Content display method and device based on intelligent model, equipment and medium

Abstract

The invention provides a content display method and device based on an intelligent model, equipment and a medium. Comprising the following steps: receiving a task requirement sent by a user, and sending the task requirement to a preset intelligent language model to obtain model original content output by the intelligent language model; keyword detection is carried out on the task requirements, and target keywords in the task requirements are determined; performing semantic similarity matching on the task requirements and the model original content to obtain a semantic similarity matching result; if the semantic similarity matching result meets the semantic similarity requirement, retrieving the original content of the model to obtain retrieval content and a preset number of content source links of the retrieval content; and displaying the original content of the model and the source link of the target content based on the retrieval content. In this way, the source of the original content generated by the intelligent language model is determined, and the use effect and the user satisfaction degree are improved.

Description

Technical Field

[0001] The present application relates to the field of natural language processing, and in particular to a content display method, device, equipment and medium based on an intelligent model. Background Art

[0002] The application of intelligent language models in the field of natural language processing is becoming more and more mature. They can reply to users with original content based on input questions or prompts. However, when users actually use them, they also need to determine the source of the content replied by the model.

[0003] When outputting original content, the existing intelligent language model will search for reply information that matches the user input information from multiple candidate information in the information library, and finally reply with the selected information content and output the content source of the information.

[0004] However, existing methods are easily limited by the information database when replying to the content source of the generated original content, and have the problem of poor performance. Summary of the invention

[0005] The present application provides a content display method, device, equipment and medium based on an intelligent model, which is used to solve the problem that the existing methods are easily restricted by the information library when replying to the content source of the generated original content, resulting in poor usage effect.

[0006] In a first aspect, the present application provides a content display method based on an intelligent model, the method comprising:

[0007] Receive task requirements sent by users, and send the task requirements to a preset intelligent language model to obtain original model content output by the intelligent language model;

[0008] Perform keyword detection on the task requirements to determine the target keywords in the task requirements, where the target keywords are keywords that represent the target query type;

[0009] Perform semantic similarity matching on the task requirements and the original content of the model to obtain semantic similarity matching results;

[0010] If the semantic similarity matching result meets the semantic similarity requirement, the original content of the model is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content;

[0011] Based on the search content, the model original content and the target content source link are displayed, wherein the target content source link is the content source link corresponding to the search content and the model original content.

[0012] In the embodiment of the present application, keyword detection is performed on the task requirements to determine the target keywords in the task requirements, including:

[0013] Identify key words in the task requirements;

[0014] Compare the semantic similarity between the keyword and the preset keyword, where the preset keyword is a preset keyword type that represents the retrieval of the original content of the model;

[0015] If the keyword and the preset keyword meet the comparison requirement of semantic similarity, the keyword is determined to be the target keyword.

[0016] In the embodiment of the present application, after comparing the semantic similarity between the keyword and the preset keyword, the method further includes:

[0017] If the keyword and the preset keyword do not meet the semantic similarity comparison requirements, the model's original content will be displayed to the user.

[0018] In the embodiment of the present application, after performing semantic similarity matching on the task requirements and the original content of the model to obtain the semantic similarity matching result, the method further includes:

[0019] If the semantic similarity matching result does not meet the semantic similarity requirement, the model original content is displayed to the user.

[0020] In the embodiment of the present application, based on the semantic similarity matching result meeting the semantic similarity requirement, the original content of the model is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0021] Determine the preset word count for text division;

[0022] Based on the preset word count of the text division, the model original content is divided to obtain multiple model original content segments, and the word count of the model original content segment is the same as the word count of the text division;

[0023] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0024] In the embodiment of the present application, based on the semantic similarity matching result meeting the semantic similarity requirement, the original content of the model is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0025] Determine the sentence composition of the model's original content based on the sentence structure of the model's original content;

[0026] Based on the composition of each sentence in the model original content, semantic extraction is performed on the model original content to obtain multiple model original content segments, wherein each model original content segment represents a complete semantic content;

[0027] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0028] In the embodiment of the present application, based on the semantic similarity matching result meeting the semantic similarity requirement, the original content of the model is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0029] Retrieve the original content of the model to obtain the retrieved content and a content source link sequence of the retrieved content;

[0030] A preset number of content source links that are ranked first are determined from the content source link sequence.

[0031] In a second aspect, the present application provides a content display device based on an intelligent model, the device comprising:

[0032] A sending module is used to receive task requirements sent by users, and send the task requirements to a preset intelligent language model to obtain the original model content output by the intelligent language model;

[0033] A detection module is used to detect keywords in the task requirements and determine the target keywords in the task requirements, where the target keywords are keywords that represent the target query type;

[0034] A matching module is used to perform semantic similarity matching between task requirements and original model content to obtain semantic similarity matching results;

[0035] A retrieval module, for retrieving the original content of the model based on the semantic similarity matching result meeting the semantic similarity requirement, and obtaining the retrieved content and a preset number of content source links of the retrieved content;

[0036] The display module is used to display the model original content and the target content source link based on the search content, wherein the target content source link is the content source link corresponding to the search content and the model original content.

[0037] In a third aspect, the present application provides a device, comprising: a processor, and a memory communicatively connected to the processor;

[0038] Memory stores computer-executable instructions;

[0039] The processor executes the computer-executable instructions stored in the memory to implement the method of the present application.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the method of the present application.

[0041] The content display method, device, equipment and medium based on the intelligent model provided in the present application receive task requirements sent by users, and send the task requirements to a preset intelligent language model to obtain the model original content output by the intelligent language model; perform keyword detection on the task requirements to determine the target keywords in the task requirements, where the target keywords are keywords that characterize the target query type; perform semantic similarity matching on the task requirements and the model original content to obtain semantic similarity matching results; based on the semantic similarity matching results satisfying the semantic similarity requirements, the model original content is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content; based on the retrieved content, the model original content and the target content source link are displayed, where the target content source link is the content source link corresponding to the retrieved content and the model original content.

[0042] In this way, the intelligent language model can be used to generate original model content according to the task requirements given by the user, and the task requirements can be detected by keywords, as well as semantic similarity matching with the model original content. Based on the detected target keywords and semantic similarity matching results, it can be determined that the original model content generated by the intelligent language model according to the task requirements needs to be retrieved, and based on the retrieval results, the content source of the original model content can be determined, and the model original content and content source link can be displayed to the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0044] Figure 1 A flowchart of a content display method based on an intelligent model provided in an embodiment of the present application;

[0045] Figure 2 A flowchart of another content display method based on an intelligent model provided in an embodiment of the present application;

[0046] Figure 3 A schematic diagram of the structure of a content display device based on an intelligent model provided in an embodiment of the present application;

[0047] Figure 4 This is a structural block diagram of a device for executing the intelligent model-based content display method according to an embodiment of the present application.

[0048] The above-mentioned drawings have shown clear embodiments of the present application, which will be described in more detail below; these drawings and textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0049] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0050] In the prior art, in the actual application process of the intelligent language model, the user will need to determine the content source of the model original content replied by the model. At this time, it is difficult for the intelligent language model to directly determine the source of the output in the training corpus, because the model original content generated by the intelligent language model is not retrieved from multiple candidate corpora, but the result is directly given by the trained model, and the training data set of the model is particularly large, and the source of the output content cannot be found in a short time.

[0051] In order to solve the above problems, an embodiment of the present application provides a content display method based on an intelligent model, which can detect keywords on the task requirements given by the user and match the semantic similarity with the generated model original content, so as to determine the content source of the model original content that the user needs to determine. At this time, the model original content is retrieved to determine the content source link of the model original content, and the model original content and the content source link are displayed to the user, so as to accurately provide the user with the content source, thereby realizing the determination of the source of the original content generated by the intelligent language model and improving the usage effect and user satisfaction.

[0052] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0053] This embodiment does not impose any special restrictions on the implementation method of the execution entity, as long as the execution entity can receive the task requirements sent by the user, and send the task requirements to the preset intelligent language model, to obtain the model original content output by the intelligent language model; perform keyword detection on the task requirements, and determine the target keywords in the task requirements, where the target keywords are keywords that characterize the target query type; perform semantic similarity matching on the task requirements and the model original content to obtain semantic similarity matching results; based on the semantic similarity matching results that meet the semantic similarity requirements, the model original content is retrieved to obtain the retrieved content and a preset number of content source links for the retrieved content; based on the retrieved content, the model original content and the target content source link are displayed, wherein the target content source link is the content source link corresponding to the retrieved content and the model original content.

[0054] Among them, the preset intelligent language model can be a robot based on the generative pre-trained model-3.5 (Generative Pre-Trained Transformer, GPT) and above principles as an intelligent language model, including but not limited to the generative pre-trained transformer model (Chat Generative Pre-trained Transformer, Chat GPT), ERNIE Bot, Google Natural Language Processing Model (Bidirectional Encoder Representations from Transformers, BERT) and other domestic and foreign chat robots.

[0055] Figure 1 A flowchart of a content display method based on an intelligent model provided in an embodiment of the present application. Figure 1 As shown, the content display method based on the intelligent model may include the following steps:

[0056] S110, receiving a task requirement sent by a user, and sending the task requirement to a preset intelligent language model to obtain original model content output by the intelligent language model.

[0057] Based on this, through the preset intelligent language model, it is possible to generate original model content according to the task requirements given by the user, thereby meeting the task requirements.

[0058] S120: Perform keyword detection on the task requirements to determine target keywords in the task requirements, where the target keywords are keywords that represent the target query type.

[0059] The keyword detection may be performed using a simple keyword detection model or other methods, as long as the keyword detection can be performed on the text content required by the task. This embodiment does not impose any limitation on this.

[0060] The target keyword represents the keyword for which the content source link needs to be displayed to the user; the target keyword can be a keyword required for tasks such as asking questions about encyclopedia knowledge and querying existing information such as the weather, while the non-target keyword can be a keyword required for tasks such as mathematical calculations, literary and artistic creation, and consulting, and this embodiment does not impose any restrictions on this; in addition, the target keyword can also be a keyword in a sentence in which the user clearly needs the content source in a given task requirement, such as a keyword in a sentence such as "I need to know the source of the content".

[0061] Based on this, by performing keyword detection on the task requirements, if the target keyword is detected, the task requirement is characterized as the type of content source that needs to be displayed to the user, so that the content source can be displayed to the user later.

[0062] In some implementations, performing keyword detection on the task requirements to determine target keywords in the task requirements includes:

[0063] Identify key words in the task requirements;

[0064] Compare the semantic similarity between the keyword and the preset keyword, where the preset keyword is a preset keyword type that represents the retrieval of the original content of the model;

[0065] If the keyword and the preset keyword meet the comparison requirement of semantic similarity, the keyword is determined to be the target keyword.

[0066] Among them, the preset keywords can be keywords required for tasks such as asking questions about encyclopedia knowledge and querying existing information such as weather, or they can be keywords for sentences in given task requirements where the user clearly requires the source of content.

[0067] The semantic similarity comparison requirement is that the semantic similarity satisfies a preset similarity threshold, which is used to determine whether the keywords required by the task and the preset keywords have a high semantic similarity; the preset similarity threshold can be initially set or can be set by the user.

[0068] Based on this, by comparing the keywords in the task requirements with the preset keywords, the target keywords in the task requirements are determined, so that the task requirements are determined to be the task type that requires displaying content source links to users.

[0069] In this implementation, after comparing the semantic similarity between the keyword and the preset keyword, the method further includes:

[0070] If the keyword and the preset keyword do not meet the semantic similarity comparison requirements, the model's original content will be displayed to the user.

[0071] Based on this, if the semantic similarity between the keywords in the task requirements and the preset keywords does not meet the preset similarity threshold, the original model content generated by the intelligent language model will be displayed to the user without the need to link to the source of the displayed content.

[0072] S130, performing semantic similarity matching on the task requirements and the original content of the model to obtain a semantic similarity matching result.

[0073] Among them, semantic similarity matching includes but is not limited to automatic methods such as comparison using existing artificial intelligence models, comparison using the angle size of word vectors, and comparison using dictionary matching queries. As long as the method can be used to compare the text similarity between the task requirements and the original content of the model, this embodiment does not impose any restrictions on this.

[0074] Based on this, by comparing the semantic similarity between the task requirements and the original content of the model, the semantic similarity between the two can be determined, which will help to retrieve the original content of the model based on the semantic similarity matching results.

[0075] In some implementations, after performing semantic similarity matching on the task requirements and the original content of the model to obtain a semantic similarity matching result, the method further includes:

[0076] If the semantic similarity matching result does not meet the semantic similarity requirement, the model original content is displayed to the user.

[0077] Among them, the semantic similarity requirement is that the semantic similarity satisfies a preset similarity threshold, which is used to determine that there is a high semantic similarity between the task requirements and the original content of the model; the preset similarity threshold can be initially set or set by the user.

[0078] Based on this, if the task requirements and the model original content do not meet the preset similarity threshold, that is, the semantic similarity between the task requirements and the model original content does not meet the similarity threshold, the model original content will be displayed to the user without the need to link to the source of the displayed content.

[0079] S140: If the semantic similarity matching result meets the semantic similarity requirement, the original content of the model is searched to obtain the searched content and a preset number of content source links of the searched content.

[0080] Among them, the search is not limited to an existing search engine. Any network search engine or search comparison technology can be used for search, including but not limited to Baidu, Google, Bing and other network search engines.

[0081] The retrieval content is obtained by dividing the original content of the model to facilitate better retrieval of the content.

[0082] The preset quantity is a pre-set quantity, which may be initially set or set by the user.

[0083] Based on this, by determining that the semantic similarity matching results meet the semantic similarity requirements, the search content is searched based on the retrieval module to obtain the content source of the original content of the model.

[0084] In some implementations, based on the semantic similarity matching result satisfying the semantic similarity requirement, the model original content is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0085] Determine the preset word count for text division;

[0086] Based on the preset word count of the text division, the model original content is divided to obtain multiple model original content segments, and the word count of the model original content segment is the same as the word count of the text division;

[0087] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0088] Among them, the preset text division word count is a pre-set text word count for dividing the model original content, and the model original content is divided into one section according to the preset text division word count, thereby obtaining multiple model original content sections; in addition, if the text word count of the last model original content section does not reach the preset text division word count, the user can search the last section as a separate model original content section, or merge the last section with the previous model original content section and then search again, which is not limited in this embodiment.

[0089] Based on this, the model original content is extracted and divided based on the preset number of words of the text division, thereby obtaining multiple model original content segments, so as to retrieve according to the model original content segments and obtain content source links.

[0090] In other implementations, based on the semantic similarity matching result satisfying the semantic similarity requirement, the model original content is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0091] Determine the sentence composition of the model's original content based on the sentence structure of the model's original content;

[0092] Based on the composition of each sentence in the model original content, semantic extraction is performed on the model original content to obtain multiple model original content segments, wherein each model original content segment represents a complete semantic content;

[0093] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0094] Among them, the sentence structure is the grammatical structure of the sentence, which may include subject, predicate, object, attributive, adverbial, complement and other sentence structures, and the sentence composition is the text content composition of the sentence; by dividing the structure of the sentence, the composition of the sentence in the original content of the model is determined.

[0095] Semantic extraction can be understood as dividing the sentences of the model's original content according to the sentence structure of the model's original content, so that the obtained model's original content segments represent complete semantic content.

[0096] Based on this, the model original content is divided according to semantic integrity to obtain multiple model original content segments. Each model original content segment represents complete semantic content, so that the model original content segments are retrieved to obtain content source links.

[0097] In other implementations, based on the semantic similarity matching result satisfying the semantic similarity requirement, the model original content is retrieved to obtain the retrieved content and a preset number of content source links of the retrieved content, including:

[0098] Retrieve the original content of the model to obtain the retrieved content and a content source link sequence of the retrieved content;

[0099] A preset number of content source links that are ranked first are determined from the content source link sequence.

[0100] The content source link sequence is a sequence composed of multiple content source links obtained through retrieval. For example, after searching for the original content of the model in a search engine, the retrieval result sequence obtained is the content source link sequence.

[0101] The content source link ranked first is closest to the original content of the model and can better meet user needs.

[0102] Based on this, a content source link sequence is obtained by retrieving the original content of the model, so as to determine a preset number of content source links that are ranked first and display them to the user.

[0103] S150. Based on the search content, display the model original content and the target content source link, wherein the target content source link is the content source link corresponding to the search content and the model original content.

[0104] The target content source link is a content source link determined from the retrieved content source links and used for displaying to the user.

[0105] Based on this, by determining the content source link of the model's original content, the model's original content and the target content source link are displayed to users.

[0106] In this embodiment, by receiving task requirements sent by users and sending the task requirements to a preset intelligent language model, the model original content output by the intelligent language model is obtained; after obtaining the model original content, the task requirements are subjected to keyword detection and compared with preset keywords; and semantic similarity matching is performed between the task requirements and the model original content, so as to determine that the target keywords in the task requirements and the semantic similarity matching results meet the semantic similarity requirements, and then the model original content is divided into model original content segments according to the preset text division word count or sentence composition and retrieved, so as to obtain the content source link, and display the model original content and the content source link to the user.

[0107] In this way, by performing keyword detection on the task requirements, the content source of the model original content that the user needs to obtain is determined, so as to meet the user's needs to view the content source during the actual application process; and the task requirements and the model original content are matched for semantic similarity, so as to determine that the generated model original content meets the given task requirements, so as to subsequently determine the content source of the model original content; and the model original content is divided according to the preset number of words in the text or the sentence composition of the model original content, so as to search the model original content segments separately, so as to ensure that the search results are more accurate; when the user gives the task requirements, the corresponding model original content can be obtained, and the source link of the content will be given when the user needs it.

[0108] Figure 2 A flowchart of another method for displaying content based on an intelligent model provided in an embodiment of the present application. Figure 2 As shown, the content display method based on the intelligent model may include the following steps:

[0109] S210. Obtain the original return result of the artificial intelligence large model chat robot according to the user input.

[0110] Among them, the artificial intelligence large-model chatbot can use robots based on the principles of GPT-3.5 and above as artificial intelligence large-model chatbots, including but not limited to Chat GPT, ERNIE Bot, Bard and other domestic and foreign chatbots.

[0111] Based on this, the original return result is determined by giving user input to the artificial intelligence large model chatbot.

[0112] S220, the user input passes through the classification module at the same time to determine whether it needs to be replied to the source, if so, proceed to the next step, otherwise directly output the original return result.

[0113] Among them, the classification module can be a natural language processing model based on deep learning network training, or it can be a simple keyword detection model. Its function is to determine whether the user's input needs to be replied to the source. Common input sentences that need to be replied to the source include but are not limited to encyclopedia knowledge questions, queries for existing information such as weather, etc. Common input sentences that do not require a reply to the source include but are not limited to mathematical calculations, literary and artistic creation, consulting, etc.

[0114] Furthermore, determining whether a reply source is needed also depends on whether the original reply result has a high degree of match or overlap with the user input content.

[0115] Based on this, if the classification module determines that the user input of the reply source does not need to be output, the original reply result is directly output as the result to the user.

[0116] S230: Use the original returned results as input to the search comparison module to obtain a preset number of result links.

[0117] Among them, the search comparison module is not limited to an existing search engine, and any network search engine or search comparison technology can be used, including but not limited to Baidu, Google, Bing and other network search engines.

[0118] The preset number may be initially set or determined by the user.

[0119] S240: Combine the original returned result with a preset number of result links, and output the result to the user as a new returned result.

[0120] Based on this, the original returned results and a preset number of search results are merged and linked to display the content to the user.

[0121] S250. The user obtains an output result including a reply source, and can view the reply source by clicking a link.

[0122] Based on this, users can query the content source of the content generated by the artificial intelligence large model chatbot by clicking on the displayed content and content links.

[0123] In this embodiment, a classification module and a search and comparison module are added to the artificial intelligence large model chat robot to form a new chat robot; when the user inputs a sentence into the new chat robot, the original output result is first obtained through the artificial intelligence large model chat robot; at the same time, the user input sentence is also classified by the classification module to determine whether the reply to the input sentence needs to be attached with a source link. If it needs to be attached, continue with the following steps, otherwise the original result is directly output; when the reply source needs to be output, the original output result will be used as the input of the search module, and the result link will be obtained through the network search engine, and the result link will be merged with the original output result as a new output result.

[0124] As a result, we get a chat robot that is mainly based on a large artificial intelligence model and supplemented by a classification module and a search and comparison module, which can output the source of replies. When the user enters a sentence, the corresponding reply can be obtained based on the input content, and the source of the reply content can be given when necessary.

[0125] Figure 3 A schematic diagram of the structure of a content display device 300 based on an intelligent model provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the intelligent model-based content display device 300 includes: a sending module 310, a detection module 320, a matching module 330, a retrieval module 340 and a display module 350.

[0126] The sending module 310 is used to receive the task requirements sent by the user, and send the task requirements to the preset intelligent language model to obtain the model original content output by the intelligent language model;

[0127] A detection module 320 is used to perform keyword detection on the task requirements and determine target keywords in the task requirements, where the target keywords are keywords that represent the target query type;

[0128] A matching module 330 is used to perform semantic similarity matching on the task requirements and the original content of the model to obtain a semantic similarity matching result;

[0129] A retrieval module 340 is used to retrieve the original content of the model based on the semantic similarity matching result meeting the semantic similarity requirement, and obtain the retrieved content and a preset number of content source links of the retrieved content;

[0130] The display module 350 is used to display the model original content and the target content source link based on the search content, wherein the target content source link is the content source link corresponding to the search content and the model original content.

[0131] In the embodiment of the present application, the detection module 320 may also be specifically used for:

[0132] Identify key words in the task requirements;

[0133] Compare the semantic similarity between the keyword and the preset keyword, where the preset keyword is a preset keyword type that represents the retrieval of the original content of the model;

[0134] If the keyword and the preset keyword meet the comparison requirement of semantic similarity, the keyword is determined to be the target keyword.

[0135] In the embodiment of the present application, the detection module 320 may also be specifically used for:

[0136] If the keyword and the preset keyword do not meet the semantic similarity comparison requirements, the model's original content will be displayed to the user.

[0137] In the embodiment of the present application, the intelligent model-based content display device 300 can also be specifically used for:

[0138] If the semantic similarity matching result does not meet the semantic similarity requirement, the model original content is displayed to the user.

[0139] In the embodiment of the present application, the search module 340 may also be specifically used for:

[0140] Determine the preset word count for text division;

[0141] Based on the preset word count of the text division, the model original content is divided to obtain multiple model original content segments, and the word count of the model original content segment is the same as the word count of the text division;

[0142] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0143] In the embodiment of the present application, the search module 340 may also be specifically used for:

[0144] Determine the sentence composition of the model's original content based on the sentence structure of the model's original content;

[0145] Based on the composition of each sentence in the model original content, semantic extraction is performed on the model original content to obtain multiple model original content segments, wherein each model original content segment represents a complete semantic content;

[0146] The original content segment of the model is searched to obtain the content source link of the original content segment of the model.

[0147] In the embodiment of the present application, the search module 340 may also be specifically used for:

[0148] Retrieve the original content of the model to obtain the retrieved content and a content source link sequence of the retrieved content;

[0149] A preset number of content source links that are ranked first are determined from the content source link sequence.

[0150] Figure 4 This is a schematic diagram of the structure of the device provided in the embodiment of the present application. Figure 4 As shown, the device 400 includes:

[0151] The device 400 may include a processor 401 with one or more processing cores, a memory 402 with one or more computer-readable storage media, a communication component 403 and other components. The processor 401 , the memory 402 and the communication component 403 are connected via a bus 404 .

[0152] In a specific implementation process, at least one processor 401 executes the computer execution instructions stored in the memory 402, so that at least one processor 401 executes the above message processing method.

[0153] The specific implementation process of the processor 401 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0154] In the above Figure 4 In the illustrated embodiment, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present application may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0155] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0156] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0157] In some embodiments, a computer program product is also proposed, including a computer program or instructions, which, when executed by a processor, implements the steps in any of the above-mentioned intelligent model-based content display methods.

[0158] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0159] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0160] To this end, an embodiment of the present application provides a computer-readable storage medium, which stores multiple computer execution instructions, and the computer execution instructions can be loaded by a processor to execute the steps in any one of the intelligent model-based content display methods provided in the embodiments of the present application.

[0161] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0162] According to one aspect of the present application, a computer program product or a computer program is provided. The computer program product or the computer program comprises computer instructions stored in a computer-readable storage medium.

[0163] Since the instructions stored in the storage medium can execute the steps in any one of the intelligent model-based content display methods provided in the embodiments of the present application, the beneficial effects that can be achieved by any one of the intelligent model-based content display methods provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0164] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0165] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A content display method based on an intelligent model, characterized in that: The method comprises: Receiving a task requirement sent by a user, and sending the task requirement to a preset intelligent language model, to obtain the model original content output by the intelligent language model; Perform keyword detection on the task requirement to determine a target keyword in the task requirement, where the target keyword is a keyword that represents a target query type; Performing semantic similarity matching on the task requirements and the original content of the model to obtain a semantic similarity matching result; Based on the semantic similarity matching result satisfying the semantic similarity requirement, the original content of the model is searched to obtain the searched content and a preset number of content source links of the searched content; Based on the search content, the model original content and the target content source link are displayed, wherein the target content source link is a content source link corresponding to the search content and the model original content.

2. The method according to claim 1, characterized in that The performing keyword detection on the task requirement to determine target keywords in the task requirement includes: Identify key words in the task requirements; Performing a semantic similarity comparison between the keyword and a preset keyword, wherein the preset keyword is a preset keyword type that represents the retrieval of the original content of the model; If the keyword and the preset keyword meet the comparison requirement of the semantic similarity, the keyword is determined to be the target keyword.

3. The method according to claim 2, characterized in that After comparing the semantic similarity between the keyword and the preset keyword, the method further includes: If the keyword and the preset keyword do not meet the comparison requirement of the semantic similarity, the original content of the model is displayed to the user.

4. The method according to claim 1, characterized in that: After performing semantic similarity matching on the task requirement and the original content of the model to obtain a semantic similarity matching result, the method further includes: If the semantic similarity matching result does not meet the semantic similarity requirement, the model original content is displayed to the user.

5. The method according to claim 1, characterized in that The method of searching the original content of the model based on the semantic similarity matching result satisfying the semantic similarity requirement to obtain the searched content and a preset number of content source links of the searched content includes: Determine the preset word count for text division; Based on the preset number of words for the text divisions, the model original content is divided to obtain a plurality of model original content segments, wherein the number of words in the model original content segments is the same as the number of words for the text divisions; The model original content segment is searched to obtain a content source link of the model original content segment.

6. The method according to claim 1, characterized in that The method of searching the original content of the model based on the semantic similarity matching result satisfying the semantic similarity requirement to obtain the searched content and a preset number of content source links of the searched content includes: Determining the sentence composition of the model original content based on the sentence structure of the model original content; Based on the composition of each sentence in the model original content, semantic extraction is performed on the model original content to obtain a plurality of model original content segments, wherein each of the model original content segments represents a complete semantic content; The model original content segment is searched to obtain a content source link of the model original content segment.

7. The method according to claim 1, characterized in that The method of searching the original content of the model based on the semantic similarity matching result satisfying the semantic similarity requirement to obtain the searched content and a preset number of content source links of the searched content includes: Retrieving the original content of the model to obtain the retrieved content and a content source link sequence of the retrieved content; The preset number of content source links ranked first are determined from the content source link sequence.

8. A content display device based on an intelligent model, characterized in that: The device comprises: A sending module, used to receive a task requirement sent by a user, and send the task requirement to a preset intelligent language model to obtain the model original content output by the intelligent language model; A detection module, used to perform keyword detection on the task requirements and determine target keywords in the task requirements, wherein the target keywords are keywords that represent target query types; A matching module, used for performing semantic similarity matching between the task requirements and the original content of the model to obtain a semantic similarity matching result; A retrieval module, configured to retrieve the original content of the model based on the semantic similarity matching result satisfying the semantic similarity requirement, and obtain the retrieved content and a preset number of content source links of the retrieved content; A display module is used to display the model original content and the target content source link based on the search content, wherein the target content source link is the content source link corresponding to the search content and the model original content.

9. A device, characterized in that: include: one or more processors; Memory; One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions can be called by a processor to execute the method according to any one of claims 1 to 7.

Images