrieved to obtain retrieved content. If there are books related to the input content in the retrieved content, the query model corresponding to the target matching intent can be invoked to generate answer content based on the book's summary content. If there are no books related to the input content in the retrieved content, and the number of retrieved content is less than or equal to a first quantity threshold, the query model corresponding to the target matching intent can be invoked to generate answer content based on the input content. If the number of retrieved content is greater than the first quantity threshold, the query model corresponding to the target matching intent can be invoked to generate answer content based on the retrieved content.

[0136] The recalled content can be retrieved from books. The books are related to the input content, meaning that the titles of the books are similar to the input content.

[0137] (2) Call the query model corresponding to the target matching intent to recall the reference content of the answer content.

[0138] In detail, if the answer content is generated based on the recalled content corresponding to the content input operation, then the recalled content with a similarity greater than the first similarity threshold to the answer content can be identified as the cited content of the answer content, and a citation relationship can be established. If the answer content is not generated based on the recalled content corresponding to the content input operation, then the query model corresponding to the target matching intent can be invoked to recall books related to the answer content based on the answer content, and the books among the recalled books with a similarity greater than the second similarity threshold to the answer content can be identified as the cited content of the answer content, and a citation relationship can be established.

[0139] (3) Call the query model corresponding to the target matching intent to generate query content that is semantically related to the content input by the content input operation.

[0140] In detail, if the answer content is generated based on the recall content corresponding to the content input operation, then the query model corresponding to the target matching intent can be called to generate query content that is semantically related to the content input operation based on the recall content and the content input operation. If there are books related to the content input operation in the recall content, then the query content associated with the books can be used as the query content that is semantically related to the content input operation.

[0141] (4) The answer content corresponding to the content entered in the content input operation, the reference information of the reference content of the answer content, and the query content that is semantically related to the content entered in the content input operation are determined as the query results of the content entered in the content input operation under the target matching intent.

[0142] S1906, Sort the query results of the content entered in the content input operation under each matching intent, and determine the first query result corresponding to the content entered in the content input operation.

[0143] In detail, the ranking process may include: obtaining the relevance between the input content and the query results under each matching intent, where relevance refers to the semantic connection between the input content and the query results under each matching intent; obtaining the completeness of the query results under each matching intent, which can be determined based on the repetition score (the ratio of the number of repeated information in the query results to the total number of characters in the query results), the context consistency score (obtained by training a proprietary model to evaluate the input content), and the human preference score (obtained by training a model simulating human preferences to evaluate the input content); determining the ranking weight of the query results under each matching intent based on the relevance between the input content and the query results under each matching intent, as well as the completeness of the query results under each matching intent; and ranking the query results under each matching intent according to the ranking weight. The higher the ranking weight of a query result under any matching intent, the higher the relevance between the query result under that matching intent and the input content, and the higher the probability that the query result under that matching intent will be the final query result of the input content.

[0144] After sorting, the query result with the highest sort weight among the query results for each matching intent can be determined as the first query result corresponding to the input content. When sorting the query results under each matching intent in descending order of sort weight, the query result ranked first is the one with the highest sort weight. When sorting the query results under each matching intent in ascending order of sort weight, the query result ranked last is the one with the highest sort weight.

[0145] S1907, Display the first query result corresponding to the content entered during the content input operation in the input method panel.

[0146] S1908, share the first query result.

[0147] In this embodiment of the application, the query model corresponding to each preset intent can be an AI big model. By combining the generation capability of the AI ​​big model with the content input process, the AI ​​big model can be used to quickly generate query results corresponding to the input content during the content generation process. Thus, the query efficiency during the content input process can be improved by using the AI ​​big model.

[0148] Figure 19 above details the query result generation logic. It's important to note that the large model's process for generating queries is typically lengthy. Waiting for the large model to generate all query results before returning them to the terminal would result in excessively long waiting times for the content input user. To avoid this, a single query request from the content input user can be split into several requests by the terminal. After each request, the backend server returns partial query results. Specifically, when the terminal initiates a query request, the backend server generates a session identifier (sessionid) to represent a complete request. For the first request in the terminal's query request, the backend server generates a session identifier (sessionid) and returns it to the terminal. Subsequent requests within the terminal's query request can carry this session identifier (sessionid).

[0149] For the backend server, the session ID can be used to determine whether the request is the first or a subsequent request. Since the process of a large model is usually quite long, we designed the entire process as an incremental update process. The first request is mainly used to retrieve content; a start flag is designed, and a response packet is sent to the terminal. The terminal can determine whether the process is complete or continues the request based on the completion flag in the response packet. See Figure 20 for details:

[0150] I. For the initial request:

[0151] (1) Legality judgment:

[0152] If it is the first request, the first request can be judged for legality. The legality judgment can include frequency judgment, blacklist judgment and anti-spam judgment. Frequency judgment refers to judging whether the number of queries by the content input user in the statistical period is within the query threshold. Blacklist judgment refers to judging whether the content input user is a blacklist object. Anti-spam judgment refers to checking whether there are illegal words in the input content (query). The legality judgment process can include: judging whether the number of queries by the content input user in the statistical period is within the query threshold; if yes, the first request is an illegal request and the first request can be rejected; if no, it can be judged whether the content input user is a blacklist object; if yes, the first request is an illegal request and the first request can be rejected; otherwise, anti-spam judgment can be performed on the input content. Anti-spam judgment refers to checking whether there are illegal words in the input content (query); if no, the first request can be determined to be a legal request and (2) can be executed; if there are, the first request can be rejected.

[0153] (2) Cache check:

[0154] The cache determination process may include: querying the input content in the content inputter's personal cache table. This personal cache table can be used to cache content that has already been queried by the content inputter, as well as the identifier data of the query results corresponding to the queried content. The purpose of this query is to check whether the input content has already been queried by the content inputter. If it has, the identifier data of the first query result corresponding to the input content can be retrieved from the personal cache table. Based on this identifier data, the first query result corresponding to the input content can be retrieved from the result cache space.

[0155] If the input content has not been queried by the content inputter, it can be queried in the result cache space. The result cache space can be used to cache the content that has been queried, as well as the query results corresponding to the queried content. The purpose of the query here is to check whether the input content has been queried by other inputters. If it has been queried by other inputters, the first query result corresponding to the input content can be read from the result cache space and returned to the content inputter. In addition, the identifier data of the first query result corresponding to the input content can also be cached in the content inputter's personal cache table. This can ensure that when the content inputter queries the same input content in the future, the same query result will be obtained. If it has not been queried by other inputters, (3) can be executed.

[0156] It should be noted that the query results need to undergo another anti-spam check. Query results that pass the anti-spam check (i.e., the query results do not contain illegal words) can be returned to the content input user. This ensures that the query results output to the content input user are legitimate content.

[0157] (3) Perform recall processing on the input content to obtain the recalled content:

[0158] For input content (queries) without cached results, we need to use a recall service (emb service) to retrieve content related to the input content (query) as the recall content. Recall content can come from sources such as books or core viewpoints.

[0159] (4) Relevance ranking:

[0160] The input query and the recalled content are sorted using a relevance model. Low-relevance recalled content is filtered out based on a certain relevance threshold, and the recalled content is then sorted from highest to lowest relevance score. Before the large model generates the answer, a flag is generated, the standard generation process begins, and the flag is incrementally written to storage. At this point, the first request generates a response packet, and the subsequent processes proceed asynchronously.

[0161] (5) Large model generates response content:

[0162] Specifically, if the number of recalled content after relevance filtering is less than or equal to the second threshold (the first threshold can be set based on experience, for example, 1), the query can be rejected. If the number of recalled content is greater than the second threshold but less than or equal to the first threshold (the first threshold can be set based on experience, for example, 5), then the large model can directly answer the input content (query) without relying on the recalled content, thus obtaining the answer. The large model here uses a specific fine-tuned model, referred to here as model1 (the first large model).

[0163] If the number of recalled content after relevance filtering is sufficient (e.g., the number of recalled content is greater than a first threshold), the retained recalled content (i.e., the recalled content after relevance filtering) can be input into the large model, allowing the large model to answer based on the given recalled content and obtain the answer content. The large model here uses a specific fine-tuned model, which is referred to here as model2 (the second large model).

[0164] If the input (query) is directly related to a book, the book's summary can be used as input to the larger model, allowing it to respond based on the summary. The larger model here is model2 (the second largest model) mentioned above.

[0165] The generated answers can be incrementally written to storage.

[0166] (6) Generate relevant query content from the large model:

[0167] In detail, based on the answers generated by the large model, sentence by sentence plus the input content (query), the fine-tuned large model model3 (the third large model) can generate query content that is semantically related to the input content.

[0168] If the input query is related to a book, or if the answer is generated based on a book's summary, we directly use pre-generated offline book-related questions as the query content that is semantically relevant to the input.

[0169] The generated query results can be incrementally written to storage.

[0170] (7) Generate references:

[0171] In detail, if there is recalled content, a similarity model can be used to associate the similarity between the generated answer content and the recalled content. For similarity reaching a certain threshold, a reference relationship can be established between the answer content and the recalled content.

[0172] If no recall content exists, books can be recalled based on the generated answer content. Then, a similarity model can be used to associate the answer content with the books based on similarity. For those with similarity reaching a certain threshold, a citation relationship between the answer content and the books can be established.

[0173] The generated reference relationships can be incrementally written to storage.

[0174] (8) Update cache:

[0175] The generated data will be written to the cache and can be reused for subsequent requests.

[0176] II. For non-first-time requests:

[0177] Based on the unique session identifier returned in the first request, the corresponding storage is read directly, and the content in the storage is returned to the content user after passing the anti-spam test. If the content triggers the anti-spam test, the entire process ends immediately; otherwise, the process ends when the completion flag is read.

[0178] In this embodiment of the application, by combining the generation capability of the AI ​​big model with the content input process, the AI ​​big model can be used to quickly generate query results corresponding to the input content during the content generation process, thereby improving the query efficiency during the content input process.

[0179] Please refer to Figure 21, which is a schematic diagram of a content processing apparatus provided in an embodiment of this application. This content processing apparatus can be installed in a computer device provided in this embodiment. The computer device can be a terminal of a content input user, or a content processing system composed of a content input user's terminal and a backend server. The apparatus shown in Figure 21 can be a computer program running in the computer device, and can be used to execute some or all of the steps in the method embodiments shown in Figures 2, 13, or 19. Please refer to Figure 21; the content processing apparatus may include the following units:

[0180] Display unit 2101 is used to display the service interface of the first client;

[0181] The display unit 2101 is also used to receive content input operations in the service interface;

[0182] The display unit 2101 is also configured to display an input method panel in response to a query symbol entered in the content input operation; and to display a first query result in the input method panel corresponding to the content entered in the content input operation.

[0183] In one implementation, when display unit 2101 displays the input method panel in response to a query symbol entered during a content input operation, it specifically performs the following steps:

[0184] The input method panel is displayed in response to the query symbol entered after the target content is entered during the content input operation.

[0185] In one implementation, the query symbol is located at any position in the target content; the display unit 2101, when displaying the first query result corresponding to the content entered in the content input operation in the input method panel, specifically performs the following steps:

[0186] The input method panel displays the first search result corresponding to the target content.

[0187] In one implementation, when the display unit 2101 displays the first query result corresponding to the content entered in the content input operation in the input method panel, it specifically performs the following steps:

[0188] When the query symbol is located within the target content, the first query result corresponding to the first content is displayed in the input method panel. The first content is the content within the target content that is located before the query symbol.

[0189] When the query symbol is at the beginning or end of the target content, the first query result corresponding to the target content is displayed in the input method panel.

[0190] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0191] In response to the query symbol entered during the content input operation, if the business type of the first client is the first business type, then the input method panel will be displayed; the first business type refers to the business type that rarely inputs query symbols.

[0192] If the first client's service type is the second service type, the input method panel will not be displayed; the second service type refers to the service type that frequently inputs query symbols.

[0193] In one implementation, the first query result includes at least one of the following:

[0194] The response content corresponding to the content entered during the content input operation;

[0195] The cited information in the answer content; the cited information in the answer content refers to the descriptive information of the cited content in the answer content;

[0196] The query content is semantically related to the content entered in the content input operation.

[0197] In one implementation, the input method to which the input method panel belongs is set with multiple preset intents; the multiple preset intents include explicit intents and general intents, where explicit intents refer to intents used to obtain specific types of query results, and general intents refer to intents used to obtain general query results;

[0198] When the query intent of the content entered in the content input operation is explicit, the first query result includes the answer content corresponding to the content entered in the content input operation;

[0199] When the query intent of the content entered in the content input operation is a general intent, the first query result includes the answer content corresponding to the content entered in the content input operation, the reference information of the answer content, and the query content that is semantically related to the content entered in the content input operation.

[0200] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0201] In response to a trigger action on the referenced information, display the original text of the referenced content.

[0202] In one implementation, the source of the referenced content is a second client; the display unit 2101, in response to a trigger operation on the referenced information, displays the original text of the referenced content, specifically performing the following steps:

[0203] In response to a triggering operation on the referenced information, the original text of the referenced content is displayed in the second client.

[0204] In one implementation, the answer content includes one or more sub-contents. Each target sub-content in the answer content that contains referenced content is marked with a reference tag. The reference information includes the reference information of each target sub-content. The reference information of each target sub-content refers to the introductory information of the referenced content referenced by the corresponding target sub-content. The reference tag of the reference information of each target sub-content corresponds to the reference tag that has been added to it.

[0205] Display unit 2101, in response to a trigger operation on reference information, displays the original text of the referenced content, specifically performing the following steps:

[0206] In response to a triggering operation on the reference information of any target sub-content, display the original text of the referenced content referenced by any target sub-content.

[0207] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0208] In response to a query operation, the second query result corresponding to the query content is displayed in the input method panel.

[0209] In one implementation, the input method to which the input method panel belongs has multiple preset intents; the display unit 2101, when displaying the first query result corresponding to the content entered in the content input operation in the input method panel, specifically performs the following steps:

[0210] The input method panel displays the search results corresponding to the content entered during the content input operation under the first preset intent;

[0211] The first preset intent is the preset intent that has the highest relevance to the query intent of the content entered in the content input operation among multiple preset intents.

[0212] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0213] Share the first query result.

[0214] In one implementation, the display unit 2101, when sharing the first query result, performs any of the following:

[0215] Share the first query result in text format;

[0216] Share the first query result as an image;

[0217] Share the first query result in voice format;

[0218] Share the first query result as a link.

[0219] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0220] Select the content to share from the first search results;

[0221] The display unit 2101 is used to perform the following steps when sharing the first query result:

[0222] Share the selected results.

[0223] In one implementation, the display unit 2101 is further configured to perform the following steps:

[0224] The input method panel is displayed in response to a triggering action on the query control; or, the input method panel is displayed in response to a query gesture.

[0225] In one implementation, the content processing apparatus further includes a processing unit 2102 for performing the following steps:

[0226] The intent of the input content is identified, and one or more matching intents of the input content are determined from multiple preset intents; each preset intent corresponds to its own query model.

[0227] Call the query model corresponding to each matching intent, query the results of the content input operation, and obtain the query results of the content input operation under each matching intent;

[0228] Sort the query results for the content entered in the content input operation under each matching intent, and determine the first query result corresponding to the content entered in the content input operation.

[0229] In one implementation, any one of one or more matching intents is represented as the target matching intent; the processing unit 2102 is used to call the query model corresponding to the target matching intent, perform result query on the content input operation, and obtain the query result of the content input operation under the target matching intent, specifically performing the following steps:

[0230] Invoke the query model corresponding to the target matching intent to generate the answer content corresponding to the content input operation;

[0231] Invoke the query model corresponding to the target matching intent to retrieve the referenced content of the answer;

[0232] Invoke the query model corresponding to the target matching intent to generate query content that is semantically related to the content input operation;

[0233] The query results for the content input in the content input operation are determined as the query results for the content input in the content input operation under the target matching intent, including the answer content corresponding to the content input operation, the reference information of the reference content of the answer content, and the query content that is semantically related to the content input operation.

[0234] In one implementation, when processing unit 2102 calls the query model corresponding to the target matching intent to generate the answer content corresponding to the content input in the content input operation, it specifically performs the following steps:

[0235] The input content is retrieved by performing a recall process to obtain the retrieved content.

[0236] If the recalled content contains books related to the content entered during the content input operation, the query model corresponding to the target matching intent will be invoked to generate the answer content based on the summary content of the books;

[0237] If there are no books related to the content entered in the content input operation in the recalled content, and the number of recalled content is less than or equal to the first quantity threshold, then the query model corresponding to the target matching intent is invoked to generate the answer content based on the content entered in the content input operation.

[0238] If the number of recalled content items exceeds the first threshold, the query model corresponding to the target matching intent is invoked to generate answer content based on the recalled content.

[0239] In one implementation, when processing unit 2102 calls the query model corresponding to the target matching intent to retrieve the referenced content of the answer, it specifically performs the following steps:

[0240] If the answer content is generated based on the recall content corresponding to the content input by the content input operation, then the recall content with a similarity greater than the first similarity threshold with the answer content will be identified as the reference content of the answer content.

[0241] If the answer content is not generated based on the recall content corresponding to the content input operation, the query model corresponding to the target matching intent is invoked to recall books related to the answer content based on the answer content, and the books among the recalled books whose similarity to the answer content is greater than the second similarity threshold are identified as the reference content of the answer content.

[0242] In one implementation, when processing unit 2102 calls the query model corresponding to the target matching intent to generate query content semantically related to the content input in the content input operation, it specifically performs the following steps:

[0243] If the answer is generated based on the recall content corresponding to the content input in the content input operation, then the query model corresponding to the target matching intent is invoked to generate query content that is semantically related to the content input in the content input operation based on the recall content and the content input in the content input operation.

[0244] If the recalled content contains books related to the content entered in the content input operation, then the query content associated with the books will be used as the query content semantically related to the content entered in the content input operation.

[0245] According to one embodiment of this application, the various units in the content processing apparatus shown in FIG22 can be individually or entirely merged into one or more other units, or some of the units can be further divided into multiple functionally smaller units. This can achieve the same operation without affecting the technical effect of the embodiment of this application. The above units are based on logical function division. In practical applications, the function of one unit can also be implemented by multiple units, or the function of multiple units can be implemented by one unit. In other embodiments of this application, the content processing apparatus may also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented collaboratively by multiple units.

[0246] According to another embodiment of this application, the content processing apparatus shown in FIG21, and the content processing method of the embodiments of this application, can be constructed and implemented by running a computer program capable of performing some or all of the steps involved in the methods shown in FIG2, 13, or 19 on a general-purpose computing device, such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access storage medium (RAM), and read-only storage medium (ROM). The computer program can be recorded on, for example, a computer-readable storage medium, loaded into the aforementioned computing device through the computer-readable storage medium, and run therein.

[0247] In this embodiment, a content input operation can be received in the service interface. In response to the query symbol entered in the content input operation, an input method panel can be displayed, and a first query result corresponding to the content entered in the content input operation can be displayed in the input method panel. That is to say, in this embodiment, the query result corresponding to the input content can be obtained by entering the query symbol. The input of the query symbol is simple and fast. Therefore, in this embodiment, the query result corresponding to the input content can be quickly obtained during the content input process by entering the query symbol.

[0248] Please refer to Figure 22, which is a schematic diagram of a computer device provided in an embodiment of this application. The computer device shown in Figure 22 includes at least a processor 2201, an input interface 2202, an output interface 2203, and a computer-readable storage medium 2204. The processor 2201, input interface 2202, output interface 2203, and computer-readable storage medium 2204 can be connected via a bus or other means.

[0249] Computer-readable storage medium 2204 can be stored in the memory of a computer device. Computer-readable storage medium 2204 is used to store computer programs, including computer instructions. Processor 2201 is used to execute the computer program stored in computer-readable storage medium 2204. Processor 2201 (or CPU (Central Processing Unit)) is the computing and control core of the computer device, suitable for implementing computer programs, specifically suitable for loading and executing computer programs to achieve corresponding methods or functions.

[0250] This application also provides a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space for storing the operating system of the computer device. Furthermore, the storage space also stores computer programs suitable for loading and execution by a processor. It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device; optionally, it can also be at least one computer-readable storage medium located remotely from the aforementioned processor.

[0251] The computer device can be a terminal for the content input user, or it can be a content processing system consisting of a terminal for the content input user and a backend server. In a specific implementation, the processor 2201 can load and execute the computer program stored in the computer-readable storage medium 2204 to implement the corresponding steps in the methods shown in Figures 2, 13, or 19.

[0252] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the content processing method described above.

Claims

1. A content processing method, characterized in that, The method is performed by a computer device, and the method includes: Display the service interface of the first client; Receive content input operations in the service interface; In response to the query symbol entered during the content input operation, the input method panel is displayed; and, The input method panel displays the first query result corresponding to the content entered during the content input operation.

2. The method as described in claim 1, characterized in that, The method of displaying an input method panel in response to a query symbol entered in response to the content input operation includes: In response to the query symbol entered after the target content is entered during the content input operation, the input method panel is displayed.

3. The method as described in claim 1 or 2, characterized in that, The query symbol is located at any position in the target content; displaying the first query result corresponding to the content entered in the content input operation in the input method panel includes: The first query result corresponding to the target content is displayed in the input method panel.

4. The method according to any one of claims 1-3, characterized in that, The display of the first query result corresponding to the content entered in the content input operation on the input method panel includes: When the query symbol is located in the target content, a first query result corresponding to the first content is displayed in the input method panel. The first content is the content in the target content that is located before the query symbol. When the query symbol is located at the beginning or end of the target content, the first query result corresponding to the target content is displayed in the input method panel.

5. The method according to any one of claims 1-4, characterized in that, The method further includes: In response to the query symbol entered during the content input operation, if the first client's business type is a first business type, then the display of the input method panel is triggered; the first business type refers to the business type that rarely inputs the query symbol. If the first client's service type is the second service type, then the input method panel will not be displayed; the second service type refers to the service type that frequently inputs the query symbol.

6. The method according to any one of claims 1-5, characterized in that, The first query result includes at least one of the following: The answer content corresponding to the content entered in the aforementioned content input operation; The cited information in the answer content; the cited information in the answer content refers to the descriptive information of the cited content referenced by the answer content; The query content is semantically related to the content entered in the content input operation.

7. The method according to any one of claims 1-6, characterized in that, The input method panel has multiple preset intents set for its input method; these preset intents include explicit intents and general intents. Explicit intents refer to intents used to obtain specific types of query results, while general intents refer to intents used to obtain general query results. When the query intent of the content entered in the content input operation is the explicit intent, the first query result includes the answer content corresponding to the content entered in the content input operation; When the query intent of the content input operation is the general intent, the first query result includes the answer content corresponding to the content input operation, the reference information of the answer content, and the query content that is semantically related to the content input operation.

8. The method according to any one of claims 1-7, characterized in that, The method further includes: In response to a trigger operation on the referenced information, the original text of the referenced content is displayed.

9. The method according to any one of claims 1-8, characterized in that, The source of the cited content is a second client; the step of displaying the original text of the cited content in response to a trigger operation on the cited information includes: In response to a trigger operation on the referenced information, the original text of the referenced content is displayed in the second client.

10. The method according to any one of claims 1-9, characterized in that, The answer content includes one or more sub-contents. Each target sub-content in the answer content that contains cited content is marked with a citation tag. The citation information includes the citation information of each target sub-content. The citation information of each target sub-content refers to the introductory information of the cited content referenced by the corresponding target sub-content. The citation tag of the citation information of each target sub-content corresponds to the citation tag that has been added to it. The step of displaying the original text of the referenced content in response to a triggering operation on the referenced information includes: In response to a trigger operation on the reference information of any target sub-content, the original text of the referenced content referenced by the target sub-content is displayed.

11. The method according to any one of claims 1-10, characterized in that, The method further includes: In response to the query operation on the query content, a second query result corresponding to the query content is displayed in the input method panel.

12. The method according to any one of claims 1-11, characterized in that, The input method panel has multiple preset intents for its input method settings; The display of the first query result corresponding to the content entered in the content input operation on the input method panel includes: The input method panel displays the query results corresponding to the content entered by the content input operation under the first preset intent; The first preset intent is the preset intent that has the highest correlation with the query intent of the content entered in the content input operation among the plurality of preset intents.

13. The method according to any one of claims 1-12, characterized in that, The method further includes: Share the results of the first query.

14. The method according to any one of claims 1-13, characterized in that, Sharing the first query result includes any of the following: Share the first query result in text format; Share the first query result as an image; Share the first query result in voice format; Share the first query result as a link.

15. The method according to any one of claims 1-14, characterized in that, The method further includes: Select the content to share from the first query result; Sharing the first query result includes: Share the selected results.

16. The method according to any one of claims 1-15, characterized in that, The method further includes: In response to a trigger action on the query control, display the input method panel; or, The input method panel is displayed in response to the input query gesture.

17. The method according to any one of claims 1-16, characterized in that, The method further includes: The intent of the content input operation is identified, and one or more matching intents of the content input operation are determined from multiple preset intents; each preset intent corresponds to its own query model. Call the query model corresponding to each matching intent, perform a result query on the content input operation, and obtain the query results of the content input operation under each matching intent; The query results of the content input in the content input operation under each matching intent are sorted to determine the first query result corresponding to the content input in the content input operation.

18. The method according to any one of claims 1-17, characterized in that, Any one of the one or more matching intents is referred to as the target matching intent; The process of calling the query model corresponding to the target matching intent, querying the content input by the content input operation, and obtaining the query results of the content input under the target matching intent includes: The query model corresponding to the target matching intent is invoked to generate the answer content corresponding to the content input in the content input operation; Invoke the query model corresponding to the target matching intent to retrieve the referenced content of the answer content; Invoke the query model corresponding to the target matching intent to generate query content that is semantically related to the content input by the content input operation; The answer content corresponding to the content input in the content input operation, the reference information of the reference content of the answer content, and the query content that is semantically related to the content input in the content input operation are determined as the query results of the content input in the content input operation under the target matching intent.

19. The method according to any one of claims 1-18, characterized in that, The step of calling the query model corresponding to the target matching intent to generate answer content corresponding to the content input in the content input operation includes: The content input by the aforementioned content input operation is retrieved to obtain the retrieved content; If the recalled content contains books related to the content input during the content input operation, then the query model corresponding to the target matching intent is invoked to generate the answer content based on the summary content of the books; If there are no books related to the content input in the content input operation among the recalled content, and the number of recalled content is less than or equal to the first quantity threshold, then the query model corresponding to the target matching intent is invoked to generate the answer content based on the content input operation. If the number of recalled content items is greater than a first quantity threshold, then the query model corresponding to the target matching intent is invoked to generate the answer content based on the recalled content.

20. The method according to any one of claims 1-19, characterized in that, The step of invoking the query model corresponding to the target matching intent and recalling the referenced content of the answer content includes: If the answer content is generated based on the recall content corresponding to the content input by the content input operation, then the recall content with a similarity greater than the first similarity threshold with the answer content is determined as the reference content of the answer content. If the answer content is not generated based on the recall content corresponding to the content input operation, then the query model corresponding to the target matching intent is invoked to recall books related to the answer content based on the answer content, and the books among the recalled books whose similarity to the answer content is greater than the second similarity threshold are identified as the reference content of the answer content.

21. The method according to any one of claims 1-20, characterized in that, The step of invoking the query model corresponding to the target matching intent to generate query content semantically related to the content input in the content input operation includes: If the answer content is generated based on the recall content corresponding to the content input in the content input operation, then the query model corresponding to the target matching intent is invoked to generate query content that is semantically related to the content input in the content input operation based on the recall content and the content input in the content input operation; If the recalled content contains books related to the content entered in the content input operation, then the query content associated with those books will be used as the query content semantically related to the content entered in the content input operation.

22. A content processing apparatus, characterized in that, include: The display unit is used to display the service interface of the first client. The display unit is also used to receive content input operations in the service interface; The display unit is further configured to display an input method panel in response to a query symbol entered in response to the content input operation; and, The display unit is also used to display the first query result corresponding to the content entered in the content input operation on the input method panel.

23. A computer device, characterized in that, The computer device includes: A processor is a tool for implementing computer programs. A computer-readable storage medium storing a computer program adapted to be loaded by the processor and executed as described in any one of claims 1-21.

24. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted to be loaded by a processor and executed by the content processing method as described in any one of claims 1-21.

25. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the content processing method as described in any one of claims 1-21.

Images