Information acquisition method and related device

Abstract

The invention discloses an information acquisition method and a related device. The method comprises the steps of obtaining a target problem through to-be-obtained information in a structured format and an executable task set and problem combination in a structured format; the target problem is input into a preset generation model for task generation corresponding to the to-be-obtained information, a to-be-executed task set in a structured format is output, and the to-be-executed task set comprises to-be-executed tasks used for obtaining the to-be-obtained information in an executable task set; performing task execution according to the to-be-executed tasks in the to-be-executed task set in the structured format to obtain a task execution result in the structured format; and performing language generation on the task execution result in the structured format to obtain an information acquisition result of the to-be-acquired information. According to the method, the whole network unpublished information matched with the to-be-acquired information can be acquired, and different and variable information acquisition results can be generated through customization of the same to-be-acquired information, so that the information acquisition results are high in accuracy and high in customization.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to an information acquisition method and related devices. Background Art

[0002] With the rapid development of computer technology, the dissemination and sharing of information have become more and more common. Based on the information acquisition requirement, a user inputs the information to be acquired through a user interface, and uses an information acquisition method to obtain an information acquisition result of the information to be acquired.

[0003] In related technologies, the information acquisition method usually searches for publicly available information across the network through the information to be acquired, and uses the publicly available information across the network that matches the information to be acquired as the information acquisition result of the information to be acquired.

[0004] However, the above method can neither acquire information that is not publicly available across the network, nor can the publicly available information across the network obtained through the same information to be acquired be different and variable, resulting in low accuracy and weak customization of the information acquisition result, thus leading to poor information acquisition effect. Summary of the Invention

[0005] To solve the above technical problems, this application provides an information acquisition method and related devices. By using a preset generation model to answer a target question formed by combining the information to be acquired and a set of executable tasks, the executable tasks for acquiring the information to be acquired can be called as needed to form a set of tasks to be executed, the tasks to be executed in the set of tasks to be executed are executed, and the information acquisition result is output in natural language. It can not only acquire information that is not publicly available across the network and matches the information to be acquired, but also customize and generate different and variable information acquisition results for the same information to be acquired, making the information acquisition result highly accurate and strongly customizable, thus improving the information acquisition effect.

[0006] Embodiments of this application disclose the following technical solutions:

[0007] On the one hand, an embodiment of this application provides an information acquisition method, and the method includes:

[0008] Combining the information to be acquired in a structured format and the set of executable tasks in the structured format to obtain a target question;

[0009] Using a preset generation model to generate tasks corresponding to the information to be acquired for the target question, to obtain the set of tasks to be executed in the structured format; the set of tasks to be executed includes the tasks to be executed in the set of executable tasks for acquiring the information to be acquired;

[0010] Executing the tasks to be executed in the set of tasks to be executed in the structured format to obtain the task execution result in the structured format;

[0011] Perform language generation on the task execution result in the structured format to obtain the information acquisition result of the information to be acquired.

[0012] On the other hand, an embodiment of the present application provides an information acquisition device, which includes: a question combination unit, a task generation unit, a task execution unit, and a language generation unit;

[0013] The question combination unit is used to combine questions for the information to be acquired in the structured format and the set of executable tasks in the structured format to obtain a target question;

[0014] The task generation unit is used to generate tasks corresponding to the information to be acquired for the target question through a preset generation model to obtain a set of tasks to be executed in the structured format; the set of tasks to be executed includes the tasks to be executed in the set of executable tasks for acquiring the information to be acquired;

[0015] The task execution unit is used to execute the tasks to be executed in the set of tasks to be executed in the structured format to obtain the task execution result in the structured format;

[0016] The language generation unit is used to perform language generation on the task execution result in the structured format to obtain the information acquisition result of the information to be acquired.

[0017] On the other hand, an embodiment of the present application provides a computer device, which includes a processor and a memory:

[0018] The memory is used to store a computer program and transmit the computer program to the processor;

[0019] The processor is used to execute the method described in any of the foregoing aspects according to the instructions in the computer program.

[0020] On the other hand, an embodiment of the present application provides a computer-readable storage medium, which is used to store a computer program. When the computer program runs on a computer device, the computer device is enabled to execute the method described in any of the foregoing aspects.

[0021] On the other hand, an embodiment of the present application provides a computer program product, which includes a computer program. When the computer program runs on a computer device, the computer device is enabled to execute the method described in any of the foregoing aspects.

[0022] It can be seen from the above technical scheme that, firstly, the target question is obtained by combining the information to be obtained in a structured format and the set of executable tasks in a structured format; this step combines the target question to be answered by the preset generation model through the information to be obtained that represents the information acquisition requirements and the executable tasks that can be called by the preset generation model. Secondly, the target question is input into the preset generation model to generate the tasks corresponding to the information to be obtained, and the set of tasks to be executed in a structured format is output, wherein the set of tasks to be executed includes the tasks to be executed in the executable task set for obtaining the information to be obtained; this step uses the executable tasks required to be called to obtain the information to be obtained as the tasks to be executed through the preset generation model to answer the target question. Then, the tasks to be executed in the set of tasks to be executed in the structured format are executed to obtain the task execution results in a structured format; this step obtains the task execution results in a structured format representing the information acquisition results by executing the tasks to be executed. Finally, the task execution results in the structured format are language generated to obtain the information acquisition results of the information to be obtained; this step converts the task execution results into information acquisition results in natural language that is easy to understand.

[0023] Based on this, the method answers the target question composed of the information to be obtained and the set of executable tasks through a preset generation model. The tasks to be executed for obtaining the information to be obtained can be called on demand to form a set of tasks to be executed, the tasks to be executed in the set of tasks to be executed can be executed, and the information acquisition results can be output in natural language. Not only can the undisclosed information on the entire network that matches the information to be obtained be obtained, but the same information to be obtained can be customized to generate different and changing information acquisition results, so that the information acquisition results are highly accurate and highly customized, thereby improving the information acquisition effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technical members in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0025] Figure 1 A system schematic diagram of an information acquisition method provided in an embodiment of the present application;

[0026] Figure 2 A flowchart of an information acquisition method provided in an embodiment of the present application;

[0027] Figure 3 A schematic diagram of the steps of an information acquisition method provided in an embodiment of the present application;

[0028] Figure 4Schematic diagram of specific steps for obtaining a task execution result in a structured format provided by an embodiment of the present application;

[0029] Figure 5 Schematic diagram of steps for obtaining a result of display information of a task execution result based on a structured format provided by an embodiment of the present application;

[0030] Figure 6 System framework diagram of an information acquisition method provided by an embodiment of the present application;

[0031] Figure 7 Specific flowchart of an information acquisition method provided by an embodiment of the present application;

[0032] Figure 8 Structural diagram of an information acquisition device provided by an embodiment of the present application;

[0033] Figure 9 Structural diagram of a server provided by an embodiment of the present application;

[0034] Figure 10 Structural diagram of a terminal provided by an embodiment of the present application. Detailed implementation manners

[0035] The embodiments of the present application will be described below with reference to the accompanying drawings.

[0036] At present, the information acquisition method usually searches the publicly available information on the whole network for the information to be acquired, and uses the publicly available information on the whole network that matches the information to be acquired as the information acquisition result of the information to be acquired.

[0037] However, through research, it is found that the above method can not only fail to acquire the information that is not publicly available on the whole network. For example, it cannot acquire customized interface information, internal account information, and non-Hypertext Transfer Protocol (HTTP) information, etc.; moreover, the publicly available information on the whole network obtained by the same information to be acquired is the same and fixed, resulting in low accuracy and weak customization of the information acquisition result, thus leading to poor information acquisition effect.

[0038] The embodiment of the present application provides an information acquisition method. By using a preset generation model to answer the target question combined by the information to be acquired and the set of executable tasks, the executable tasks for acquiring the information to be acquired can be called as needed to form a set of executable tasks, execute the executable tasks in the set of executable tasks, and output the information acquisition result in natural language; not only can it acquire the information that is not publicly available on the whole network and matches the information to be acquired, but also the same information to be acquired can be customized to generate different and variable information acquisition results, making the information acquisition result have high accuracy and strong customization, thus improving the information acquisition effect.

[0039] Next, the system architecture of the information acquisition method will be introduced. Refer to Figure 1 , Figure 1 FIG. is a schematic diagram of a system for an information acquisition method provided by an embodiment of the present application. The system includes a computer device 100, and the computer device 100 is used to execute the information acquisition method.

[0040] The computer device 100 combines problems of the information to be acquired in a structured format and a set of executable tasks in a structured format to obtain a target problem.

[0041] As an example, the structured format is the JSON format, which is a lightweight data exchange format. The information to be acquired is goal, and the set of executable tasks is tasks. Then, the computer device 100 combines problems through the JSON-format goal and the JSON-format tasks to obtain the target problem as {goal, tasks}.

[0042] The computer device 100 generates tasks corresponding to the information to be acquired for the target problem through a preset generation model to obtain a set of tasks to be executed in a structured format. The set of tasks to be executed includes the tasks to be executed in the set of executable tasks for acquiring the information to be acquired.

[0043] As an example, the preset generation model is a Generative Pre-Trained Transformer (GPT). Based on the above example, the computer device 100 inputs {goal, tasks} into the GPT to generate tasks corresponding to goal, and outputs the set of tasks to be executed in the JSON format as the JSON-format tasks'. Among them, the JSON-format tasks' includes the task to be executed task' in the JSON-format tasks for acquiring goal.

[0044] The computer device 100 executes the tasks to be executed in the set of tasks to be executed in a structured format to obtain a task execution result in a structured format.

[0045] As an example, based on the above example, the computer device 100 executes tasks according to the task to be executed task' in the JSON-format tasks' to obtain the task execution result in the JSON format as the JSON-format results.

[0046] The computer device 100 generates a language for the task execution result in a structured format to obtain an information acquisition result of the information to be acquired.

[0047] As an example, based on the above example, the computer device 100 generates language for results in JSON format to obtain natural language information acquisition results of the goal.

[0048] That is to say, the target question to be answered by the preset generation model is combined through the information to be obtained that represents the information acquisition needs and the executable tasks that can be called by the preset generation model; the executable tasks required to obtain the information to be obtained are used as the tasks to be executed through the preset generation model to answer the target question; the task execution results in a structured format representing the information acquisition results are obtained by executing the tasks to be executed; the task execution results are converted into information acquisition results in natural language that are easy to understand. Based on this, the method answers the target question composed of the information to be obtained and the executable task set through the preset generation model, and can call the tasks to be executed to obtain the information to be obtained on demand to form a set of tasks to be executed, execute the tasks to be executed in the set of tasks to be executed, and output the information acquisition results in natural language; not only can the undisclosed information on the entire network that matches the information to be obtained be obtained, but also the same information to be obtained can be customized to generate different and changing information acquisition results, so that the information acquisition results are highly accurate and customized, thereby improving the information acquisition effect.

[0049] It should be noted that the information acquisition method in the embodiment of the present application involves artificial intelligence. Artificial intelligence is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that the machines have the functions of perception, reasoning and decision-making.

[0050] Artificial intelligence technology is a comprehensive discipline that covers a wide range of fields, including both hardware-level and software-level technologies. Basic artificial intelligence technologies generally include sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, pre-trained model technology, operation / interaction systems, mechatronics, etc. Among them, the pre-trained model is also called a large model or a basic model. After fine-tuning, it can be widely used in downstream tasks in various major directions of artificial intelligence. In the embodiments of this application, artificial intelligence technology mainly involves natural language processing technology and machine learning / deep learning technology.

[0051] Natural Language Processing (NLP) is an important direction in the fields of computer science and artificial intelligence. It studies various theories and methods that can achieve effective communication between humans and computers in natural language. Natural language processing involves natural language, that is, the language people use in daily life, and is closely related to linguistic research; at the same time, it involves computer science and mathematics. The pre-trained model, an important technology for model training in the field of artificial intelligence, evolved from the large language model in the field of NLP. After fine-tuning, the large language model can be widely applied to downstream tasks. Natural language processing technologies usually include text processing, semantic understanding, machine translation, robot question answering, knowledge graphs and other technologies. Machine learning / deep learning is an interdisciplinary subject that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers simulate or implement human learning behaviors to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve their own performance. Machine learning / deep learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications cover all fields of artificial intelligence. Machine learning / deep learning technologies usually include artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, teaching learning and other technologies. The pre-trained model is the latest development result of deep learning, integrating the above technologies.

[0052] It should be noted that in the embodiments of the present application, the computer device can be a server or a terminal. The method provided in the embodiments of the present application can be executed independently by the terminal or the server, or can be executed in cooperation by the terminal and the server. Among them, when the method provided in the embodiments of the present application is executed independently by the terminal or the server, its execution method is similar to that of Figure 1 the corresponding embodiment, mainly replacing the computer device with the terminal or the server. In addition, when the method provided in the embodiments of the present application is executed in cooperation by the terminal and the server, the steps that need to be reflected on the front-end interface can be executed by the terminal, while some steps that require background calculation and do not need to be reflected on the front-end interface can be executed by the server.

[0053] Among them, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart voice interaction device, a vehicle-mounted terminal, an extended reality device or an aircraft, etc., but is not limited thereto. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services, but is not limited thereto. The terminal and the server can be directly or indirectly connected through wired or wireless communication methods, and the present application does not make any restrictions here. For example, the terminal and the server can be connected through a network, and the network can be a wired or wireless network.

[0054] In addition, the embodiments of the present application can be applied to various scenarios, including but not limited to cloud technology, artificial intelligence, intelligent transportation, assisted driving, autonomous driving, digital humans, virtual humans, virtual reality, augmented reality, mixed reality, audio and video, etc.

[0055] Next, taking a computer device executing the method provided by the embodiments of the present application as an example, the information acquisition method provided by the embodiments of the present application will be introduced in detail in combination with the accompanying drawings. Refer to Figure 2 , Figure 2 which is a flowchart of an information acquisition method provided by the embodiments of the present application. The method includes:

[0056] S201: Combine the information to be acquired in a structured format and the set of executable tasks in a structured format to obtain a target problem.

[0057] In the related art, the information acquisition method usually searches the publicly available information on the entire network through the information to be acquired, and uses the publicly available information on the entire network that matches the information to be acquired as the information acquisition result of the information to be acquired. However, the above method can not only not acquire the information that is not publicly available on the entire network, such as customized interface information, internal account information, and non-HTTP information; but also the publicly available information obtained through the same information to be acquired is the same and fixed, resulting in low accuracy and weak customization of the information acquisition result, thus resulting in poor information acquisition effect.

[0058] Therefore, in the embodiments of the present application, in order to solve the above problems, considering that the preset generation model can generate a set of tasks to be executed for acquiring the information to be acquired in the set of executable tasks based on the information to be acquired representing the information acquisition requirement and the set of executable tasks called by the preset generation model, and execute the tasks to be executed in the set of tasks to be executed to meet the information acquisition requirement, it can not only acquire the information that is not publicly available on the entire network that matches the information to be acquired, but also the same information to be acquired can be customized to generate different and variable information acquisition results.

[0059] Based on this, first, considering that it is necessary to interact with the preset generation model using data in a structured format for convenient data processing, the target problem including the information to be acquired in a structured format and the set of executable tasks in a structured format is obtained by combining problems through the information to be acquired in a structured format and the set of executable tasks in a structured format. Among them, the information to be acquired refers to the information representing the information acquisition requirement, and the information to be acquired includes keywords or prompt words representing the information acquisition requirement, etc.; the set of executable tasks refers to the set of executable tasks that can be executed locally and called by the preset generation model, and the set of executable tasks includes tasks for acquiring customized interface information, tasks for acquiring internal account information, and tasks for acquiring non-HTTP information, etc.; the structured format refers to the field format reflecting the data structure, and the structured format includes JSON format, etc.

[0060] The S201 combines the information to be obtained representing the information acquisition requirement and the executable tasks that can be called by a preset generation model to form a target question to be answered by the preset generation model; it provides a data basis for subsequently using the executable tasks required to obtain the information to be obtained as tasks to be executed to form a set of tasks to be executed, and executing the tasks to be executed in the set of tasks to be executed to meet the information acquisition requirement, so as to be able to obtain non-public information across the network that matches the information to be obtained.

[0061] As an example of the S201, the structured format is JSON format, the information to be obtained is goal, and goal is specifically "obtain X news", and the set of executable tasks is tasks, and tasks includes task0, task1,...; then the computer device combines the {goal: obtain X news} and {tasks: task0, task1,...} to get the target question as {goal: obtain X news, tasks: task0, task1,...}.

[0062] S202: Generate a set of tasks to be executed in a structured format corresponding to the information to be obtained for the target question through a preset generation model; the set of tasks to be executed includes the tasks to be executed in the set of executable tasks for obtaining the information to be obtained.

[0063] In the embodiment of the present application, after executing S201 to combine the target question including the information to be obtained in a structured format and the set of executable tasks in a structured format, since the target question is to be answered by the preset generation model, in order to meet the information acquisition requirement, it is necessary to input the target question into the preset generation model to generate the tasks corresponding to the information to be obtained, and output a structured format including the tasks to be executed in the set of executable tasks for obtaining the information to be obtained to form a set of tasks to be executed. Among them, the preset generation model is used to generate the tasks to be executed in the set of executable tasks for obtaining the information to be obtained based on the information to be obtained and the set of executable tasks to form a set of tasks to be executed.

[0064] The S202 uses the executable tasks required to obtain the information to be obtained as tasks to be executed through the preset generation model to form a set of tasks to be executed to answer the target question; it provides an execution basis for subsequently executing the tasks to be executed in the set of tasks to be executed to meet the information acquisition requirement, so as to not only be able to obtain non-public information across the network that matches the information to be obtained, but also the same information to be obtained can be customized to generate different and variable information acquisition results.

[0065] As an example of S202, based on the above S201 example, the preset generation model is GPT, and the computer device inputs {goal: get X news, tasks: task0, task1, ...} into GPT to generate tasks corresponding to {goal: get X news}, and outputs a set of tasks to be executed in JSON format as tasks' in JSON format, where tasks' includes task0, task1, ..., that is, tasks' in JSON format is {task1', task0', ...}. Among them, tasks' includes the task task' to be executed in tasks for obtaining {goal: get X news}.

[0066] S203: Execute the tasks to be executed in the set of tasks to be executed in a structured format to obtain task execution results in a structured format.

[0067] In the embodiment of the present application, after executing S202 to generate a set of tasks to be executed in a structured format, since the set of tasks to be executed includes tasks to be executed in the set of executable tasks for obtaining information to be obtained, in order to meet the information acquisition requirements, it is necessary to perform task execution according to the tasks to be executed in the set of tasks to be executed in a structured format to obtain task execution results in a structured format. Among them, the task execution results in a structured format refer to the information acquisition results in a structured format, and the task execution results in a structured format include the execution results of each task to be executed in the set of tasks to be executed.

[0068] S203 obtains a task execution result in a structured format representing the information acquisition result by executing the task to be executed to meet the information acquisition needs; not only can the undisclosed information on the entire network matching the information to be acquired be acquired, but the same information to be acquired can also be customized to generate different and changing information acquisition results.

[0069] As an example of S203, based on the above S202 example, the computer device executes the task task' to be executed in {task1', task0', ...}, and obtains the task execution result in JSON format as results in JSON format, where results includes result1, result0, ..., that is, the results in JSON format are {result1, result0, ...}.

[0070] S204: Perform language generation on the task execution result in a structured format to obtain an information acquisition result of the information to be acquired.

[0071] In an embodiment of the present application, after executing S203 to obtain the task execution result in a structured format, since the task execution result in a structured format refers to the information acquisition result in a structured format, in order to easily understand the information acquisition result, it is also necessary to generate language for the task execution result in the structured format to obtain the information acquisition result in the natural language of the information to be acquired.

[0072] This S204 generates the task execution results in a structured format into information acquisition results in natural language that are easy to understand, so that users can not only obtain undisclosed information on the entire network in natural language that matches the information to be obtained, but also customize the same information to be obtained to generate different, changing, natural language information acquisition results.

[0073] As an example of S204, based on the above S203 example, the computer device generates language from {result1, result0, ...} to obtain a natural language information acquisition result of {goal: obtain X news}.

[0074] For the above content, see Figure 3 , Figure 3 A schematic diagram of the steps of an information acquisition method provided in an embodiment of the present application, the information acquisition method includes asking questions to a preset generation model, the preset generation model answering, executing tasks to be executed, and summarizing the task execution results. Among them, asking questions to the preset generation model means: through the information to be acquired in a structured format and the set of executable tasks in a structured format, the question combination obtains a target question including the information to be acquired in a structured format and the set of executable tasks in a structured format, and asks the target question to the preset generation model. The preset generation model answering means: generating tasks corresponding to the information to be acquired for the target question through the preset generation model, and outputting a structured format including the tasks to be executed in the executable task set for acquiring the information to be acquired to form a set of tasks to be executed. Executing tasks to be executed means: executing tasks according to the tasks to be executed in the set of tasks to be executed in a structured format, and obtaining task execution results in a structured format. Summarizing the task execution results means: performing language generation on the task execution results in a structured format to obtain the information acquisition results of the natural language of the information to be acquired.

[0075] It can be seen from the above technical scheme that, firstly, the target question is obtained by combining the information to be obtained in a structured format and the set of executable tasks in a structured format; this step combines the target question to be answered by the preset generation model through the information to be obtained that represents the information acquisition requirements and the executable tasks that can be called by the preset generation model. Secondly, the target question is input into the preset generation model to generate the tasks corresponding to the information to be obtained, and the set of tasks to be executed in a structured format is output, wherein the set of tasks to be executed includes the tasks to be executed in the executable task set for obtaining the information to be obtained; this step uses the executable tasks required to be called to obtain the information to be obtained as the tasks to be executed through the preset generation model to answer the target question. Then, the tasks to be executed in the set of tasks to be executed in the structured format are executed to obtain the task execution results in a structured format; this step obtains the task execution results in a structured format representing the information acquisition results by executing the tasks to be executed. Finally, the task execution results in the structured format are language generated to obtain the information acquisition results of the information to be obtained; this step converts the task execution results into information acquisition results in natural language that is easy to understand.

[0076] Based on this, the method answers the target question composed of the information to be obtained and the set of executable tasks through a preset generation model. The tasks to be executed for obtaining the information to be obtained can be called on demand to form a set of tasks to be executed, the tasks to be executed in the set of tasks to be executed can be executed, and the information acquisition results can be output in natural language. Not only can the undisclosed information on the entire network that matches the information to be obtained be obtained, but the same information to be obtained can be customized to generate different and changing information acquisition results, so that the information acquisition results are highly accurate and highly customized, thereby improving the information acquisition effect.

[0077] In an embodiment of the present application, considering that the executable task set in the target problem needs to uniquely identify the executable tasks, clarify the task description and parameter format of the executable tasks, the executable task set includes task identification data of the executable tasks, task description data of the executable tasks and parameter format data of the executable tasks; so that the tasks corresponding to the information to be obtained can be subsequently generated through a preset generation model, and the tasks to be executed in the executable task set for obtaining the information to be obtained can be obtained.

[0078] Based on this, when specifically implementing the above S201, on the one hand, it is necessary to obtain the information to be obtained in a structured format through the object interface, and on the other hand, it is necessary to obtain the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format through the task interface; thus, the information to be obtained in a structured format, the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format are combined into a target problem through the problem combiner. Therefore, the present application provides a possible implementation manner, where the executable task set includes the task identification data of the executable task, the task description data of the executable task, and the parameter format data of the executable task; the above S201 includes the following S2011-S2013 (not shown in the figure):

[0079] S2011: Obtain the information to be obtained in a structured format through the object interface.

[0080] S2012: Obtain the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format through the task interface.

[0081] S2013: Perform problem combination on the information to be obtained in a structured format, the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format through the problem combiner to obtain the target problem.

[0082] The executable task set of S2011-S2013 includes the task identification data of the executable task, the task description data of the executable task, and the parameter format data of the executable task, so that the combined target problem includes the task identification data of the executable task, the task description data of the executable task, and the parameter format data of the executable task; it provides a more accurate and more customized data basis for subsequently generating the task corresponding to the information to be obtained through the preset generation model to obtain the to-be-executed task for obtaining the information to be obtained in the executable task set.

[0083] As an example of S2011 - S2013, the object interface is user interface, the task interface is task interface, and the question combinator is prompt combinator. Based on the above S201 example, the computer device obtains {goal: obtain X news} through user interface, and obtains the task identification data of task in JSON - formatted tasks as id, the task description data of task as des, and the parameter format data of task as params type through task interface; combines {goal: obtain X news}, {tasks: id, des, params type,...} into a target question through the question combinator, that is, {goal: obtain X news, tasks: id, des, params type,...}.

[0084] Among them, {goal: obtain X news, tasks: id, des, params type,...} is specifically as follows:

[0085]

[0086]

[0087] In the embodiment of the present application, based on the executable task set including the task identification data of the executable task, the task description data of the executable task, and the parameter format data of the executable task, the to - be - executed task set needs to uniquely identify the task identification data of the to - be - executed task for obtaining the to - be - obtained question in the executable task set, and clarify the parameter data of the to - be - executed task; then the to - be - executed task set includes the task identification data of the to - be - executed task and the parameter generation data of the to - be - executed task.

[0088] Based on this, when the above S202 is specifically implemented, first, on the basis of the information to be obtained in the target problem, the task identification data of the executable task, and the task description data of the executable task, the executable task for obtaining the information to be obtained is selected from the executable task set as the task to be executed through the preset generation model, and the task identification data of the task to be executed is obtained; then, on the basis of the parameter format data of the executable task in the target problem and the task identification data of the task to be executed, the parameter generation data of the task to be executed is generated for the task to be executed through the preset generation model, and the parameter generation data of the task to be executed is obtained; finally, on the basis of the task identification data of the task to be executed and the parameter generation data of the task to be executed, the task to be executed is generated for the task to be executed through the preset generation model, and the tasks to be executed in a structured format are obtained to form a set of tasks to be executed. Therefore, the present application provides a possible implementation method, and the set of tasks to be executed includes the task identification data of the task to be executed and the parameter generation data of the task to be executed; the above S202 includes the following S2021-S2023 (not shown in the figure):

[0089] S2021: Select tasks based on the information to be obtained in the target problem, the task identification data of the executable tasks, and the task description data of the executable tasks through a preset generation model to obtain the task identification data of the task to be executed.

[0090] S2022: Generate parameters according to the parameter format data of the executable tasks in the target problem and the task identification data of the tasks to be executed through a preset generation model to obtain parameter generation data of the tasks to be executed.

[0091] S2023: Generate tasks according to the task identification data of the tasks to be executed and the parameter generation data of the tasks to be executed by using a preset generation model to obtain a set of tasks to be executed in a structured format.

[0092] S2021-S2023 first selects the task identification data of the task to be executed for obtaining the information to be obtained, and then generates the parameter generation data of the task to be executed for obtaining the information to be obtained, so that the set of tasks to be executed includes the task identification data of the task to be executed and the parameter generation data of the task to be executed; for the subsequent execution of the tasks to be executed in the set of tasks to be executed to meet the information acquisition needs, a more accurate, customized and feasible execution basis is provided, so that not only the undisclosed information on the entire network that matches the information to be obtained can be obtained, but also the same information to be obtained can be customized to generate different and changing information acquisition results.

[0093] As an example of S2021 - S2023, based on the above example of S2011 - S2013, the computer device selects tasks through GPT according to {goal: obtain X news, tasks: id, des, params type,...} for {goal: obtain X news} and {tasks: id, des,...}, and the task identification data of the task to be executed is the id of tasks'; through GPT, according to {tasks: params type} in {goal: obtain X news, tasks: id, des, params type,...} and {tasks': id,...}, parameter generation is performed, and the parameter generation data of the task to be executed is the params of tasks'; through GPT, task generation is performed according to the id of tasks' and the params of tasks', and the tasks' in JSON format is [{id, params},...].

[0094] Among them, [{id, params},...] is specifically as follows:

[0095]

[0096]

[0097] In the embodiment of the present application, when the above S203 is specifically implemented, considering that the executable tasks in the set of executable tasks are not memory resident, first, the task executor needs to load the executable tasks in the set of executable tasks according to the task identification data of the tasks to be executed in the structured format of the set of tasks to be executed, and obtain the loaded tasks to be executed; then, the task executor executes the loaded tasks to be executed according to the parameter generation data of the tasks to be executed in the structured format of the set of tasks to be executed, and obtains the task execution result in the structured format. Therefore, the present application provides a possible implementation manner, and the above S203 includes the following S2031 - S2032 (not shown in the figure):

[0098] S2031: Load tasks through the task executor according to the task identification data of the tasks to be executed in the structured format of the set of tasks to be executed, and obtain the loaded tasks to be executed.

[0099] S2032: Execute the loaded tasks to be executed through the task executor according to the parameter generation data of the tasks to be executed in the structured format of the set of tasks to be executed, and obtain the task execution result in the structured format.

[0100] S2031-S2032 first loads the task to be executed through the task identification data of the task to be executed, and then generates data to execute the loaded task to be executed through the parameters of the task to be executed. It can execute the tasks to be executed in the set of tasks to be executed more correctly and customized, so that the task execution results in a structured format are more correct and customized; not only can it obtain the undisclosed information on the entire network that matches the information to be obtained, but also the same information to be obtained can be customized to generate different and changing information acquisition results.

[0101] As an example of S2031-S2032, the task executor is task implementations. Based on the above S2021-S2023 examples, the computer device loads tasks according to the id of tasks' in [{id, params}, ...] through task implementations to obtain the loaded tasks'; then, through tasksimplementations, according to the params of tasks' in [{id, params}, ...], the loaded tasks' are executed to obtain {result1, result0, ...}.

[0102] Among them, {result1, result0, ...} is as follows:

[0103]

[0104]

[0105] See also Figure 4 , Figure 4 A schematic diagram of specific steps for obtaining a task execution result in a structured format provided for an embodiment of the present application; the specific steps include: parsing a set of tasks to be executed in a structured format to obtain the tasks to be executed in the set of tasks to be executed; determining whether there are any unexecuted tasks to be executed, and if so, loading the tasks to be executed according to the task identification data of the tasks to be executed; generating data according to the parameters of the tasks to be executed, executing the loaded tasks to be executed, and returning to determine whether there are any unexecuted tasks to be executed; if not, outputting the task execution result in a structured format.

[0106] In the embodiment of the present application, considering that the preset generation model has a powerful natural language processing capability, when the above S204 is specifically implemented, the task execution result in a structured format can be input into the preset generation model for language generation, and the information acquisition result of the natural language of the information to be acquired is output. Therefore, the present application provides a possible implementation method, and the above S204 includes S2040 (not shown in the figure): language generation is performed on the task execution result in a structured format by the preset generation model to obtain the information acquisition result of the information to be acquired.

[0107] The S2040 can generate the task execution result in a structured format into an information acquisition result in natural language that is easier to understand more accurately and more customized through a preset generation model, making the information acquisition result more accurate and more customized.

[0108] As an example of the S2040, based on the above S203 example, the computer device inputs {result1, result0,...} into GPT for language generation to obtain an information acquisition result in natural language of {goal: obtain X news}.

[0109] The above S2040 can at least adopt the following two specific implementation manners:

[0110] One implementation manner means that in order to reduce the complexity of generating the task execution result in a structured format into an information acquisition result in natural language through a preset generation model, first generate the task execution result in a structured format into an information acquisition result in a preset text format with lower complexity through the preset generation model, and then convert the information acquisition result in the preset text format into an information acquisition result in a target text format with higher complexity. Therefore, the present application provides a possible implementation manner, and S2040 includes the following S1-S2 (not shown in the figure):

[0111] S1: Generate the task execution result in a structured format into an information acquisition result in a preset text format according to the preset text format through a preset generation model.

[0112] S2: Perform format conversion on the information acquisition result in the preset text format according to the target text format to obtain an information acquisition result in the target text format; the complexity of the target text format is greater than that of the preset text format.

[0113] The S1-S2 generates the task execution result in a structured format into an information acquisition result in a preset text format with lower complexity and then converts it into an information acquisition result in a target text format with higher complexity, which can reduce the language generation complexity of the preset generation model and improve the language generation accuracy of the preset generation model.

[0114] As an example of S1 - S2, the preset text format is Markdown format. Markdown is a lightweight markup language that allows writing in a plain - text format that is easy to read and write. The target text format is HTML format. HTML is a markup language that includes a series of tags. Through these tags, the document formats on the network can be unified, making scattered network resources connected into a logical whole. Based on the above S2040 example, first, GPT is used to generate language for {result1, result0, …} according to the Markdown format to obtain the information acquisition result in Markdown format. Then, the information acquisition result in Markdown format is converted into the information acquisition result in HTML format.

[0115] Specifically, the information acquisition result in Markdown format "#This is a title" is converted into the information acquisition result in HTML format as " <h1>This is a title< / h1> ".

[0116] Another implementation method means that in order to reduce the operation process of generating the information acquisition result in natural language from the task execution result in structured format through a preset generation model, the task execution result in structured format is directly generated into the information acquisition result in the target text format with higher complexity through the preset generation model. Therefore, the present application provides a possible implementation method. The above S2040 includes S3 (not shown in the figure): The preset generation model is used to generate language for the task execution result in structured format according to the target text format to obtain the information acquisition result in the target text format.

[0117] As an example of S3, the target text format is HTML format. Based on the above S2040 example, GPT is used to generate language for {result1, result0, …} according to the HTML format to obtain the information acquisition result in HTML format.

[0118] In addition, in the embodiments of the present application, after the information acquisition result of the information to be acquired is generated in S201 - S204 above, in order for the user who inputs the information to be acquired to clearly and intuitively obtain the information acquisition result of the information to be acquired, it is also necessary to display the information acquisition result of the information to be acquired corresponding to the information to be acquired. Therefore, the present application provides a possible implementation method. The method further includes S4 (not shown in the figure): Display the information acquisition result of the information to be acquired corresponding to the information to be acquired.

[0119] As an example of S4, based on the above S201 - S204 examples, the computer device displays the information acquisition result in natural language of {goal: Obtain X news} corresponding to {goal: Obtain X news}.

[0120] In addition, in the embodiments of the present application, when considering displaying the information acquisition result of the information to be acquired corresponding to the information to be acquired, page information with rich highlighting, colors, images, links, or tables is more friendly and attractive; based on this, when the information acquisition result of the information to be acquired is an information acquisition result in a target text format, and the target text format is a hypertext format, it is also necessary to generate a corresponding information page for the information acquisition result in the hypertext format; so as to display the information page corresponding to the information to be acquired, thereby improving the information acquisition effect and information acquisition experience in the visual dimension. Therefore, the present application provides a possible implementation manner. When the information acquisition result of the information to be acquired is an information acquisition result in a target text format, and the target text format is a hypertext format, the method further includes S5 (not shown in the figure): generating an information page corresponding to the information acquisition result according to the information acquisition result in the hypertext format; correspondingly, the above S4 includes S40 (not shown in the figure): displaying the information page corresponding to the information to be acquired.

[0121] As an example of the above S5 and S40, the hypertext format is the html format. Based on the above examples of S201 - S204, when the information acquisition result of {goal: obtain X news} is an information acquisition result in the html format, the computer device generates an information page corresponding to the information acquisition result for the information acquisition result in the html format; and displays the information page corresponding to {goal: obtain X news}.

[0122] See Figure 5 , Figure 5 FIG. is a schematic diagram of steps for displaying the information acquisition result based on the structured format task execution result provided by the embodiments of the present application. The steps include summarizing the task execution result, converting the format, generating the page, and displaying the page. Among them, summarizing the task execution result means: obtaining the information acquisition result in the plain text format by performing language generation on the structured format task execution result according to the preset generation model in the plain text format. Converting the format means: performing format conversion on the information acquisition result in the plain text format according to the hypertext format to obtain the information acquisition result in the hypertext format. Generating the page means: generating an information page corresponding to the information acquisition result according to the information acquisition result in the hypertext format. Displaying the page means: displaying the information page corresponding to the information to be acquired.

[0123] In addition, in the embodiments of the present application, when performing the above-mentioned S202 to input the target problem into the preset generation model to generate tasks corresponding to the information to be obtained, there may be a situation where the executable task set lacks the to-be-executed tasks for obtaining the information to be obtained; based on this, the preset generation model can also generate the required tasks outside the executable task set for obtaining the information to be obtained as the to-be-expanded tasks to form a to-be-expanded task set; thereby expanding the executable task set and updating the executable task set; so as to return to the above-mentioned S201 and re-execute the information acquisition method of the above-mentioned S201-S204. Therefore, the present application provides a possible implementation manner, and the method further includes the following S6-S9 (not shown in the figure):

[0124] S6: Generate tasks corresponding to the information to be obtained for the target problem through the preset generation model to obtain a to-be-expanded task set in a structured format; the to-be-expanded task set includes the to-be-expanded tasks outside the executable task set for obtaining the information to be obtained.

[0125] S7: Expand the executable task set according to the to-be-expanded task set in the structured format to obtain an expanded executable task set.

[0126] S8: Update the executable task set according to the expanded executable task set.

[0127] S9: Return to execute the combination of the problem for the information to be obtained in the structured format and the executable task set in the structured format to obtain the target problem.

[0128] In S6-S9, the required tasks outside the executable task set called for obtaining the information to be obtained are used as the to-be-expanded tasks through the preset generation model to form a to-be-expanded task set to expand and update the executable task set, so as to re-execute the information acquisition method, which can avoid the problem of information acquisition failure in the case where the executable task set lacks the to-be-executed tasks for obtaining the information to be obtained, and further improve the information acquisition effect and information acquisition experience.

[0129] Among them, on the basis that the executable task set includes the task identification data of the executable task, the task description data of the executable task, and the parameter format data of the executable task, the to-be-expanded task set needs to clarify the task description of the to-be-expanded tasks outside the executable task set for obtaining the to-be-obtained problem; then the to-be-expanded task set includes the task description data of the to-be-expanded tasks.

[0130] Based on this, when specifically implementing S6, first, based on the task description data of the information to be obtained and the executable tasks in the target problem, the required tasks for obtaining the information to be obtained outside the set of executable tasks are expanded through a preset generation model as the tasks to be expanded, and the task description data of the tasks to be expanded is obtained; then, the task description data of the tasks to be expanded is used by the preset generation model for task generation to obtain the tasks to be expanded in a structured format to form a set of tasks to be expanded. Therefore, the present application provides a possible implementation manner, and the set of tasks to be expanded includes the task description data of the tasks to be expanded; the above S6 includes the following S61 - S62 (not shown in the figure):

[0131] S61: Expand tasks through a preset generation model according to the task description data of the information to be obtained and the executable tasks in the target problem, and obtain the task description data of the tasks to be expanded;

[0132] S62: Generate tasks through a preset generation model according to the task description data of the tasks to be expanded, and obtain a set of tasks to be expanded in a structured format.

[0133] By expanding the task description data of the tasks to be expanded for obtaining the information to be obtained outside the set of executable tasks in S61 - S62, the set of tasks to be expanded includes the task description data of the tasks to be expanded; this provides an expansion basis for subsequent expansion of the set of executable tasks.

[0134] Based on the above content, refer to Figure 6 , Figure 6 which is a system framework diagram of an information acquisition method provided by an embodiment of the present application. The system includes a computer device, and the framework of the computer device includes a user interface, a scheduler executor, a prompt combiner, a task interface, and task implementations. Among them, the user interface obtains the goal input by the user and sends the goal in JSON format to the executor.

[0135] The executor receives the goal in JSON format sent by the user interface and sends the goal in JSON format to the prompt combiner.

[0136] On the one hand, the prompt combinator receives the goal in JSON format sent by the executor. On the other hand, it obtains the tasks in JSON format from the task interface. Through the goal in JSON format and the tasks in JSON format, the target problem {goal, tasks} is combined and sent to GPT. Among them, the task interface stores tasks, and the tasks include the id, des, and params type of the task.

[0137] GPT receives {goal, tasks} sent by the prompt combinator, generates the tasks corresponding to the goal for {goal, tasks}, and the output is tasks' in JSON format. It returns tasks' in JSON format to the executor through the prompt combinator.

[0138] The executor receives tasks' in JSON format returned by GPT through the prompt combinator and calls taskimplementations.

[0139] Task implementations execute tasks according to task' in tasks' in JSON format, obtain results in JSON format, and return results in JSON format to the executor. Among them, task implementations store tasks, and the tasks include the id and params type of the task.

[0140] The executor receives results in JSON format returned by task implementations and sends results in JSON format to GPT through the prompt combinator.

[0141] GPT receives results in JSON format sent by the executor through the prompt combinator, generates the natural language information acquisition result of the goal from the results in JSON format, and returns the information acquisition result to the user interface through the prompt combinator and the executor.

[0142] The user interface receives the information acquisition result returned by GPT through the prompt combinator and the executor and displays the information acquisition result to the user.

[0143] In addition, the dashed line between the task interface and the task implementation indicates that the tasks in the task interface and the task implementation are in one-to-one correspondence, that is, each task in the task implementation has a corresponding task in the task interface.

[0144] The dashed line between the executor and the task interface indicates that when GPT returns tasks stored in the task interface to the executor through the prompt combinator and the tasks need to be expanded, the executor expands the tasks stored in the task interface, making the information acquisition ability of the information acquisition method system more and more powerful.

[0145] In addition, the framework of the computer device includes a base library base, which implements some basic capabilities, such as network requests, database connections, account management, etc.

[0146] In summary, see Figure 7 , Figure 7 which is the specific flowchart of an information acquisition method provided by an embodiment of the present application. The method includes:

[0147] S701: Obtain the information to be acquired in a structured format through the object interface.

[0148] S702: Obtain the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format through the task interface.

[0149] S703: Combine the information to be acquired in a structured format, the task identification data of the executable task in a structured format, the task description data of the executable task in a structured format, and the parameter format data of the executable task in a structured format through the question combinator to obtain the target question.

[0150] S704: Select tasks according to the information to be acquired, the task identification data of the executable task, and the task description data of the executable task in the target question through the preset generation model to obtain the task identification data of the task to be executed.

[0151] S705: Generate parameters according to the parameter format data of the executable task and the task identification data of the task to be executed in the target question through the preset generation model to obtain the parameter generation data of the task to be executed.

[0152] S706: Generate tasks according to the task identification data of the tasks to be executed and the parameter generation data of the tasks to be executed by using a preset generation model to obtain a set of tasks to be executed in a structured format.

[0153] S707: Loading tasks according to the task identification data of the tasks to be executed in the set of tasks to be executed in a structured format by the task executor to obtain the loaded tasks to be executed.

[0154] S708: The task executor generates data according to the parameters of the tasks to be executed in the set of tasks to be executed in a structured format, and executes the loaded tasks to be executed, to obtain task execution results in a structured format.

[0155] S709: Performing language generation on the task execution result in a structured format according to a plain text format through a preset generation model to obtain an information acquisition result in a preset text format.

[0156] S710: Convert the information acquisition result in plain text format according to the hypertext format to obtain the information acquisition result in hypertext format.

[0157] S711: Generate an information page corresponding to the information acquisition result according to the information acquisition result in a hypertext format.

[0158] S712: Display an information page corresponding to the information to be obtained.

[0159] It can be seen from the above technical solution that by presetting the generation model to answer the target question composed of the information to be obtained and the executable task set, the tasks to be executed for obtaining the information to be obtained can be called on demand to form a set of tasks to be executed, the tasks to be executed in the set of tasks to be executed can be executed, and the information acquisition results can be displayed on an information page; not only can the undisclosed information on the entire network that matches the information to be obtained be obtained, but also the same information to be obtained can be customized to generate different and changing information acquisition results, so that the information acquisition results are highly accurate and highly customized, thereby improving the information acquisition effect.

[0160] It should be noted that, based on the implementation methods provided in the above aspects, this application can also be further combined to provide more implementation methods.

[0161] based on Figure 2 Corresponding to the information acquisition method provided in the embodiment, the embodiment of the present application also provides an information acquisition device, see Figure 8 , Figure 8 This is a structural diagram of an information acquisition device provided in an embodiment of the present application. The information acquisition device 800 includes: a question combination unit 801, a task generation unit 802, a task execution unit 803 and a language generation unit 804;

[0162] A problem combination unit 801 is configured to combine the information to be obtained in a structured format and a set of executable tasks in a structured format to obtain a target problem;

[0163] A task generation unit 802 is configured to generate tasks corresponding to the information to be obtained for the target problem through a preset generation model to obtain a set of tasks to be executed in a structured format; the set of tasks to be executed includes tasks to be executed in the set of executable tasks for obtaining the information to be obtained;

[0164] A task execution unit 803 is configured to execute the tasks to be executed in the set of tasks to be executed in a structured format to obtain a task execution result in a structured format;

[0165] A language generation unit 804 is configured to generate a language for the task execution result in a structured format to obtain an information acquisition result of the information to be obtained.

[0166] In a possible implementation manner, the set of executable tasks includes task identification data of the executable tasks, and the problem combination unit 801 is specifically configured to:

[0167] Obtain the information to be obtained in a structured format through an object interface;

[0168] Obtain the task identification data of the executable tasks in a structured format, the task description data of the executable tasks in a structured format, and the parameter format data of the executable tasks in a structured format through a task interface;

[0169] Combine the information to be obtained in a structured format, the task identification data of the executable tasks in a structured format, the task description data of the executable tasks in a structured format, and the parameter format data of the executable tasks in a structured format through a problem combiner to obtain a target problem.

[0170] In a possible implementation manner, the set of tasks to be executed includes task identification data of the tasks to be executed and parameter generation data of the tasks to be executed; the task generation unit 802 is specifically configured to:

[0171] Select tasks according to the information to be obtained in the target problem, the task identification data of the executable tasks, and the task description data of the executable tasks through a preset generation model to obtain the task identification data of the tasks to be executed;

[0172] Generate parameters according to the parameter format data of the executable tasks in the target problem and the task identification data of the tasks to be executed through a preset generation model to obtain the parameter generation data of the tasks to be executed;

[0173] Generate tasks by using a preset generation model to generate data based on the task identification data of the task to be executed and the parameters of the task to be executed, and obtain a set of tasks to be executed in a structured format.

[0174] In a possible implementation, the task execution unit 803 is specifically configured to:

[0175] Load the tasks to be executed through a task executor according to the task identification data of the tasks to be executed in the set of tasks to be executed in a structured format, and obtain the tasks to be executed after loading;

[0176] Execute the tasks to be executed after loading through a task executor according to the data generated from the parameters of the tasks to be executed in the set of tasks to be executed in a structured format, and obtain a task execution result in a structured format.

[0177] In a possible implementation, the language generation unit 804 is specifically configured to:

[0178] Generate language for the task execution result in a structured format through a preset generation model to obtain an information acquisition result of the information to be obtained.

[0179] In a possible implementation, the language generation unit 804 is specifically configured to:

[0180] Generate language for the task execution result in a structured format through a preset generation model according to a preset text format to obtain an information acquisition result in the preset text format;

[0181] Convert the format of the information acquisition result in the preset text format according to the target text format to obtain an information acquisition result in the target text format; the complexity of the target text format is greater than the complexity of the preset text format.

[0182] In a possible implementation, the language generation unit 804 is specifically configured to:

[0183] Generate language for the task execution result in a structured format through a preset generation model according to the target text format to obtain an information acquisition result in the target text format.

[0184] In a possible implementation, the device further includes: a display unit;

[0185] The information display unit is configured to display the information acquisition result of the information to be obtained corresponding to the information to be obtained.

[0186] In a possible implementation, when the information acquisition result of the information to be obtained is an information acquisition result in the target text format and the target text format is a hypertext format, the device further includes: a page generation unit;

[0187] A page generation unit, configured to generate an information page corresponding to the information acquisition result according to the information acquisition result in hypertext format;

[0188] An information display unit, specifically configured to:

[0189] Display the information page corresponding to the information to be acquired.

[0190] In a possible implementation, the apparatus further includes: a task expansion unit, a task update unit, and a step return unit;

[0191] The task generation unit 802 is further configured to generate a set of tasks to be expanded corresponding to the information to be acquired for the target problem through a preset generation model, and obtain a set of tasks to be expanded in a structured format; the set of tasks to be expanded includes tasks to be expanded for acquiring the information to be acquired outside the set of executable tasks;

[0192] The task expansion unit is configured to expand the set of executable tasks according to the set of tasks to be expanded in a structured format, and obtain an expanded set of executable tasks;

[0193] The task update unit is configured to update the set of executable tasks according to the expanded set of executable tasks;

[0194] The step return unit is configured to return to perform a problem combination on the information to be acquired in a structured format and the set of executable tasks in a structured format, and obtain the target problem.

[0195] In a possible implementation, the set of tasks to be expanded includes task description data of the tasks to be expanded; the task generation unit 802 is specifically configured to:

[0196] Expand the task description data of the tasks to be expanded through a preset generation model according to the task description data of the information to be acquired and the executable tasks in the target problem, and obtain the task description data of the tasks to be expanded;

[0197] Generate a set of tasks to be expanded in a structured format through a preset generation model according to the task description data of the tasks to be expanded.

[0198] It can be seen from the above technical solution that the information acquisition device includes a question combination unit, a task generation unit, a task execution unit and a language generation unit. The question combination unit obtains the target question by combining the information to be acquired in a structured format and the executable task set in a structured format; the unit combines the target question to be answered by the preset generation model through the information to be acquired that represents the information acquisition requirements and the executable tasks that can be called by the preset generation model. The task generation unit inputs the target question into the preset generation model to generate tasks corresponding to the information to be acquired, and outputs a set of tasks to be executed in a structured format, wherein the set of tasks to be executed includes the tasks to be executed in the executable task set for acquiring the information to be acquired; the unit uses the executable tasks required to be called to acquire the information to be acquired as the tasks to be executed through the preset generation model to answer the target question. The task execution unit executes the tasks to be executed in the set of tasks to be executed in a structured format to obtain the task execution results in a structured format; the unit obtains the task execution results in a structured format representing the information acquisition results by executing the tasks to be executed. The language generation unit generates language from the task execution result in a structured format to obtain an information acquisition result of the information to be acquired; this unit converts the task execution result into an information acquisition result in a natural language that is easy to understand.

[0199] Based on this, the device answers the target questions composed of the information to be obtained and the set of executable tasks through a preset generation model. It can call the tasks to be executed for obtaining the information to be obtained on demand to form a set of tasks to be executed, execute the tasks to be executed in the set of tasks to be executed, and output the information acquisition results in natural language. It can not only obtain the undisclosed information on the entire network that matches the information to be obtained, but also the same information to be obtained can be customized to generate different and changing information acquisition results, so that the information acquisition results are highly accurate and customized, thereby improving the information acquisition effect.

[0200] The present application also provides a computer device, which may be a server. Figure 9 , Figure 9 A structural diagram of a server provided in an embodiment of the present application, the server 900 may have relatively large differences due to different configurations or performances, and may include one or more processors, such as a CPU 922, and a memory 932, one or more storage media 930 (such as one or more mass storage devices) storing application programs 942 or data 944. Among them, the memory 932 and the storage medium 930 can be short-term storage or permanent storage. The program stored in the storage medium 930 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the server. Furthermore, the central processing unit 922 can be configured to communicate with the storage medium 930 and execute a series of instruction operations in the storage medium 930 on the server 900.

[0201] The server 900 may also include one or more power supplies 926, one or more wired or wireless network interfaces 950, one or more input / output interfaces 958, and / or one or more operating systems 941, such as Windows Server TM , Mac OS X TM , Unix TM , Linux TM , FreeBSD TM and so on.

[0202] In this embodiment, the central processing unit 922 in the server 900 may execute the methods provided in various alternative implementation manners of the above embodiment.

[0203] The computer device provided by the embodiment of the present application may also be a terminal. Refer to Figure 10 , Figure 10 , which is a structural diagram of a terminal provided by the embodiment of the present application. Taking the terminal as a smart phone as an example, the smart phone includes: a radio frequency (RF) circuit 1010, a memory 1020, an input unit 1030, a display unit 1040, a sensor 1050, an audio circuit 1060, a wireless fidelity (WiFi) module 1070, a processor 1080, and a power supply 1090 and other components. The input unit 1030 may include a touch panel 1031 and other input devices 1032. The display unit 1040 may include a display panel 1041. The audio circuit 1060 may include a speaker 1061 and a microphone 1062. Those skilled in the art can understand that Figure 10 the structure of the smart phone shown in

[0204] does not limit the smart phone, and may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0205] The processor 1080 is the control center of the smart phone, connecting various parts of the entire smart phone using various interfaces and circuits. By running or executing software programs and / or modules stored in the memory 1020, and by invoking data stored in the memory 1020, it performs various functions of the smart phone and processes data. Optionally, the processor 1080 may include one or more processing units; preferably, the processor 1080 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor may not be integrated into the processor 1080 either.

[0206] In this embodiment, the processor 1080 in the smart phone can execute the methods provided in various optional implementation manners of the above embodiment.

[0207] According to one aspect of the present application, there is provided a computer-readable storage medium for storing a computer program. When the computer program runs on a computer device, the computer device is caused to execute the methods provided in various optional implementation manners of the above embodiment.

[0208] According to one aspect of the present application, there is provided a computer program product. The computer program product includes a computer program stored in a computer-readable storage medium. The processor of the computer device reads the computer program from the computer-readable storage medium, and the processor executes the computer program, causing the computer device to execute the methods provided in various optional implementation manners of the above embodiment.

[0209] The descriptions of the processes or structures corresponding to the above respective drawings each have their own focuses. For parts not detailed in a certain process or structure, reference can be made to the relevant descriptions of other processes or structures.

[0210] The term "including" in the specification of the present application and any of its variations in the above drawings is intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0211] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections between each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0212] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0213] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0214] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memory (ROM), RAM, magnetic disks, or optical discs and other various media that can store computer programs.

[0215] As described above, the above embodiments are only used to illustrate the technical solution of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. An information acquisition method, characterized in that, the method includes: Combining problems of the information to be acquired in structured format and the set of executable tasks in the structured format to obtain a target problem; Generating tasks corresponding to the information to be acquired for the target problem through a preset generation model to obtain a set of executable tasks in the structured format; the set of executable tasks includes executable tasks for acquiring the information to be acquired in the set of executable tasks; Executing the executable tasks in the set of executable tasks in the structured format to obtain a task execution result in the structured format; Generating a language for the task execution result in the structured format to obtain an information acquisition result of the information to be acquired.

2. The method according to claim 1, characterized in that, the set of executable tasks includes task identification data of executable tasks, task description data of the executable tasks, and parameter format data of the executable tasks; The combining problems of the information to be acquired in structured format and the set of executable tasks in the structured format to obtain a target problem includes: Obtaining the information to be acquired in the structured format through an object interface; Obtaining the task identification data of the executable tasks in the structured format, the task description data of the executable tasks in the structured format, and the parameter format data of the executable tasks in the structured format through a task interface; Combining problems of the information to be acquired in the structured format, the task identification data of the executable tasks in the structured format, the task description data of the executable tasks in the structured format, and the parameter format data of the executable tasks in the structured format through a problem combiner to obtain the target problem.

3. The method according to claim 2, characterized in that, the set of executable tasks includes task identification data of the executable tasks and parameter generation data of the executable tasks; The generating tasks corresponding to the information to be acquired for the target problem through a preset generation model to obtain a set of executable tasks in the structured format includes: Selecting tasks through the preset generation model according to the information to be acquired, task identification data of executable tasks, and task description data of executable tasks in the target problem to obtain task identification data of the executable tasks; Generating parameters through the preset generation model according to the parameter format data of the executable tasks and the task identification data of the executable tasks in the target problem to obtain parameter generation data of the executable tasks; Generating tasks through the preset generation model according to the task identification data of the executable tasks and the parameter generation data of the executable tasks to obtain a set of executable tasks in the structured format.

4. The method according to claim 3, characterized in that, The executing the executable tasks in the set of executable tasks in the structured format to obtain a task execution result in the structured format includes: The task executor loads tasks based on the task identification data of the tasks to be executed in the set of tasks to be executed in the structured format, and obtains the tasks to be executed after loading; The task executor executes the loaded tasks to be executed based on the parameter generation data of the tasks to be executed in the set of tasks to be executed in the structured format, and obtains the task execution results in the structured format.

5. The method according to claim 1, wherein, performing language generation on the task execution results in the structured format to obtain the information acquisition results of the information to be acquired includes: performing language generation on the task execution results in the structured format through the preset generation model to obtain the information acquisition results of the information to be acquired.

6. The method according to claim 5, wherein, performing language generation on the task execution results in the structured format through the preset generation model to obtain the information acquisition results of the information to be acquired includes: performing language generation on the task execution results in the structured format through the preset generation model according to the preset text format to obtain the information acquisition results in the preset text format; performing format conversion on the information acquisition results in the preset text format according to the target text format to obtain the information acquisition results in the target text format; the complexity of the target text format is greater than the complexity of the preset text format.

7. The method according to claim 5, wherein, performing language generation on the task execution results in the structured format through the preset generation model to obtain the information acquisition results of the information to be acquired includes: performing language generation on the task execution results in the structured format through the preset generation model according to the target text format to obtain the information acquisition results in the target text format.

8. The method according to any one of claims 1-7, wherein, the method further includes: displaying the information acquisition results of the information to be acquired corresponding to the information to be acquired.

9. The method according to claim 8, wherein, when the information acquisition results of the information to be acquired are the information acquisition results in the target text format and the target text format is the hypertext format, the method further includes: generating an information page corresponding to the information acquisition results according to the information acquisition results in the hypertext format; displaying the information acquisition results of the information to be acquired corresponding to the information to be acquired includes: displaying the information page corresponding to the information to be acquired.

10. The method according to claim 2, wherein, the method further includes: performing task generation for the information to be acquired on the target problem through the preset generation model to obtain a set of tasks to be expanded in the structured format; the set of tasks to be expanded includes tasks to be expanded for acquiring the information to be acquired outside the set of executable tasks; performing task expansion on the set of executable tasks according to the set of tasks to be expanded in the structured format to obtain an expanded set of executable tasks; Update the executable task set according to the expanded executable task set; Return to perform the problem combination of the information to be obtained in the structured format and the executable task set in the structured format to obtain a target problem.

11. The method according to claim 10, wherein, the task set to be expanded includes the task description data of the task to be expanded; the generating the task set to be expanded in the structured format by performing, by the preset generation model, the task corresponding to the information to be obtained on the target problem includes: performing, by the preset generation model, task expansion according to the task description data of the information to be obtained and the executable task in the target problem to obtain the task description data of the task to be expanded; performing, by the preset generation model, task generation according to the task description data of the task to be expanded to obtain the task set to be expanded in the structured format.

12. An information acquisition device, wherein, the device includes: a problem combination unit, a task generation unit, a task execution unit, and a language generation unit; the problem combination unit is configured to perform problem combination on the information to be obtained in the structured format and the executable task set in the structured format to obtain a target problem; the task generation unit is configured to perform, by a preset generation model, task generation corresponding to the information to be obtained on the target problem to obtain the task set to be executed in the structured format; the task set to be executed includes the tasks to be executed in the executable task set for obtaining the information to be obtained; the task execution unit is configured to execute the tasks to be executed in the task set to be executed in the structured format to obtain the task execution result in the structured format; the language generation unit is configured to perform language generation on the task execution result in the structured format to obtain the information acquisition result of the information to be obtained.

13. A computer device, wherein, the computer device includes a processor and a memory: the memory is configured to store a computer program and transmit the computer program to the processor; the processor is configured to execute the method according to any one of claims 1-11 according to the instructions in the computer program.

14. A computer-readable storage medium, wherein, the computer-readable storage medium is configured to store a computer program, and when the computer program runs on a computer device, the computer device is caused to execute the method according to any one of claims 1-11.

15. A computer program product, including a computer program, wherein, when the computer program runs on a computer device, the computer device is caused to execute the method according to any one of claims 1-11.

Images