A problem recommendation method, device, computer device and storage medium

Abstract

The present disclosure provides a question recommendation method, apparatus, computer device, and storage medium. Among them, the method includes: in response to meeting the display condition for displaying recommended question information on a target dialogue page, determining the target scenario type corresponding to the target dialogue page; based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type, determining the target model auxiliary information for generating the recommended question information; the second model auxiliary information corresponding to different scenario types is different; inputting the target model auxiliary information into an artificial intelligence model to generate the recommended question information.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology. Specifically, it relates to a question recommendation method, device, computer device, and storage medium. Background Art

[0002] When using artificial intelligence technology to conduct dialogue and answer questions on a dialogue page, the user can ask questions, and then the artificial intelligence technology can provide answer content in response to the questions. For the user, it is relatively difficult to continuously input question information. After receiving the feedback answer content, the user cannot easily initiate a new question, resulting in the user being unable to continue the dialogue to further obtain information. Summary of the Invention

[0003] Embodiments of the present disclosure at least provide a question recommendation method, device, computer device, and storage medium.

[0004] In a first aspect, embodiments of the present disclosure provide a question recommendation method, including: in response to meeting the display condition for displaying recommended question information on a target dialogue page, determining the target scenario type corresponding to the target dialogue page; based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type, determining the target model auxiliary information for generating the recommended question information; the second model auxiliary information corresponding to different scenario types is different; inputting the target model auxiliary information into an artificial intelligence model to generate the recommended question information.

[0005] In an optional implementation, the second model auxiliary information includes multiple example question information under the target scenario type; and / or, includes question description information for the target scenario type, and the question description information includes content limitation information and / or format limitation information for the recommended question information.

[0006] In an optional implementation, the question description information is determined according to the following steps: obtaining the historical dialogue content in the target dialogue page; based on the historical dialogue content, determining at least one target keyword associated with the target scenario type; based on the target keyword and the model auxiliary configuration information corresponding to the target scenario type, determining the question description information.

[0007] In an optional implementation, before inputting the target model auxiliary information into the artificial intelligence model, the method further includes: obtaining the historical dialogue content in the target dialogue page; the inputting the target model auxiliary information into the artificial intelligence model to generate the recommended question information includes: inputting the target model auxiliary information and the historical dialogue content into the artificial intelligence model to generate the recommended question information.

[0008] In an alternative embodiment, inputting the target model auxiliary information and the historical conversation content into an artificial intelligence model to generate the recommended question information includes: screening out the most recent first historical conversation content from the obtained multiple historical conversation contents, and second historical conversation contents that have a contextual association with the most recent first historical conversation content; and inputting the target model auxiliary information, the first historical conversation content, and the second historical conversation content into the artificial intelligence model to generate the recommended question information.

[0009] In an alternative embodiment, the display conditions for displaying the recommended question information on the target conversation page include at least one of the following: determining that the target conversation page starts to be displayed currently; determining that a reply message is received from the artificial intelligence module in response to the question information, or no new question information is received within a preset time period after receiving the reply message.

[0010] In an alternative embodiment, the method further includes: determining the question feature information of the user according to the trigger data of the user for the recommended question information; and updating the target model auxiliary information based on the question feature information, so as to assist the artificial intelligence model in generating new recommended question information based on the updated target model auxiliary information.

[0011] In a second aspect, an embodiment of the present disclosure further provides a question recommendation device, including: a first determination module, configured to determine the target scenario type corresponding to the target conversation page in response to meeting the display conditions for displaying the recommended question information on the target conversation page; a second determination module, configured to determine the target model auxiliary information for generating the recommended question information based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type; the second model auxiliary information corresponding to different scenario types is different; and a generation module, configured to input the target model auxiliary information into an artificial intelligence model to generate the recommended question information.

[0012] In an alternative embodiment, the second model auxiliary information includes multiple question information examples in the target scenario type; and / or includes question description information for the target scenario type, and the question description information includes content limitation information and / or format limitation information for the recommended question information.

[0013] In an alternative embodiment, the question description information is determined according to the following steps: obtaining the historical conversation content in the target conversation page; determining at least one target keyword associated with the target scenario type based on the historical conversation content; and determining the question description information based on the target keyword and the model auxiliary configuration information corresponding to the target scenario type.

[0014] In an alternative embodiment, before inputting the target model auxiliary information into the artificial intelligence model, the generation module is further configured to: obtain the historical conversation content in the target conversation page; when the generation module inputs the target model auxiliary information into the artificial intelligence model to generate the recommended question information, it is configured to: input the target model auxiliary information and the historical conversation content into the artificial intelligence model to generate the recommended question information.

[0015] In an alternative embodiment, when the generation module inputs the target model auxiliary information and the historical conversation content into the artificial intelligence model to generate the recommended question information, it is configured to: screen out the most recent first historical conversation content from the obtained multiple historical conversation contents, and a second historical conversation content that has a contextual association with the most recent first historical conversation content; input the target model auxiliary information, the first historical conversation content, and the second historical conversation content into the artificial intelligence model to generate the recommended question information.

[0016] In an alternative embodiment, the display conditions for displaying the recommended question information on the target conversation page include at least one of the following: determining that the display of the target conversation page starts currently; determining that a reply message to the question information is received from the artificial intelligence module or no new question information is received within a preset time period after receiving the reply message.

[0017] In an alternative embodiment, the device further includes a processing module, configured to: determine the question feature information of the user according to the trigger data of the user for the recommended question information; based on the question feature information, update the target model auxiliary information to assist the artificial intelligence model in generating new recommended question information based on the updated target model auxiliary information.

[0018] In a third aspect, an alternative implementation manner of the present disclosure further provides a computer device, including a processor and a memory. The memory stores machine-readable instructions executable by the processor. The processor is configured to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the machine-readable instructions execute the steps in the first aspect or any possible implementation manner in the first aspect.

[0019] In a fourth aspect, an alternative implementation manner of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run, it executes the steps in the first aspect or any possible implementation manner in the first aspect.

[0020] A problem recommendation method, apparatus, computer device, and storage medium provided by an embodiment of the present disclosure are specifically selected to automatically generate recommended question information for a user to select and reference, thereby solving the problem that it is difficult for the user to initiate a new question. When generating the recommended question information, specifically, the target model auxiliary information for generating the recommended question information can be determined by adapting the first model auxiliary information of each scenario type and the second model auxiliary information matching the current target scenario type, so that the artificial intelligence model uses the target model auxiliary information to generate the recommended question information. In this way, the generated recommended question information can not only meet the requirements applicable in each scenario, but also specifically reflect the characteristics of the scenario type reflected in the current dialogue page.

[0021] To make the above objects, features, and advantages of the present disclosure more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the following briefly introduces the accompanying drawings required for use in the embodiments. The accompanying drawings are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present disclosure and, together with the specification, are used to explain the technical solutions of the present disclosure. It should be understood that the following drawings only illustrate certain embodiments of the present disclosure and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 Shows a flowchart of a problem recommendation method provided by an embodiment of the present disclosure;

[0024] Figure 2a Shows a schematic diagram of a target dialogue page provided by an embodiment of the present disclosure;

[0025] Figure 2b Shows a schematic diagram of another target dialogue page provided by an embodiment of the present disclosure;

[0026] Figure 3 Shows a schematic diagram of a problem recommendation apparatus provided by an embodiment of the present disclosure;

[0027] Figure 4 Shows a schematic diagram of a computer device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Apparently, the described embodiments are only a part rather than all of the embodiments of the present disclosure. The components of the embodiments of the present disclosure described and illustrated herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure is not intended to limit the scope of the present disclosure claimed, but merely represents selected embodiments of the present disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of the present disclosure without creative efforts fall within the scope of protection of the present disclosure.

[0029] It has been found through research that when using artificial intelligence technology to conduct dialogue and answer questions on a dialogue page, the user can ask questions, and then the artificial intelligence technology can provide answer content in response to the questions raised. Therefore, if one wants to obtain the answer content in response, the user needs to provide information first. However, it is relatively difficult for the user to continuously initiate new questions, which may cause the user to be unable to continue the dialogue to further obtain information.

[0030] Based on the above research, the present disclosure provides a question recommendation method, specifically selecting to automatically generate recommended question information for the user for the user to select and refer to, so as to solve the problem that it is difficult for the user to initiate new questions. When generating the recommended question information, specifically, the target model auxiliary information for generating the recommended question information can be determined by adapting the first model auxiliary information of each scenario type and the second model auxiliary information matching the current target scenario type, so that the artificial intelligence model can use the target model auxiliary information to generate the recommended question information. In this way, the generated recommended question information can not only meet the requirements applicable in each scenario, but also specifically reflect the characteristics of the scenario type reflected in the current dialogue page.

[0031] Regarding the defects existing in the above solutions, they are all the results obtained by the inventors through practice and careful research. Therefore, the process of discovering the above problems and the solutions proposed by the present disclosure below for the above problems should be the contributions made by the inventors to the present disclosure during the process of the present disclosure.

[0032] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0033] To facilitate the understanding of this embodiment, a problem recommendation method disclosed in the embodiments of the present disclosure will be introduced in detail first. The execution subject of the problem recommendation method provided in the embodiments of the present disclosure is generally a computer device with certain computing capabilities, and this computer device includes, for example: a terminal device, a server, or other processing devices. The terminal device can be a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some possible implementation manners, the problem recommendation method can be implemented by a processor calling computer-readable instructions stored in a memory.

[0034] The problem recommendation method provided in the embodiments of the present disclosure will be described below. The problem recommendation method provided in the embodiments of the present disclosure can be applied to a platform that relies on an artificial intelligence model to complete dialogue and question answering. In this platform, the content input by the user and the answer content replied by the artificial intelligence model according to the content input by the user are specifically presented in the form of a dialogue page.

[0035] In the existing dialogue page, the user can successively input information such as questions. After the artificial intelligence model makes a response according to the information input by the user, a dialogue is completed. If a new dialogue is to be started, the user needs to input information again. However, for the user, this continuous questioning is difficult, and the user may not be able to provide further questions to continue exploring the topic. For example, in the previous dialogue, the user asked a question about a physical concept. After the artificial intelligence model explains this physical concept, the user cannot easily determine the content that can be further explored. In the embodiments of the present disclosure, the recommended question information can be specifically provided to the user in a targeted manner based on the scene matched in the dialogue to assist the user in further dialogue, thereby reducing the usage difficulty of the user on this platform.

[0036] See Figure 1 As shown, it is a flowchart of a problem recommendation method provided in the embodiments of the present disclosure. The method includes steps S101 to S103, where:

[0037] S101: In response to satisfying the display condition for displaying recommended question information on the target dialogue page, determine the target scene type corresponding to the target dialogue page;

[0038] S102: Based on the first model auxiliary information adapted to each scene type and the second model auxiliary information adapted to the target scene type, determine the target model auxiliary information for generating the recommended question information; the second model auxiliary information corresponding to different scene types is different;

[0039] S103: Input the target model auxiliary information into the artificial intelligence model to generate the recommended question information.

[0040] Regarding S101 above, first, the target dialogue page is described. Under the above-mentioned platform, a dialogue page is specifically provided to present the content input by the user and the response content replied by the artificial intelligence model according to the content input by the user in the form of a dialogue page. In a possible case, the dialogue page may specifically correspond to a specific vertical category or topic. For example, under a dialogue page, it is specifically associated with only one vertical category. For example, in a certain dialogue page, a dialogue is specifically carried out for vertical categories such as sports, medicine, or performing arts. Or, it can also be reflected that multiple recent dialogue contents in the dialogue page specifically revolve around the same topic, such as a dialogue around a certain physical concept, which is not limited here. In the embodiments of the present disclosure, the determined categories such as vertical categories and topics that can indicate the same type are referred to as target scenario types.

[0041] For the dialogue page selected by the user for the dialogue, that is, the target dialogue page. In a possible case, if it is determined that the condition for displaying the recommended question information is met under the target dialogue page, the target scenario type corresponding to the target dialogue page can be first determined, and then the recommended question information can be generated according to the determined target scenario type.

[0042] Specifically, the display conditions for displaying the recommended question information on the target dialogue page specifically include at least one of the following: determining that the display of the target dialogue page starts currently; determining that the response information feedback by the artificial intelligence module for the question information is received or no new question information is received within a preset duration after receiving the response information.

[0043] Next, the above-listed several possible display conditions are described.

[0044] In a possible case, for new users on the platform or users who open a new dialogue page for the first time, they may not clearly know how to send information to make the artificial intelligence model reply. Therefore, in this case, it is appropriate to select the method of first providing the recommended question information to the user as a teaching example so that the user can refer to these recommended question information to determine the information to be sent.

[0045] Exemplarily, see Figure 2aAs shown, it is a schematic diagram of a target dialogue page provided by an embodiment of the present disclosure. In this schematic diagram, the specific target scenario type corresponding to the target dialogue page, for example, indicates physics. It is currently determined to start displaying the target dialogue page. Therefore, the information sent by the user or the information fed back by the artificial intelligence model is not included in this target dialogue page. However, multiple recommended question information can be specifically displayed below the target dialogue page, and it can be guided by "You can ask:", and multiple questions as shown in the example are displayed.

[0046] In another possible case, after the user sends information in the target dialogue page and receives the reply information fed back by the artificial intelligence module, there may also be a need to obtain the recommended question information. Therefore, here it can be determined to display the recommended question information after receiving the reply information fed back by the artificial intelligence module for the question information, so that the user can use the recommended question information to assist in continuing the conversation.

[0047] Or, considering that the user may be able to organize the language by himself to determine the next question information after viewing the reply information fed back by the artificial intelligence model, and thus does not need to provide the recommended question information, it can also be determined that the display condition includes that no new question information is received within a preset time after receiving the reply information. For example, after the reply information fed back by the artificial intelligence module is completely displayed for one minute and no new question information is received, it can be considered that the user cannot easily think about how to ask the next question at present, so it is also appropriate to display the recommended question information.

[0048] Exemplarily, refer to Figure 2b As shown, it is a schematic diagram of another target dialogue page provided by an embodiment of the present disclosure. In this schematic diagram, the specific target scenario type corresponding to the target dialogue page, for example, also indicates physics, and there is already conversation content in the dialogue page, including the question information proposed by the user and the reply information fed back by the artificial intelligence model. Below the page, multiple recommended question information is displayed, specifically arranged and displayed in the form of multiple questions.

[0049] For the target dialogue page, in the case of determining the display condition of the recommended question information according to the method in the above example, the target scenario type corresponding to the target dialogue page can be determined. Here, as described in the above example, if the target dialogue page is created only for a certain scenario type, such as specifically created for discussing physics, the scenario type can be directly determined to indicate physics. Or, if the target dialogue page is a dialogue page that can comprehensively discuss various topics, the current target scenario type under the target dialogue page can be specifically determined according to the recent several pieces of conversation content, such as sports, learning, etc.

[0050] In this way, according to the above steps, it is possible to determine whether to display the recommended question information on the target conversation page and the specific target scenario type corresponding to the target conversation page.

[0051] Regarding the above S102, after determining the target scenario type under the target conversation page, recommended question information for display to the user can be generated for the target conversation page. Here, when generating the recommended question information, an artificial intelligence model can be specifically selected. In order for the artificial intelligence model to generate the recommended question information by itself, it is necessary to input the target model auxiliary information (prompt) for generating the recommended question information to it. The target model auxiliary information can specifically be a sentence used to guide and constrain the artificial intelligence model to input appropriate recommended question information.

[0052] In specific implementation, when determining the target model auxiliary information, two aspects of factors can be specifically considered. On the one hand, for the recommended question information, there are the same requirements under different scenario types. For example, in order for the user to understand the text, the generated recommended question information should be of the same language type as the question information sent by the user; or in terms of format, in order for the user to have more question references, specifically three recommended question information are selected for display, etc. In the embodiments of the present disclosure, the model auxiliary information that can adapt to each scenario type is referred to as the first model auxiliary information.

[0053] On the other hand, under different target scenario types, there may specifically be special requirements under that target scenario type. For example, in the target scenario type of painting generation, the recommended question information is specifically, for example, "Generate a seascape painting", which is specifically displayed in the form of a declarative sentence, while in the target scenario type indicating physics as described in the above example, it can specifically be selected to be displayed in the form of an interrogative sentence. Therefore, when determining the target model auxiliary information, in the embodiments of the present disclosure, the second model auxiliary information that adapts to the target scenario type is specifically selected to generate the target model auxiliary information.

[0054] Therefore, in specific implementation, based specifically on the first model auxiliary information that adapts to each scenario type and the second model auxiliary information that adapts to the target scenario type, the target model auxiliary information for generating the recommended question information is determined; the second model auxiliary information corresponding to different scenario types is different.

[0055] The first model auxiliary information and the second model auxiliary information will be described separately below.

[0056] First, for the first model auxiliary information, its specific description targets all target scenario types. Referring to the above example, it may specifically include descriptions of language types and question formats. For example, for all dialogue pages under each scenario type, the language type of the recommended question information determined should be the same as that of the user's question information, and three different questions should be proposed for each.

[0057] Secondly, for the second model auxiliary information, since the second model auxiliary information adapted is different under different target scenario types, it is necessary to determine different second model auxiliary information for different target scenario types respectively. Since there are many possible target scenario types, in one possible case, the second model auxiliary information may specifically include multiple question information examples under the target scenario type, so that the artificial intelligence model can determine the recommended question information to be fed back to the user under this target scenario type through learning from multiple question examples.

[0058] Here, in order for the artificial intelligence model to generate recommended question information under the target scenario type through learning from question information examples, when selecting question information examples, the question information examples will select examples that conform to the question characteristics under the target scenario type. For example, in the above example, in the target scenario type of painting generation, declarative sentences will be specifically used when asking questions, and the content of the painting required will be indicated. Therefore, the question information examples may specifically include, for example, "Generate a seascape painting", "Create an illustration for a romantic novel", "Draw several consecutive life comics", and so on.

[0059] In another possible case, the second model auxiliary information can also be directly determined, that is, the specific description for generating recommended question information is determined for the target scenario type. In this case, the second model auxiliary information includes question description information for the target scenario type, and the question description information includes content limitation information and / or format limitation information for the recommended question information.

[0060] Here, specifically, for the content limitation information, it may specifically include descriptions of the content of the recommended question information. Taking the above-mentioned target scenario of painting generation as an example, the corresponding content limitation information includes, for example, descriptions such as "You should describe a specific scenario" and "The different question information you generate indicates different painting scenarios", which inform the artificial intelligence model on how to generate recommended question information. In addition, there is also format limitation information to limit the format of the output recommended question information, such as "You should output three different recommended questions and display them in the form of 1, 2, 3", and so on. In this way, based on the question description information including content limitation information and / or format limitation information, the artificial intelligence model can also output corresponding question information examples.

[0061] Further, when determining the question description information, the historical conversation content on the target conversation page can also be referred to, so that the question is more coherent and the recommended question information is more valuable for reference. Therefore, in the embodiments of the present disclosure, the following method can specifically be selected to determine the question description information:

[0062] Obtain the historical conversation content on the target conversation page; based on the historical conversation content, determine at least one target keyword associated with the target scenario type; based on the target keyword and the model-assisted configuration information corresponding to the target scenario type, determine the question description information.

[0063] Here, in the target conversation content, multiple target keywords can specifically be determined by means of semantic processing. Here, the target conversation content includes both the question information sent by the user and the received reply information. Since there may have been multiple rounds of conversations on the target conversation page, when selecting the target conversation content, multiple conversation contents generated recently can be selected. For example, the content of the last three conversations can be referred to, or all conversation contents related to the target scenario type can be used as the target conversation content, and no limitation is made here. Taking Figure 2b as an example, in the case where the two displayed conversation contents are used as the target conversation content, the obtained keywords are such as "string theory", "quantum mechanics", "general relativity", "explanation", "universe", "elementary particle", etc., and the keywords associated with the target scenario type can be determined as the target keywords. For example, the keywords "string theory", "quantum mechanics", "general relativity", etc. involving specific physical concepts can be determined as the target keywords.

[0064] It is easy to understand according to the above description that the question description information is specifically a paragraph of descriptive content. Therefore, for the above-determined target keywords, when used to determine the question description information, the model-assisted configuration information corresponding to the target scenario type can specifically be used to determine. For example, the model-assisted configuration information specifically includes that "<target keyword>" must be included when generating the question information. Here, "<target keyword>" indicates the part that must be filled in the model-assisted configuration information, and specifically indicates filling in the target keywords determined in the above steps.

[0065] In this way, for example Figure 2b as shown, in the first recommended question information displayed, the target keywords "string theory", "elementary particle", "fundamental force", and "unified theory" determined from the target conversation content are specifically included. In this way, the historical conversation content can specifically also affect the recommended question information, and the displayed recommended question information also better meets the requirements of the content involved in the continued conversation in the current conversation stage to generate a new conversation.

[0066] In this way, according to the above steps, the target model auxiliary information can be obtained so that the artificial intelligence model can generate recommended question information.

[0067] Regarding the above S103, using the target model auxiliary information, the artificial intelligence model can complete the task of generating recommended question information to display the recommended question information to the user on the target dialogue page.

[0068] In a possible situation, in addition to directly influencing the generation of recommended question information by determining the model auxiliary information in the above steps, it is also possible to provide historical dialogue content as a reference during the process of the artificial intelligence model generating recommended question information using the target model auxiliary information. Here, when providing historical dialogue content, different from specifically focusing on capturing target keywords in the dialogue that match the target scenario type when determining the model auxiliary information above, here it can specifically focus on learning the user's questioning habits in the historical dialogue.

[0069] In specific implementation, before inputting the target model auxiliary information into the artificial intelligence model, it is specifically also possible to obtain the historical dialogue content in the target dialogue page. When inputting the target model auxiliary information into the artificial intelligence model to generate the recommended question information, specifically, the target model auxiliary information and the historical dialogue content can be input into the artificial intelligence model to generate the recommended question information.

[0070] Here, the historical dialogue content can include multiple pieces, specifically determined according to the actual number of conversations generated under the target dialogue page. It specifically includes the question information sent by the user and can also include the reply information feedback by the artificial intelligence model. In a possible situation, the question information can directly reflect the user's questioning habits. In another possible situation, through the user's feedback on the reply information, the user's questioning habits in the Q&A can also be reflected indirectly.

[0071] In specific implementation, when using the historical dialogue content to enable the artificial intelligence model to generate recommended question information, specifically, the most recent piece of historical dialogue content can be used as a reference, and other relevant historical dialogue content can be referred to so that the model can generate recommended question information by referring to this dialogue content. In this way, it is also possible to avoid other dialogue content with low relevance to the current dialogue content on the target dialogue page from interfering with the generation of recommended question information by the artificial intelligence model under the current dialogue topic.

[0072] Therefore, in the embodiments of the present disclosure, specifically, the most recent first historical conversation content can be selected from the obtained multiple historical conversation contents, and the second historical conversation content contextually related to the most recent first historical conversation content can be selected; the target model auxiliary information, the first historical conversation content, and the second historical conversation content are input into the artificial intelligence model to generate the recommended question information.

[0073] For example, in Figure 2b , the second historical conversation content replied by the artificial intelligence model can be used as the most recent first historical conversation content, and the second historical conversation content contextually related to the most recent first historical conversation content can be determined, that is, Figure 2b the first question information sent by the user in Figure 2b . The determined first historical conversation content and second historical conversation content can both be input into the artificial intelligence model to generate the recommended question information. Specifically, through the input historical conversation content, it can be determined that in the user's conversation, there is a preference for asking questions about theoretical proofs. Therefore, recommended question information focusing on proofs can be obtained. For example, in Figure 2b , the first two recommended question information related to the experimental proofs of the theory shown below are obtained.

[0074] In addition, for the recommended question information displayed each time on the target conversation page, according to whether the user selects it and which recommended question information the user specifically selects, etc., the trigger data of the recommended question information can also reflect the user's question preferences.

[0075] In specific implementation, the following method can be adopted: according to the trigger data of the user for the recommended question information, the user's question feature information is determined; based on the question feature information, the target model auxiliary information is updated to assist the artificial intelligence model in generating new recommended question information based on the updated target model auxiliary information.

[0076] For example, in the above target scenario type, according to the user's trigger data, it can be determined, for example, that the user prefers to ask questions about physical concepts and inquire about specific experimental contents, and is not concerned about the historical origin of physical concepts and the specific methods of experimental proofs. That is, through the trigger data of the recommended question information, the user's question feature information can be obtained, so as to help the artificial intelligence model generate recommended question information that more conforms to the user's question preferences according to the specific question feature information of the user.

[0077] Therefore, for the trigger data of the recommended question information, it can specifically indicate whether the recommended question information displayed on the target conversation page is selected, the number of selections, the time elapsed from the display to the trigger, the specific recommended question information selected, as well as the content features corresponding to the selected recommended question information, and the semantic relevance between the selected recommended question information and the most recent conversation content (used to determine whether the user likes to continue the current question or expand to other aspects of the question), and so on. In this way, through the trigger data, the user's question preferences can be learned.

[0078] In addition, since the user's conversation is a dynamic process, for example, there may be changes or differences in question preferences in the early and late stages of the conversation. Therefore, the user's question feature information can be dynamically determined under the continuously updated conversation content, and the target model auxiliary information can be updated using the question feature information, so as to assist the artificial intelligence model in generating new recommended question information according to the dynamically updated target model auxiliary information.

[0079] Those skilled in the art can understand that in the above method of the specific implementation manner, the writing order of each step does not mean a strict execution order and does not constitute any limitation to the implementation process. The specific execution order of each step should be determined according to its function and possible internal logic.

[0080] Based on the same inventive concept, the embodiments of the present disclosure also provide a question recommendation device corresponding to the question recommendation method. Since the principle of solving problems by the device in the embodiments of the present disclosure is similar to the above question recommendation method in the embodiments of the present disclosure, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0081] Refer to Figure 3 As shown, it is a schematic diagram of a question recommendation device provided by an embodiment of the present disclosure. The device includes: a first determination module 31, a second determination module 32, and a generation module 33; wherein,

[0082] The first determination module 31 is configured to determine the target scenario type corresponding to the target conversation page in response to meeting the display condition for displaying the recommended question information on the target conversation page;

[0083] The second determination module 32 is configured to determine the target model auxiliary information for generating the recommended question information based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type; the second model auxiliary information corresponding to different scenario types is different;

[0084] The generation module 33 is configured to input the target model auxiliary information into the artificial intelligence model to generate the recommended question information.

[0085] In an alternative embodiment, the second model assistance information includes multiple example question information under the target scenario type; and / or, includes question description information for the target scenario type, where the question description information includes content limitation information and / or format limitation information for the recommended question information.

[0086] In an alternative embodiment, the question description information is determined according to the following steps: obtaining the historical conversation content in the target conversation page; based on the historical conversation content, determining at least one target keyword associated with the target scenario type; based on the target keyword and the model assistance configuration information corresponding to the target scenario type, determining the question description information.

[0087] In an alternative embodiment, before inputting the target model assistance information into the artificial intelligence model, the generation module 33 is further configured to: obtain the historical conversation content in the target conversation page; when the generation module 33 inputs the target model assistance information into the artificial intelligence model to generate the recommended question information, it is configured to: input the target model assistance information and the historical conversation content into the artificial intelligence model to generate the recommended question information.

[0088] In an alternative embodiment, when the generation module 33 inputs the target model assistance information and the historical conversation content into the artificial intelligence model to generate the recommended question information, it is configured to: screen out the most recent first historical conversation content from the obtained multiple historical conversation contents, and a second historical conversation content having a context association with the most recent first historical conversation content; input the target model assistance information, the first historical conversation content, and the second historical conversation content into the artificial intelligence model to generate the recommended question information.

[0089] In an alternative embodiment, the display condition for displaying the recommended question information on the target conversation page includes at least one of the following: determining that the target conversation page starts to be displayed currently; determining that a reply information is received from the artificial intelligence module for the question information or no new question information is received within a preset duration after receiving the reply information.

[0090] In an alternative embodiment, the device further includes a processing module 34, configured to: determine the question feature information of the user according to the trigger data of the user for the recommended question information; based on the question feature information, update the target model assistance information to assist the artificial intelligence model in generating new recommended question information based on the updated target model assistance information.

[0091] Descriptions of the processing flows of the various modules in the device and the interaction flows between the modules may refer to the relevant descriptions in the foregoing method embodiments and will not be elaborated here.

[0092] Embodiments of the present disclosure also provide a computer device, as Figure 4 shown, which is a schematic structural diagram of the computer device provided by the embodiments of the present disclosure, including:

[0093] a processor 10 and a memory 20; the memory 20 stores machine-readable instructions executable by the processor 10, and the processor 10 is configured to execute the machine-readable instructions stored in the memory 20. When the machine-readable instructions are executed by the processor 10, the processor 10 performs the following steps:

[0094] In response to meeting the display condition for displaying recommended question information on the target dialogue page, determine the target scenario type corresponding to the target dialogue page; based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type, determine the target model auxiliary information for generating the recommended question information; the second model auxiliary information corresponding to different scenario types is different; input the target model auxiliary information into the artificial intelligence model to generate the recommended question information.

[0095] The foregoing memory 20 includes an internal memory 210 and an external memory 220; the internal memory 210 here is also called the main memory, which is used to temporarily store the operation data in the processor 10 and the data exchanged with the external memory 220 such as a hard disk. The processor 10 exchanges data with the external memory 220 through the internal memory 210.

[0096] The specific execution process of the foregoing instructions may refer to the steps of the question recommendation method described in the embodiments of the present disclosure and will not be elaborated here.

[0097] Embodiments of the present disclosure also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the steps of the question recommendation method described in the foregoing method embodiments. Among them, the storage medium may be a volatile or non-volatile computer-readable storage medium.

[0098] Embodiments of the present disclosure also provide a computer program product, which carries program codes. The instructions included in the program codes can be used to execute the steps of the question recommendation method described in the foregoing method embodiments. Specifically, reference may be made to the foregoing method embodiments and will not be elaborated here.

[0099] Among them, the above computer program product can be specifically implemented in the form of hardware, software, or a combination thereof. In an alternative embodiment, the computer program product is specifically embodied as a computer storage medium. In another alternative embodiment, the computer program product is specifically embodied as a software product, such as a Software Development Kit (SDK), etc.

[0100] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems and devices can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein. In several embodiments provided by the present disclosure, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. Also, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some communication interfaces, and the indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0101] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0102] In addition, in each embodiment of the present disclosure, the functional units can be integrated in one processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0103] When the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such an understanding, the technical solution of the present disclosure, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present disclosure. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0104] Finally, it should be noted that the above-mentioned embodiments are only specific implementation manners of the present disclosure, used to illustrate the technical solutions of the present disclosure, rather than limiting them. The protection scope of the present disclosure is not limited thereto. Although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present disclosure can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A problem recommendation method, characterized in that, including: responding to meeting the display condition of displaying recommended question information on the target conversation page, determining the target scenario type corresponding to the target conversation page; based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type, determining the target model auxiliary information for generating the recommended question information; the second model auxiliary information corresponding to different scenario types is different; inputting the target model auxiliary information into an artificial intelligence model to generate the recommended question information; wherein, the second model auxiliary information includes multiple example question information under the target scenario type; and / or, includes question description information for the target scenario type, and the question description information includes content limitation information and / or format limitation information for the recommended question information.

2. The method according to claim 1, characterized in that The question description information is determined according to the following steps: obtaining the historical conversation content in the target conversation page; based on the historical conversation content, determining at least one target keyword associated with the target scenario type; based on the target keyword and the model auxiliary configuration information corresponding to the target scenario type, determining the question description information.

3. The method according to claim 1, wherein Before inputting the target model auxiliary information into the artificial intelligence model, the method further includes: obtaining the historical conversation content in the target conversation page; The inputting the target model auxiliary information into the artificial intelligence model to generate the recommended question information includes: inputting the target model auxiliary information and the historical conversation content into the artificial intelligence model to generate the recommended question information.

4. The method according to claim 3, characterized in that, The inputting the target model auxiliary information and the historical conversation content into the artificial intelligence model to generate the recommended question information includes: screening out the most recent first historical conversation content from the obtained multiple historical conversation contents, and a second historical conversation content having a context association with the most recent first historical conversation content; inputting the target model auxiliary information, the first historical conversation content and the second historical conversation content into the artificial intelligence model to generate the recommended question information.

5. The method according to claim 1, wherein The display condition of displaying the recommended question information on the target conversation page includes at least one of the following: determining that the display of the target conversation page starts currently; determining that a reply information is received for the question information from the artificial intelligence model or no new question information is received within a preset duration after receiving the reply information.

6. The method according to claim 1, wherein The method further includes: determining the question feature information of the user according to the trigger data of the user for the recommended question information; updating the target model auxiliary information based on the question feature information, so as to assist the artificial intelligence model to generate new recommended question information based on the updated target model auxiliary information.

7. A problem recommendation device, characterized in that, including: a first determination module, configured to determine the target scenario type corresponding to the target conversation page in response to meeting the display condition of displaying the recommended question information on the target conversation page; A second determination module, configured to determine target model auxiliary information for generating the recommended question information based on the first model auxiliary information adapted to each scenario type and the second model auxiliary information adapted to the target scenario type; the second model auxiliary information corresponding to different scenario types is different; A generation module, configured to input the target model auxiliary information into an artificial intelligence model to generate the recommended question information; Wherein, the second model auxiliary information includes multiple question information examples in the target scenario type; and / or, includes question description information for the target scenario type, and the question description information includes content limitation information and / or format limitation information for the recommended question information.

8. A computer device, characterized in that, Comprising: A processor and a memory, the memory stores machine-readable instructions executable by the processor, the processor is configured to execute the machine-readable instructions stored in the memory, and when the machine-readable instructions are executed by the processor, the processor executes the steps of the question recommendation method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, and when the computer program is run by a computer device, the computer device executes the steps of the question recommendation method according to any one of claims 1 to 6.

Images