Methods, apparatus, devices, and storage media for generating target information
By receiving and recognizing information generation instructions, generating question information and combining it with response information and candidate files, the problem of low accuracy of target information in existing technologies is solved, and more efficient and accurate target information generation is achieved.
Patent Information
- Application Number
- CN202310621513.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-29
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-05-29
AI Technical Summary
In existing technologies, due to the significant differences in sample data across different types and application scenarios, the accuracy of target information generated by the target information generation model after training is not high.
By receiving information generation instructions, identifying the purpose of information generation and/or candidate documents, generating question information, and generating target information based on the response information, candidate documents, and question information.
It improves the accuracy of target information, reduces the cost and time wastage caused by blind generation, and expands the applicability of the generation method.
Smart Images

Figure CN116775959B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of computer technology, and in particular to artificial intelligence technology, intelligent search, and other fields. Background Technology
[0002] With the continuous development and popularization of artificial intelligence and computer technologies, generating target information via the internet has become one of the main methods for target information generation. Existing technologies utilize trained target information generation models to achieve target information generation. However, these trained target information generation models need to be trained based on a large amount of pre-acquired sample data.
[0003] However, since there are significant differences between different types of sample data and sample data from different application scenarios, directly using the target information trained above to generate the target information generated by the model often results in low accuracy. Summary of the Invention
[0004] This disclosure provides a method, apparatus, device, and storage medium for generating target information.
[0005] According to one aspect of this disclosure, a method for generating target information is provided, comprising:
[0006] Receive information generation instructions, which indicate the purpose of information generation and / or candidate files;
[0007] Identify information generation instructions to determine the purpose of information generation and / or candidate files;
[0008] Based on the information generation purpose and / or candidate files, generate the question information for the information generation instruction;
[0009] Receive response information in response to the question information, and generate target information based on at least one of the response information, candidate documents, and question information.
[0010] According to another aspect of this disclosure, an apparatus for generating target information is provided, comprising:
[0011] The receiving module is used to receive information generation instructions, which indicate the purpose of information generation and / or candidate files.
[0012] An identification module is used to identify information generation instructions to determine the information generation purpose and / or candidate files; a first generation module is used to generate question information for the information generation instructions based on the information generation purpose and / or candidate files; receive response information in response to the question information, and generate target information based on at least one of the response information, candidate files, and question information.
[0013] According to another aspect of this disclosure, an electronic device is provided, comprising:
[0014] At least one processor; and
[0015] The memory is communicatively connected to the at least one processor; wherein,
[0016] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the methods of any embodiment of the present disclosure.
[0017] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform a method according to any embodiment of this disclosure.
[0018] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements a method according to any embodiment of this disclosure.
[0019] This disclosure allows for the generation of information purpose and / or candidate documents based on information generation instructions; and the determination of target information using the information generation purpose and / or candidate documents. The target information generation method proposed in this disclosure can reduce the problems of low accuracy and high cost associated with blindly generating target information.
[0020] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description
[0021] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:
[0022] Figure 1 This is an application diagram according to an embodiment of the present disclosure;
[0023] Figure 2 This is an implementation flow of a target information generation method according to an embodiment of the present disclosure. Figure 1 ;
[0024] Figure 3A This is a schematic block diagram of a method for generating target information according to an embodiment of the present disclosure. Figure 1 ;
[0025] Figure 3B This is a schematic block diagram of a method for generating target information according to an embodiment of the present disclosure. Figure 2 ;
[0026] Figure 3C This is an implementation flow of a target information generation method according to an embodiment of the present disclosure. Figure 2 ;
[0027] Figure 3D This is a schematic block diagram three of a method for generating target information according to an embodiment of the present disclosure;
[0028] Figure 3E This is a schematic block diagram of a method for generating target information according to an embodiment of the present disclosure. Figure 4
[0029] Figure 4 This is a schematic diagram illustrating the application field of a method for generating target information according to an embodiment of the present disclosure;
[0030] Figure 5 This is a flowchart of the implementation of a method for generating target information according to an embodiment of the present disclosure;
[0031] Figure 6 This is a schematic block diagram of a method for generating target information according to an embodiment of the present disclosure. Figure 5 ;
[0032] Figure 7 This is a schematic diagram of the structure of a target information generation apparatus 700 according to an embodiment of the present disclosure;
[0033] Figure 8 This is a schematic diagram of the structure of a target information generation apparatus 800 according to an embodiment of the present disclosure;
[0034] Figure 9 A schematic block diagram of an example electronic device 900 that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation
[0035] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.
[0036] In this document, the term "and / or" merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The term "at least one" in this document indicates any combination of at least two of a plurality of elements. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C. The terms "first" and "second" in this document refer to and distinguish between multiple similar technical terms, not implying a specific order or limiting the number of elements to two. For example, "first feature" and "second feature" refer to two categories / two features, where the first feature can be one or more, and the second feature can also be one or more.
[0037] To better illustrate this disclosure, some of the terms or nouns used in the embodiments of this application will be explained first:
[0038] (1) Artificial Intelligence (AI): Artificial intelligence is a new technological science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence.
[0039] (2) Large Language Model (LLM): Also known as a large-scale language model, it is a massive artificial intelligence model. Large language models aim to understand and generate human language. They can be trained on deep learning frameworks and large amounts of sample data to perform various types of tasks. For example, large language models can be applied to tasks such as text summarization, text translation, and text generation.
[0040] With the continuous development and popularization of artificial intelligence and computer technologies, the methods of generating target information are gradually becoming more intelligent. Currently, generating target information via the internet has become one of the main methods for generating target information. For example, in existing technologies, if it is necessary to generate multiple test questions (i.e., target information) for any given test point, multiple test questions for that test point can be generated manually or via the internet. It should be noted that compared to manually generating multiple test questions for a given test point, generating multiple test questions for that test point via the internet has the advantages of requiring less time, fewer resources, and higher accuracy.
[0041] Currently, in existing technologies, target information can generally be generated using a trained target information generation model. For example, people can send text and / or voice commands to a trained target information generation model to generate target information. Specifically, text and / or voice commands (i.e., information generation commands) such as "generate an article on the topic of my father," "generate an English dialogue," and "generate test questions related to power series" can be sent to the trained target information generation model. After receiving the text and / or voice commands, the trained target information generation model can perform speech recognition on the voice commands or text recognition on the text commands to obtain the recognition results corresponding to the text and / or voice commands, and generate target information based on the recognition results. In existing technologies, to obtain a trained target information generation model, it is generally necessary to acquire a large amount of sample data in advance, train the target information generation model based on the sample data, and then use the trained target information generation model to generate target information. However, due to the significant differences between different types of sample data and sample data from different application scenarios, the target information generated by the target information generation model trained above often suffers from low accuracy.
[0042] Therefore, to solve the above problems, this disclosure proposes a method for generating target information. The target information generation method proposed in this disclosure can be applied to, for example... Figure 1In the system architecture 100 shown, the system architecture 100 may include a terminal device 110 and a server 120. The terminal device 110 and the server 120 can communicate via any type of wired or wireless network. The terminal device 110 can be used to run applications; in this embodiment, these applications may include cloud storage applications, file management applications, download tool applications, and search applications, etc. The server 120 can be used to provide background management for the applications running on the terminal device 110. For example, the server 120 can respond to information generation instructions sent by the terminal device 110 and generate target information. This embodiment does not impose a specific limitation on the number of terminal devices 110. For example, the terminal devices 110 proposed in this embodiment include, but are not limited to, mobile phones, computers, smart voice interaction devices, smart home appliances, in-vehicle terminals, game consoles, e-book readers, multimedia playback devices, wearable devices, and other electronic devices. In some implementations, server 120 may be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.
[0043] Figure 2 This is an implementation flow of a target information generation method according to an embodiment of the present disclosure. Figure 1 ,include:
[0044] S210. Receive information generation instruction, which is used to indicate the purpose of information generation and / or candidate files;
[0045] S220, Identify information generation instructions to determine the purpose of information generation and / or candidate documents;
[0046] S230. Generate the problem information for the information generation instruction based on the information generation purpose and / or candidate files;
[0047] S240, Receive response information in response to the question information, and generate target information based on at least one of the response information, candidate files, and question information.
[0048] It should be noted that the method for generating target information proposed in this embodiment can be applied to cloud storage. Cloud storage, also known as network USB flash drive or network hard drive, can be used for file storage, file backup, file access, and file sharing.
[0049] The target information generation method proposed in this disclosure can cover a wider range of long-tail learning content (such as generating highly accurate target information for both limited and abundant information in candidate files); and saves the cost and time required to generate target information.
[0050] In one example of S210-S240, the information generation instruction can be sent by a terminal device, and the information generation instruction includes a voice instruction and / or a first text instruction; the voice instruction and / or the first text instruction are used to indicate the purpose of the information generation and / or candidate files.
[0051] For example, the information generation instruction may include a first text instruction input by the user terminal, such as "Generate an article on the topic of my father." Alternatively, the information generation instruction may include a voice instruction input by the user terminal; for example, the information generation instruction may include a 5-second voice message.
[0052] Using voice commands and / or first text commands as information generation commands not only makes it more convenient, faster, and more timely to generate target information, but also improves the efficiency of acquiring target information through timely processing of information generation commands.
[0053] It should be noted that, due to the diversity and complexity of information generation instructions, the embodiments of this disclosure can either directly parse the information generation instructions to identify the information generation purpose and / or candidate file indicated by the instructions; or they can first perform further processing on the information generation instructions, and then obtain the information generation purpose and / or candidate file by parsing the processed result.
[0054] Specifically, the identification information generation instruction, which determines the purpose of information generation and / or candidate files, may include:
[0055] The voice command is converted into a second text command; the second text command is parsed to determine the purpose of information generation and / or candidate files; and / or,
[0056] Parse the first text instruction to determine the purpose of information generation and / or candidate files.
[0057] The voice command can include any language, such as Chinese, English, and German; or, the voice command can include any dialect.
[0058] In one example, a voice command can be converted into a second text command in the following way:
[0059] Method 1: Use a speech-to-text algorithm to convert the speech command into a second text command corresponding to the speech command.
[0060] Method 2: Implement voice conversion of voice commands by calling existing voice conversion models and / or applications (APP, Application). For example, use the voice conversion function of a large language model (LLM) to convert the voice command into a second text command corresponding to the voice command.
[0061] It should be noted that the embodiments disclosed herein do not limit how voice commands are converted into speech; the above are merely examples.
[0062] By using voice commands and / or first text commands to generate information generation objectives and / or candidate files, the recognition accuracy of information generation commands is improved to some extent, and the applicable scenarios for information generation commands are expanded.
[0063] Of course, it should be noted that the embodiments of this disclosure do not limit whether the information generation instruction includes candidate files. For example, the information generation instruction proposed in the embodiments of this disclosure may include candidate files; or, the information generation instruction proposed in the embodiments of this disclosure may not include candidate files, but may include information that can directly indicate the candidate file, such as the storage address, storage address link, identification information, and name of the candidate file. In addition, the embodiments of this disclosure do not limit the file type of the candidate files included or indicated in the information generation instruction.
[0064] Specifically, in this embodiment of the disclosure, the candidate file may include at least one of a first text file, an image file, an audio file, and a video file.
[0065] By employing multiple types of candidate files, the scope of candidate files proposed in this disclosure can be expanded. In other words, the target information generation method proposed in this disclosure can generate target information for multiple types of candidate files, thereby improving the accuracy of the target information.
[0066] For example, such as Figure 3A As shown, if the information generation instruction includes "generate relevant test questions for the video file 'The Fundamental Laws of Materialist Dialectics'", then the candidate files for the information generation instruction can include the video file "The Fundamental Laws of Materialist Dialectics"; or, if the information generation instruction includes "generate relevant test questions for the text file 'An Introduction to the Basic Principles of Marxism, Chapter 1'", then the candidate files for the information generation instruction can include the text file "An Introduction to the Basic Principles of Marxism, Chapter 1".
[0067] In some embodiments, the candidate files involved in this disclosure may be stored in devices and / or servers with storage capabilities, such as cloud storage, cloud drives, terminal devices, and the Internet. In this disclosure, to broaden the scope of candidate files covered by the embodiments, for files not stored in devices and / or servers with storage capabilities such as cloud storage, cloud drives, terminal devices, and the Internet, this disclosure may also employ a scanning method to obtain the corresponding candidate files. Specifically, for files not stored in devices and / or servers with storage capabilities such as cloud storage, cloud drives, terminal devices, and the Internet, obtaining the corresponding candidate files by scanning includes:
[0068] Receive scan request;
[0069] Perform a scan based on the scan request to identify the candidate file.
[0070] In one example, embodiments of this disclosure can be performed by receiving an address such as Figure 3A Clicking the scan icon shown will trigger a scan request; alternatively, a scan request can be received via voice and / or text commands.
[0071] By using a scanning method to obtain candidate files, the candidate files can be processed even when there are no corresponding text files, image files, audio files, or video files, thereby further expanding the scope of application of the target information generation method proposed in this disclosure.
[0072] Furthermore, after determining the information generation purpose and / or candidate documents indicated by the information generation instruction, embodiments of this disclosure can also generate problem information for the information generation instruction based on the information generation purpose and / or candidate documents.
[0073] Still Figure 3A As shown, if the information generation instruction includes "generate test questions related to the video file 'The Fundamental Laws of Materialist Dialectics'", that is, in this information generation instruction, the information generation purpose includes "generating test questions", and the candidate file includes "the video file 'The Fundamental Laws of Materialist Dialectics'", then the question information of the information generation instruction can include any test questions related to the video text "The Fundamental Laws of Materialist Dialectics", for example, such as... Figure 3A The question may include: "How do you understand that if you grasp the principal contradiction, all problems will be solved?"; or any multiple-choice, true / false, or short-answer questions related to the video text "The Fundamental Laws of Materialist Dialectics".
[0074] Furthermore, in this embodiment of the disclosure, after generating the question information of the information generation instruction according to the information generation purpose and / or candidate file, it is also possible to receive the response information in response to the question information, and use the response information to generate the target information in subsequent operations.
[0075] In one example, the response information may include information sent by the terminal device in response to the question. It should be noted that the embodiments of this disclosure do not limit the specific form of the response information; for example, the response information may include a first text response and / or a voice response.
[0076] Still Figure 3A As shown, in response to the question "How can we understand that grasping the principal contradiction will solve all problems?", the answer could include information such as... Figure 3A The statement indicates that "within a group of contradictions, there exist principal contradictions and secondary contradictions. The principal contradiction occupies a dominant position in the contradiction system and plays a decisive role in the development of things; the principal contradiction and the secondary contradiction are interdependent and mutually influential, and can transform into each other under certain conditions."
[0077] Regarding the response information to the question, this embodiment of the disclosure may combine candidate files and question information to further process the response information in order to generate target information. Specifically, this embodiment of the disclosure may include: generating target information based on at least one of the response information, candidate files, and question information.
[0078] Furthermore, since the response information in this embodiment may include first text response information and / or voice response information, and the target information generation method proposed in this embodiment differs in its specific processing flow for the first text response information and / or voice response information (e.g., compared to obtaining target information using at least one of the first text response information, candidate question files, and question information; the specific processing flow for obtaining target information using at least one of the voice response information, candidate question files, and question information is more complex, i.e., obtaining target information using at least one of the voice response information, candidate question files, and question information can add a step of converting the voice response information into second text response information). Therefore, for the case where the response information includes first text response information and / or voice response information, this embodiment proposes methods for generating target information for both first text response information and voice response information.
[0079] Specifically, the method for generating target information based on at least one of response information, candidate documents, and question information, as proposed in this embodiment of the disclosure, includes:
[0080] Generate target information based on at least one of the first text response information, candidate documents, and question information; and / or,
[0081] The voice response information is converted to obtain the corresponding second text response information; and the target information is generated based on at least one of the second text response information, candidate files, and question information.
[0082] For example, such as Figure 3B As shown, if the information generation instruction includes "generate exam questions related to the text file 'Advanced Mathematics'", then the candidate file for this information generation instruction can include the text file "Advanced Mathematics". In this case, the question information (such as...) can be generated first based on the text file "Advanced Mathematics". Figure 3B The concept of "explanation vector" is shown, and secondly, the first text response information (such as...) can be received regarding the question. Figure 3B The target information can be generated from at least one of the following: the first text response information (i.e., the "quantity with size and direction"), the candidate file, and the question information.
[0083] Alternatively, if the question information is "the concept of an explanation vector" and the received voice response information is "3 seconds of voice", then the voice response information (i.e. "3 seconds of voice") can be converted to obtain the second text response information corresponding to the "3 seconds of voice"; and the target information can be generated based on at least one of the second text response information, the candidate file and the question information.
[0084] The target information generation method proposed in this disclosure can receive various types of response information and generate target information based on these various types of response information, thereby improving the efficiency of target information generation and expanding the applicability of the target information generation method.
[0085] Furthermore, embodiments of this disclosure can also generate target information using different generation methods based on different types of response information. Therefore, the target information generation method proposed in embodiments of this disclosure can perform targeted processing on various types of response information, thereby improving the efficiency and accuracy of generating target information using any type of response information.
[0086] In addition, besides the response information which may include various types of information (such as text information and voice information), the candidate files proposed in this disclosure embodiment may also include various types of files, such as at least one of a first text file, an image file, an audio file, and a video file.
[0087] Therefore, corresponding to the above-mentioned response information including first text response information and / or text information, the target information generation method proposed in this disclosure differs in its specific processing flow for first text candidate files and non-text candidate files (such as image files, audio files, or video files). Specifically, compared to generating target information using at least one of response information, first text files, and question information, the specific processing flow for generating target information using at least one of response information, non-text candidate files, and question information is more complex. That is, generating target information using at least one of response information, non-text candidate files, and question information may add a step of converting non-text candidate files into text candidate files. Therefore, similar to the above-mentioned processing method for response information including first text response information and / or text information, this disclosure proposes a method for generating target information for both first text candidate files and non-text candidate files.
[0088] Specifically, in this embodiment of the disclosure, target information is generated based on at least one of response information, candidate documents, and question information, including:
[0089] The purpose of information generation and / or a first text file is used to generate question information; response information in response to the question information is received, and target information is generated based on at least one of the response information, the first text file, and the question information; and / or,
[0090] Convert at least one of the image file, audio file, and video file to obtain a corresponding second text file; generate question information using the information generation target and / or the second text file; receive response information in response to the question information, and generate target information based on at least one of the response information, the second text file, and the question information.
[0091] For example, such as Figure 3B As shown, if the information generation instruction includes "generate exam questions related to the text file 'Advanced Mathematics'", then the candidate file for this information generation instruction can include the text file "Advanced Mathematics". In this case, question information can be generated based on the text file "Advanced Mathematics", for example, as... Figure 3B The system displays the "concept of an explanation vector" and receives the response information for that question (e.g., ...). Figure 3B The target information is generated based on at least one of the following: the quantity with size and direction (as shown in the response information, the text file "Advanced Mathematics"), and the question information.
[0092] Alternatively, if the information generation instruction includes "generating relevant exam questions for the video file 'The Fundamental Development of Materialist Dialectics'", then the candidate files for the information generation instruction can include the video file "The Fundamental Development of Materialist Dialectics". In this case, the video file "The Fundamental Development of Materialist Dialectics" can be converted into a corresponding second text file, and question information can be generated based on the second text file, such as "How to understand that grasping the principal contradiction solves all problems?", and the response information to the question information can be received, such as "In a group of contradictions, there are principal and secondary contradictions. The principal contradiction occupies a dominant position in the contradiction system and plays a decisive role in the development of things; the principal and secondary contradictions are interdependent and mutually influential, and can transform into each other under certain conditions." Based on the response information, the second text file corresponding to the video file "The Fundamental Development of Materialist Dialectics", and at least one of the question information, the target information can be generated.
[0093] This disclosure allows for the generation of target information by employing various types of candidate files. Therefore, this disclosure improves the accuracy of target information by expanding the range of candidate files involved in the target information generation method.
[0094] This disclosure allows for the generation of target information using different methods for different types of candidate files. In other words, the target information generation method proposed in this disclosure can process various types of candidate files in a targeted manner, thereby improving the applicability of the target information generation method and the accuracy of the target information.
[0095] It should be noted that, since the embodiments of this disclosure do not limit the specific content of the generated target information, the target information generation method proposed in the embodiments of this disclosure can be applied to multiple fields.
[0096] In one example, the method for generating target information proposed in this disclosure can be applied to the fields of intelligent question answering, intelligent testing, and test question determination.
[0097] Specifically, when the target information generation method proposed in this embodiment is applied to the fields of intelligent question answering, intelligent testing, and test question determination, the generation of target information based on at least one of response information, candidate files, and question information includes:
[0098] Determine the degree of match between the response information and the candidate documents;
[0099] Based on the matching degree and the first threshold, the first evaluation information of the response information is generated; the target information includes the first evaluation information.
[0100] The first threshold may include a pre-set matching degree between the response information and the candidate file. For example, the first threshold may include a matching degree of 90% between the response information and the candidate file. The response information includes first text response information and / or voice response information.
[0101] Specifically, such as Figure 3C As shown, if the response information includes first text response information and / or voice response information, the target information is generated based on at least one of the response information, candidate documents, and question information, including:
[0102] Step 1: Determine the response information, which includes the first text response and / or voice response;
[0103] Step 2: Analyze the first text response information and the candidate text to determine the matching degree between the first text response information and the candidate text; and / or, convert the voice response information to obtain the second text response information corresponding to the voice response information, analyze the second text response information and the candidate text to determine the matching degree between the second response information and the candidate text;
[0104] Step 3: Based on the matching degree and the first threshold, generate the first evaluation information for evaluating the response.
[0105] Accordingly, in some implementations, such as Figure 3D As shown, if the information generation instruction includes "Check the memorization results according to the video file 'The Fundamental Law of Materialist Dialectics'", that is, in this information generation instruction, the information generation purpose includes "checking the memorization results", the question information includes "how to understand that grasping the principal contradiction will solve all problems", and the response information includes the first text-voice response information "In a group of contradictions, there are principal contradictions and secondary contradictions. The principal contradiction occupies a dominant position in the contradiction system and plays a decisive role in the development of things; the principal contradiction and the secondary contradiction are interdependent and mutually influential, and can transform into each other under certain conditions", then the response information can be matched with the relevant content of "grasping the principal contradiction will solve all problems" in the video file "The Fundamental Law of Materialist Dialectics" to obtain the matching degree between the response information and the candidate file, and the first evaluation information of the response information is judged based on the matching degree.
[0106] The target information generation method proposed in this disclosure can generate target information (i.e., first evaluation information) based on the matching degree and a first threshold. This method of generating target information using the matching degree and the first threshold can reduce the interference of non-relevant information on the target information during generation, thereby effectively improving the accuracy of the target information.
[0107] Furthermore, when generating the first evaluation information of the response information based on the matching degree and the first threshold, the content contained in the target information (i.e., the first evaluation information) can be generated based on the relationship between the matching degree and the first threshold (such as the difference between the matching degree and the first threshold).
[0108] Specifically, when the matching degree is greater than or equal to the first threshold, the first evaluation information includes: the response information is correct.
[0109] Furthermore, when the matching degree is less than or equal to the first threshold, the first evaluation information includes at least one of the following:
[0110] The reply message is incorrect;
[0111] The differences between the response information and the candidate documents;
[0112] The location of the distinguishing content in the candidate file;
[0113] Distinguish the links in the candidate files.
[0114] For example, still Figure 3D As shown, with a first threshold of 90%, the information generation instruction includes "Check your memorization results based on the video file 'The Fundamental Law of Materialist Dialectics'"; the question information includes "How to understand that grasping the principal contradiction solves all problems"; and the response information includes "In a group of contradictions, there are principal and secondary contradictions. The principal contradiction occupies a dominant position in the contradiction system and plays a decisive role in the development of things; the principal and secondary contradictions are interdependent and mutually influential, and can transform into each other under certain conditions." For example, if the match between the response information and the relevant content of "Grasp the principal contradiction, and all problems will be solved" in the video file "The Fundamental Law of Materialist Dialectics" is less than or equal to 90%, then the first information includes an incorrect response; the response information differs from the candidate file (e.g., Figure 3D The statement "Secondary contradictions are subordinate in the system of contradictions and do not play a decisive role in the development of things" is shown; the position of the content in the candidate documents is distinguished (e.g., Figure 3D The “Video 00:35” shown); or links to distinguish content in the candidate files (such as...). Figure 3D (The link shown indicates that "secondary contradictions are subordinate in the system of contradictions and do not play a decisive role in the development of things"). Alternatively, if the response information matches the relevant content in the video file "The Fundamental Law of Materialist Dialectics" ("Grasp the principal contradiction, and all problems will be solved") with a degree of 90% or higher, then the response information is considered correct.
[0115] The target information generation method proposed in this disclosure can generate the specific content of the target information by comparing the matching degree with a first threshold. The method for generating the specific content of the target information proposed in this disclosure can determine the response information whose matching degree with the candidate file is within a certain range as the correct response information, thereby avoiding resource waste and time loss caused by excessively high requirements for the matching degree between the response information and the candidate file.
[0116] Furthermore, when the matching degree between the response information and the candidate document is less than or equal to the first threshold, the response information can be evaluated comprehensively and in multiple dimensions to obtain the first evaluation information (i.e., the target information). When the matching degree between the response information and the candidate document is less than or equal to the first threshold, relevant knowledge of the response information (such as extended information of the response information) can be quickly obtained based on the target information of the response information, thereby improving the efficiency of obtaining relevant knowledge of the response information.
[0117] In addition, in this embodiment of the disclosure, the location information of the target information in the candidate file may include at least one of the following:
[0118] The number of pages in the text file where the target information appears;
[0119] The number of lines in the text file in which the target information appears;
[0120] The time when the target information appears in the video file;
[0121] The time when the target information appears in the audio file.
[0122] It should be noted that the embodiments of this disclosure can also utilize the links to the target information in the candidate files for further processing, in order to quickly obtain the target information.
[0123] Specifically, embodiments of this disclosure also include:
[0124] Receive clicks on links;
[0125] Based on the click action, the candidate file will open and you will be redirected to the location corresponding to the link.
[0126] For example, if the information instruction includes "determine the relevant knowledge of power series in the cloud drive" and the target information includes a link to "page 12 of the text file 'Complex Functions and Integral Transforms'", then a click operation on the link to "page 12 of the text file 'Complex Functions and Integral Transforms'" is received. Based on this click operation, the candidate file, namely the text file "Complex Functions and Integral Transforms", is opened, and the user is redirected to the view text file "page 12 of the text file 'Complex Functions and Integral Transforms'".
[0127] By using links to jump to the target information, the accuracy of the target information can be improved while reducing the time and resources spent on acquiring related knowledge (such as extended content of the target information).
[0128] It should be noted that the method for generating first evaluation information (i.e., target information) of response information based on matching degree and a first threshold proposed in this disclosure is merely an example. For instance, the method for generating target information based on at least one of response information, candidate documents, and question information proposed in this disclosure may further include:
[0129] The pre-trained large language model generates the first evaluation information of the response information based on the question information and the response information;
[0130] The first evaluation information includes at least one of the following: the response is correct, the response is incorrect, the differences between the response and the candidate document, the location of the differences in the candidate document, and the link to the differences in the candidate document.
[0131] Among them, a pre-trained large language model can refer to an artificial intelligence model designed to understand and generate human language.
[0132] Pre-trained large language models have the characteristics of high generalization ability and high stability. Therefore, the method of generating target information using pre-trained large language models can improve the accuracy and reliability of target information.
[0133] It should be noted that when the target information generation method proposed in this disclosure is applied to the fields of intelligent question answering, intelligent testing, and test question determination, the target information generated and the specific process for generating the target information described above are merely examples. This disclosure does not limit the target information generated and the specific process for generating the target information when the target information generation method is applied to the fields of intelligent question answering, intelligent testing, and test question determination. For example, when the target information generation method proposed in this disclosure is applied to the fields of intelligent question answering, intelligent testing, and test question determination, generating target information based on at least one of the answer information, candidate documents, and question information may further include:
[0134] Based on the question information, the response information is evaluated to generate second evaluation information for the response information; the target information includes the second evaluation information.
[0135] The second evaluation information includes at least one of the following: grammatical accuracy of the response, content relevance of the response to the question, lexical accuracy of the response, and phonetic accuracy of the response.
[0136] For example, if the question information includes questions in English, i.e. Figure 3EThe question "Can you tell me about your favorite type of cuisine?" can be answered by providing a response in English. Figure 3E The response reads, "I like Chinese food. It is healthy and delicious." In this embodiment, the present disclosure can generate second evaluation information based on this response, such as... Figure 3E The message displayed includes phrases such as "Great job!"
[0137] Still Figure 3E As shown in this embodiment, the second evaluation information may include whether the English grammar of the response is correct, the relevance of the response to the question information, and the accuracy of the vocabulary of the response (e.g., ...). Figure 3E The text includes phrases like "When we are talking about a general type of food, we use 'foods' instead of 'food'" and the accuracy of the speech in the response.
[0138] It should be noted that the second evaluation information proposed in this embodiment may include text evaluation information and / or voice evaluation information. For example, this embodiment may display the text evaluation information for the response information on the terminal server; or, this embodiment may directly play the voice evaluation information for the response information; or, this embodiment may play the voice evaluation information for the response information after receiving an instruction to "play voice evaluation".
[0139] The target information generation method proposed in this disclosure can intelligently generate question information and evaluate the answer information based on the question information, thereby improving the convenience of knowledge point practice and language learning, and reducing the cost required for knowledge point practice and language learning.
[0140] It should be noted that the application field of the target information generation method proposed in this disclosure is merely illustrative, and this disclosure does not limit the specific application field of the target information generation method. For example, the target information generation method proposed in this disclosure can also be applied to, for example... Figure 4 The areas shown include target information generation, tool invocation, target information retrieval, event reminder (monitoring and care), learning plan generation, learning information generation, effect verification, information storage, and information reception.
[0141] For ease of understanding, the following embodiments of this disclosure will specifically explain how the target information generation method is applied to the above-mentioned fields:
[0142] I. Target Information Generation Domain.
[0143] Figure 5 This is a flowchart of the implementation of a method for generating target information according to an embodiment of the present disclosure, specifically, as shown in Figure 3. Figure 5 As shown, when this method for generating target information is applied to the field of target information generation, the method for generating target information includes:
[0144] S510. Based on the purpose of information generation, parse the candidate files to determine their characteristic information.
[0145] S520. Generate target text based on feature information and information generation purpose; target text includes text that is consistent with the features of candidate documents and satisfies the information generation purpose.
[0146] For example, if the information generation instruction includes "write an article on the theme of my father, following the writing style of the files in the cloud drive", then in this information generation instruction, the information generation purpose includes "write an article on the theme of my father", and the candidate files include "files in the cloud drive". Then, the files in the cloud drive can be parsed to determine the characteristic information of the files in the cloud drive, and the target information, namely an article on the theme of my father, can be generated according to the information generation purpose "write an article on the theme of my father".
[0147] The target information generation method proposed in this disclosure can intelligently generate target information associated with candidate files; and, when generating target information associated with candidate files, it can improve the accuracy of target information to a certain extent.
[0148] In one example, the feature information proposed in this disclosure embodiment includes at least one of the following:
[0149] The writing style of candidate documents;
[0150] The writing direction of candidate documents;
[0151] The font used to write candidate documents;
[0152] The format for writing candidate documents.
[0153] The feature information proposed in this embodiment can learn and understand candidate documents from multiple perspectives, thereby improving the accuracy of target information and the relevance between target information and candidate text.
[0154] II. Tool Invocation Area.
[0155] The method for generating target information proposed in this disclosure can also be applied to the field of tool invocation. Tool invocation can refer to converting at least one of audio, video, or image files into a text file.
[0156] Specifically, when the target information generation method proposed in the embodiments of this disclosure is applied to the field of tool invocation, the target information generation method includes:
[0157] Based on the purpose of information generation, the second text file corresponding to at least one of the image file, audio file, and video file is identified as the target information.
[0158] For example, if the information generation instruction includes "generate a text file corresponding to the video file 'The Fundamental Law of Materialist Dialectics'", then the information generation purpose of the information generation instruction can include "generating a text file corresponding to the video file". In this case, the video file "The Fundamental Law of Materialist Dialectics" can be converted into its corresponding second text file, and the text file can be identified as the target information.
[0159] By using embodiments of this disclosure to determine a second text file (i.e., target information) corresponding to at least one of an image file, audio file, and video file, the accuracy of the second text file (i.e., target information) can be improved.
[0160] III. Target Information Retrieval Field.
[0161] The method for generating target information proposed in this disclosure can also be applied to the field of target information retrieval.
[0162] For example, if the information generation instruction includes "generate relevant knowledge of power series in the text file 'Advanced Mathematics'", that is, in the information generation instruction, the information generation purpose includes "generating relevant knowledge of power series", the candidate file includes "text file 'Advanced Mathematics'", and the target information can include "relevant knowledge of power series" retrieved from the candidate file "text file 'Advanced Mathematics'".
[0163] IV. Event Reminder (Supervision and Care) Area.
[0164] The target information generation method proposed in this disclosure can also be applied to the field of event alerts (monitoring and care). Specifically, when the target information generation method is applied to the field of event alerts (monitoring and care), the target information generation method includes:
[0165] Based on the relevant information of the reminder event, determine whether the reminder event meets the triggering conditions;
[0166] When the triggering conditions of a reminder event are met, a reminder message for the reminder event is generated based on the reminder event, and the reminder message for the reminder event is sent.
[0167] The relevant information for the reminder event can be keywords and related events.
[0168] In some implementations, the reminder events proposed in this disclosure may include events pre-configured by the user terminal and / or the server. For example, server-pre-configured "late-to-bed reminders," "early-to-rise reminders," and "postgraduate entrance exam registration reminders," etc.
[0169] For example, if the reminder event includes the "postgraduate entrance examination registration reminder" pre-configured on the server, then the relevant information of the reminder event can include searching for postgraduate entrance examination time, the rate of collecting postgraduate entrance examination information, and searching for postgraduate entrance examination practice questions. In this case, if the relevant information meets the triggering conditions, then the reminder event "postgraduate entrance examination registration reminder" will be initiated during the postgraduate entrance examination registration time.
[0170] By using the embodiments of this disclosure to remind the event, the possibility of forgetting or delaying the reminder event can be reduced, thereby reducing the possibility of adverse effects caused by forgetting or delaying the reminder event.
[0171] In addition, in this embodiment of the disclosure, the triggering conditions for the reminder event may include:
[0172] The difference between the time of the reminder event and the current time is less than or equal to a second threshold; and / or,
[0173] The number of relevant information items for the reminder event is greater than or equal to the third threshold.
[0174] For example, if the reminder event includes a "late bedtime reminder" pre-configured by the server, and if the corresponding time of the pre-configured reminder event is 10 PM and the second threshold is one hour, then the reminder event meets the triggering conditions when the current time is 9 PM.
[0175] For example, if the reminder event includes the "postgraduate entrance examination registration reminder" pre-configured by the server, then the reminder event meets the triggering condition when the relevant information of the reminder event, such as the number of searches for postgraduate entrance examination time, the rate of collecting postgraduate entrance examination information, and the number of searches for postgraduate entrance examination practice questions, is greater than or equal to the third threshold; or, the reminder event meets the triggering condition when the keywords of the reminder event "postgraduate entrance examination registration reminder", such as the number of postgraduate entrance examination candidates and / or graduate students, are greater than or equal to the third threshold.
[0176] The embodiments of this disclosure can utilize information about any reminder event from multiple aspects to determine whether the reminder event meets the trigger information, and send reminder information for the reminder event if the reminder event meets the trigger information, thereby reducing the possibility of incorrectly determining whether the reminder event meets the trigger information, as well as the possibility of delaying or forgetting the reminder event.
[0177] Furthermore, in this embodiment of the disclosure, the reminder information includes at least one of the following:
[0178] The content of the reminder event;
[0179] Remind the time of the event;
[0180] The type of reminder event.
[0181] For example, if the reminder event "Postgraduate Entrance Examination Registration Reminder" meets the triggering conditions, then targeted information (i.e., reminder information) can be sent to the "Postgraduate Entrance Examination Registration Reminder", such as the content of the "Postgraduate Entrance Examination Registration Reminder", the time when the "Postgraduate Entrance Examination Registration Reminder" occurs, or the type of the "Postgraduate Entrance Examination Registration Reminder".
[0182] The reminder information proposed in this embodiment can determine the reminder information corresponding to the reminder event based on information from any one or more angles, thereby improving the comprehensiveness of the reminder information obtained for the reminder event.
[0183] V. Learning plan generation, learning information generation, and effectiveness evaluation.
[0184] The method for generating target information proposed in this disclosure can also be applied to the fields of learning plan generation, learning information generation, and effect verification.
[0185] Specifically, when the target information generation method proposed in the embodiments of this disclosure is applied to the fields of learning plan generation, learning information generation, and effect verification, the target information generation method includes:
[0186] Receive information generation instructions, which indicate the purpose of information generation and / or candidate files;
[0187] Identify information generation instructions to determine the purpose of information generation and / or candidate files;
[0188] Based on the information generation purpose and / or candidate documents, generate target information; wherein, target information includes learning plan generation and / or learning information.
[0189] For example, if the information generation instruction includes "to develop a photography learning plan based on the knowledge related to photography in the files in the cloud drive", that is, in the information generation instruction, the purpose of information generation includes "knowledge related to photography" and "developing a plan", and the candidate files include "files in the cloud drive", then the target information can include a photography learning plan and / or any learning information related to the purpose of information generation in the cloud drive files.
[0190] Furthermore, the embodiments of this disclosure can also be used to verify the above-mentioned learning effect by applying the method for generating the target information to the fields of intelligent question answering and / or intelligent testing.
[0191] VI. Information storage and information reception areas.
[0192] The method for generating target information proposed in this disclosure can also be applied to the fields of information storage and information reception. Specifically, when the method for generating target information proposed in this disclosure is applied to the fields of information storage and information reception, the method for generating target information may further include:
[0193] Receive information saving instructions, which specify the purpose of saving the information and the information to be saved.
[0194] Identify information saving instructions to generate the information saving purpose and the information to be saved;
[0195] Save the information to be saved to the cloud and / or the user terminal.
[0196] The embodiments of this disclosure can achieve the purpose of quickly obtaining historical target information and related information of the target information by saving target information and / or related information of the target information.
[0197] Furthermore, when the target information generation method proposed in this disclosure is applied to the fields of information storage and information reception, it can also be combined with the target information generation method in the fields of intelligent question answering and intelligent testing. That is, the target information generated in the process of intelligent question answering and intelligent testing can be saved to relevant files.
[0198] Specifically, in this embodiment of the disclosure, at least one of the following may be saved to the relevant file:
[0199] The differences between the response information and the candidate documents;
[0200] The location of the distinguishing content in the candidate file;
[0201] Distinguish the links in the candidate files.
[0202] For example, embodiments of this disclosure may include the target information, such as the differences between the response information and the candidate document; the location of the differences in the candidate document; or the link of the differences in the candidate document (i.e., the link between the candidate document and the response information). Figure 6 Save the links corresponding to "Secondary contradictions are subordinate in the system of contradictions and do not play a decisive role in the development of things" and "Secondary contradictions are subordinate in the system of contradictions and do not play a decisive role in the development of things". You can use... Figure 6 The above information can be saved by adding it to a knowledge card as shown, or by receiving voice and / or text commands.
[0203] Saving multiple historical target information generated using the embodiments of this disclosure to relevant files or knowledge cards can facilitate the subsequent retrieval and review of historical target information; and avoid the waste of resources caused by the secondary generation of the historical target information.
[0204] Of course, the embodiments of this disclosure can also save the newly received information, thereby enabling rapid acquisition and retrieval of the newly received information.
[0205] This disclosure also proposes an apparatus for generating target information. Figure 7 This is a schematic diagram of the structure of a target information generation apparatus 700 according to an embodiment of the present disclosure, including:
[0206] The receiving module 710 is used to receive information generation instructions, which are used to indicate the purpose of information generation and / or candidate files;
[0207] The identification module 720 is used to identify information generation instructions in order to determine the purpose of information generation and / or candidate files;
[0208] The first generation module 730 is configured to generate question information for an information generation instruction based on the information generation purpose and / or candidate documents; receive response information in response to the question information; and generate target information based on at least one of the response information, candidate documents, and question information.
[0209] In some implementations, the first generation module 730 is used for:
[0210] Determine the degree of match between the response information and the candidate document;
[0211] Based on the matching degree and a first threshold, first evaluation information of the response information is generated; the target information includes the first evaluation information.
[0212] In some implementations, when the matching degree is greater than or equal to a first threshold, the first evaluation information includes: the response information is correct.
[0213] In some implementations, when the matching degree is less than or equal to a first threshold, the first evaluation information includes at least one of the following:
[0214] The reply is incorrect.
[0215] The differences between this response and the candidate document;
[0216] The location of this distinguishing content in the candidate file;
[0217] The distinction is indicated by a link in the candidate file.
[0218] In some implementations, the first generation module 730 is further configured to:
[0219] The pre-trained large language model generates the first evaluation information of the response information based on the question information and the response information;
[0220] The first evaluation information includes at least one of the following: the response is correct, the response is incorrect, the difference between the response and the candidate document, the location of the difference in the candidate document, and the link of the difference in the candidate document.
[0221] Figure 8 This is a schematic diagram of the structure of a target information generation apparatus 800 according to an embodiment of the present disclosure, as shown below. Figure 8 As shown, in some embodiments, a jump module 840 is also included, for:
[0222] Receive clicks on links;
[0223] Based on the click action, the candidate file will open and you will be redirected to the location corresponding to the link.
[0224] In some implementations, a first storage module 850 is also included for saving at least one of the following to a relevant file:
[0225] The differences between this response and the candidate document
[0226] The location of this distinguishing content in the candidate file;
[0227] The distinction is indicated by a link in the candidate file.
[0228] In some implementations, the first generation module 730 is used for:
[0229] Based on the question information, the response information is evaluated to generate second evaluation information for the response information; the target information includes the second evaluation information.
[0230] The second evaluation information includes at least one of the following: grammatical accuracy of the response, content relevance of the response to the question, lexical accuracy of the response, and phonetic accuracy of the response.
[0231] In some implementations, the response information includes a first text response and / or a voice response.
[0232] In some implementations, the first generation module 730 is further configured to:
[0233] The target information is generated based on at least one of the first text response information, candidate documents, and question information; and / or,
[0234] The voice response information is converted to obtain the corresponding second text response information; and the target information is generated based on at least one of the second text response information, candidate files, and question information.
[0235] In some implementations, a second generation module 860 is also included, for:
[0236] Based on the purpose of information generation, the candidate file is parsed to determine its characteristic information.
[0237] Based on the feature information and the purpose of information generation, target text is generated; the target text includes text that is consistent with the features of the candidate file and satisfies the purpose of information generation.
[0238] In some implementations, the feature information includes at least one of the following:
[0239] The writing style of the candidate document;
[0240] The writing direction of the candidate file;
[0241] The font used to write the candidate file;
[0242] The format for writing candidate documents.
[0243] In some implementations, the candidate file includes at least one of a first text file, an image file, an audio file, and a video file.
[0244] In some implementations, the first generation module 730 is used for:
[0245] Generate question information based on the information generation purpose and / or a first text file; receive response information in response to the question information, and generate target information based on at least one of the response information, the first text file, and the question information; and / or,
[0246] Convert at least one of the image file, audio file, and video file to obtain a corresponding second text file; generate question information using the information generation target and / or the second text file; receive response information in response to the question information, and generate target information based on at least one of the response information, the second text file, and the question information.
[0247] In some implementations, when the candidate files include image files, audio files, and video files, a determining module 870 is also included, for:
[0248] Based on the purpose of information generation, the second text file corresponding to at least one of the image file, audio file, and video file is identified as the target information.
[0249] In some implementations, a third generation module 880 is also included, for:
[0250] Based on the relevant information of the reminder event, determine whether the reminder event meets the triggering conditions;
[0251] When the triggering conditions of a reminder event are met, a reminder message for the reminder event is generated based on the reminder event, and the reminder message for the reminder event is sent.
[0252] In some implementations, the triggering condition includes:
[0253] The difference between the time of the notification event and the current time is less than or equal to the second threshold; and / or,
[0254] The number of relevant information items for the alert event is greater than or equal to the third threshold.
[0255] In some implementations, the alert message includes at least one of the following:
[0256] The content of the notification event;
[0257] The reminder should indicate the time of the event.
[0258] The type of notification event.
[0259] In some embodiments, a second storage module 890 is also included, for:
[0260] Receive information saving instructions, which specify the purpose of saving the information and the information to be saved.
[0261] Identify information saving instructions to determine the purpose of information saving and the information to be saved;
[0262] Save the information to be saved to the cloud and / or the user terminal.
[0263] The specific functions and examples of each module and submodule of the apparatus in this disclosure can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.
[0264] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.
[0265] Figure 9 A schematic block diagram of an example electronic device 900 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.
[0266] like Figure 9 As shown, device 900 includes a computing unit 901, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 902 or a computer program loaded from storage unit 908 into random access memory (RAM) 903. RAM 903 may also store various programs and data required for the operation of device 900. The computing unit 901, ROM 902, and RAM 903 are interconnected via bus 904. Input / output (I / O) interface 905 is also connected to bus 904.
[0267] Multiple components in device 900 are connected to I / O interface 905, including: input unit 906, such as keyboard, mouse, etc.; output unit 907, such as various types of monitors, speakers, etc.; storage unit 908, such as disk, optical disk, etc.; and communication unit 909, such as network card, modem, wireless transceiver, etc. Communication unit 909 allows device 900 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0268] The computing unit 901 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 901 performs the various methods and processes described above, such as the method for generating target information. For example, in some embodiments, the method for generating target information may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on device 900 via ROM 902 and / or communication unit 909. When the computer program is loaded into RAM 903 and executed by the computing unit 901, one or more steps of the method for generating target information described above may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform the method for generating target information by any other suitable means (e.g., by means of firmware).
[0269] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0270] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0271] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0272] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0273] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.
[0274] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.
[0275] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.
[0276] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.
Claims
1. A method for generating target information, comprising: Receive information generation instructions, which are used to indicate the purpose of information generation and / or candidate files; The candidate file is stored in a device and / or server with storage capabilities; the information generation instruction includes at least one of the following: the storage address of the candidate file, the storage address link, identification information, and the name; Identify the information generation instructions to determine the purpose of information generation and / or candidate files; Based on the information generation purpose and / or the candidate file, generate the problem information for the information generation instruction; Receive response information in response to the question information, and generate target information based on at least one of the response information, the candidate file, and the question information; The step of generating target information based on at least one of the response information, the candidate file, and the question information includes: determining the matching degree between the response information and the candidate file; generating first evaluation information of the response information based on the matching degree and a first threshold; the target information includes the first evaluation information; wherein, when the matching degree is less than or equal to the first threshold, the first evaluation information includes: the response information is incorrect, the content that distinguishes the response information from the candidate file, the position of the content that distinguishes the response information from the candidate file, and the link of the content that distinguishes the response information from the candidate file; Receive click operations on the link; Based on the click operation, the candidate file is opened, and the user is redirected to the location corresponding to the link. Save the differences between the response information and the candidate file, the location of the differences in the candidate file, and the link of the differences in the candidate file to the relevant file; The method further includes: Based on the purpose of information generation, the candidate files are parsed to determine the feature information of the candidate files; Based on the feature information and the information generation purpose, target text is generated; the target text includes text that is consistent with the features of the candidate files and satisfies the information generation purpose; the information generation purpose includes the topic of the target text, the candidate files include files in the cloud drive, and the feature information includes the writing style of the candidate files.
2. The method according to claim 1, wherein, The step of generating target information based on at least one of the response information, the candidate file, and the question information further includes: A pre-trained large language model generates first evaluation information of the response information based on the question information and the response information; The first evaluation information includes at least one of the following: the response information is correct, the response information is incorrect, the difference between the response information and the candidate file, the position of the difference in the candidate file, and the link of the difference in the candidate file.
3. The method according to claim 1 or 2, wherein, The step of generating target information based on at least one of the response information, the candidate file, and the question information includes: Based on the question information, the response information is evaluated to generate second evaluation information for the response information; the target information includes the second evaluation information. The second evaluation information includes at least one of the following: grammatical accuracy of the response information, content relevance of the response information to the question information, lexical accuracy of the response information, and phonetic accuracy of the response information.
4. The method according to claim 1 or 2, wherein, The response information includes a first text response and / or a voice response.
5. The method according to claim 4, wherein, The step of generating target information based on at least one of the response information, the candidate file, and the question information includes: The target information is generated based on at least one of the first text response information, the candidate file, and the question information; and / or, The voice response information is converted to obtain the corresponding second text response information; and the target information is generated based on at least one of the second text response information, the candidate file, and the question information.
6. The method according to claim 1, wherein, The feature information also includes at least one of the following: The writing direction of the candidate files; The font used to write the candidate files; The writing format of the candidate documents.
7. The method according to claim 1 or 2, wherein, The candidate files include at least one of a first text file, an image file, an audio file, and a video file.
8. The method according to claim 7, wherein, The problem information for generating the information generation instruction is generated based on the information generation purpose and / or the candidate file; Receiving a response to the question information, and generating target information based on at least one of the response information, the candidate file, and the question information, including: Generate the question information based on the stated information and / or the first text file; receive response information in response to the question information, and generate the target information based on at least one of the response information, the first text file, and the question information; and / or, Convert at least one of the image file, audio file, and video file to obtain a corresponding second text file; generate the question information using the information to generate a purpose and / or the second text file; receive a response to the question information, and generate the target information based on at least one of the response information, the second text file, and the question information.
9. The method according to claim 7, wherein, If the candidate files include the image files, audio files, and video files, then the following are also included: Based on the stated purpose of information generation, a second text file corresponding to at least one of the image file, audio file, and video file is determined as the target information.
10. The method according to claim 1 or 2, further comprising: Based on the relevant information of the reminder event, determine whether the reminder event meets the triggering conditions; If the reminder event meets the triggering condition, a reminder message for the reminder event is generated based on the reminder event, and the reminder message for the reminder event is sent.
11. The method according to claim 10, wherein, The triggering conditions include: The difference between the time corresponding to the reminder event and the current time is less than or equal to a second threshold; and / or, The number of relevant information items for the reminder event is greater than or equal to the third threshold.
12. The method according to claim 10, wherein, The reminder information includes at least one of the following: The content of the reminder event; The time of occurrence of the reminder event; The type of the reminder event.
13. The method according to claim 1 or 2, further comprising: Receive an information saving instruction, the information saving instruction being used to indicate the purpose of information saving and the information to be saved; The information saving instruction is identified to determine the purpose of information saving and the information to be saved. The information to be saved is saved to the cloud and / or the user terminal.
14. An apparatus for generating target information, comprising: A receiving module is configured to receive an information generation instruction, wherein the information generation instruction indicates the purpose of information generation and / or candidate files; The candidate file is stored in a device and / or server with storage capabilities; the information generation instruction includes at least one of the following: the storage address of the candidate file, the storage address link, identification information, and the name; The identification module is used to identify the information generation instruction in order to determine the purpose of information generation and / or candidate files; The first generation module is used to generate the problem information of the information generation instruction based on the information generation purpose and / or the candidate file; Receive response information in response to the question information, and generate target information based on at least one of the response information, the candidate file, and the question information; determine the matching degree between the response information and the candidate file; and generate first evaluation information of the response information based on the matching degree and a first threshold. The target information includes the first evaluation information; wherein, when the matching degree is less than or equal to the first threshold, the first evaluation information includes: the response information is incorrect, the difference between the response information and the candidate file, the position of the difference in the candidate file, and the link of the difference in the candidate file; The jump module is used to: receive a click operation on the link; open the candidate file according to the click operation, and jump to the position corresponding to the link; The first saving module is used to save the differences between the response information and the candidate file, the position of the differences in the candidate file, and the link of the differences in the candidate file to a relevant file; The device further includes: The second generation module is used to parse the candidate file according to the information generation purpose to determine the feature information of the candidate file; Based on the feature information and the information generation purpose, target text is generated; the target text includes text that is consistent with the features of the candidate files and satisfies the information generation purpose; the information generation purpose includes the topic of the target text, the candidate files include files in the cloud drive, and the feature information includes the writing style of the candidate files.
15. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-13.
16. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-13.
17. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-13.
Citation Information
Patent Citations
Event reminding method and mobile terminal
CN107483747A
Speech training method and device based on language model and computer equipment
CN110309279A
Conversational query answering system
CN110647618A
Question-answer scoring method, device and equipment based on artificial intelligence, and storage medium
CN110797010A
Jump link editing method, jump link generating method, jump link editing device, device and system in intelligent teaching materials
CN111881383A
