Information processing method, information processing system and related equipment
Through the collaborative work of the RPA subsystem and LLM, the problems of low efficiency and lack of accuracy in traditional information processing methods have been solved, efficient and accurate information processing has been achieved, the technical threshold has been lowered, and human operational errors have been avoided.
Patent Information
- Application Number
- CN202510835164.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-20
- Publication Date
- 2025-09-19
AI Technical Summary
Traditional information processing methods rely on manual search and processing, which is inefficient and difficult to ensure the comprehensiveness and accuracy of information. Robotic process automation systems find it difficult to judge the logic of information sources, resulting in inaccurate operations and high technical barriers.
The RPA subsystem and the Large Language Model (LLM) work together. The RPA subsystem receives user instructions, the LLM analyzes the web page logic and generates an action instruction set, and the RPA subsystem executes the operation, lowering the technical threshold for users to write computer instructions and improving operation accuracy.
It achieves efficient and accurate information processing, lowers the technical threshold, avoids human operational errors, and improves the efficiency and accuracy of information processing.
Smart Images
Figure CN120670693A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer technology, and in particular to an information processing method, an information processing system, a computer device, a computer-readable storage medium, and a computer program product. Background Art
[0002] With the rapid development of the information age, users need to process large amounts of information every day. For example, companies need to quickly retrieve relevant information for industry analysis, and employees need to process large amounts of data in office automation (OA) systems every day in their daily work.
[0003] Traditional information processing relies primarily on manual search and processing, which is not only inefficient but also difficult to ensure the comprehensiveness and accuracy of information. To address this, the industry has developed tools such as Robotic Process Automation (RPA) systems to help automate some information processing.
[0004] However, it is difficult for related tools to determine the inherent logic of the information source (such as a web page), making related operations less precise; at the same time, related operations need to be triggered based on computer instructions, which has a high technical threshold. Summary of the Invention
[0005] In view of this, the present application provides an information processing method, which can process information efficiently, accurately and simply.
[0006] In a first aspect, the present application provides an information processing method, which is applied to an information processing system, wherein the information processing system includes an RPA subsystem and a large language model (LLM). The method includes:
[0007] The RPA subsystem receives instruction information input by a user, where the instruction information is natural language information used to indicate information to be obtained;
[0008] The large language model determines the information to be acquired, the first target webpage, and feature information of the information to be acquired according to the instruction information sent by the RPA subsystem;
[0009] The RPA subsystem opens the first target webpage and sends all elements on the first target webpage to the large language model, where the elements include at least one of text information or image information of the first target webpage;
[0010] The large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem;
[0011] The RPA subsystem operates the target operation object according to the action instruction set to obtain the information to be obtained.
[0012] In some possible implementations, the large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including:
[0013] The large language model determines that the target operation object is an input box, and generates a feature information input instruction so that the RPA subsystem inputs the feature information of the information to be acquired into the input box.
[0014] In some possible implementations, the large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including:
[0015] The large language model determines that the target operation object is an index control whose text information matches the feature information, and generates a click instruction to enable the RPA subsystem to click the index control to load a second target web page including the information to be obtained.
[0016] In some possible implementations, the large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including:
[0017] The large language model determines that the target operation object is at least one page turning control, and generates a click instruction and a traversal instruction to enable the RPA subsystem to traverse each web page indicated by the at least one page turning control to obtain the information to be obtained.
[0018] In some possible implementations, the large language model determines the first target webpage according to the instruction information sent by the RPA subsystem, including:
[0019] The large language model obtains the information to be acquired according to the indication information, and when the information to be acquired matches a preset keyword, determines the first target webpage according to a preset correspondence relationship between the preset keywords.
[0020] In some possible implementations, the method further includes:
[0021] The RPA subsystem sends the information to be acquired to the large language model;
[0022] The large language model generates target information according to the information to be acquired, and sends the target information to the RPA subsystem;
[0023] The RPA subsystem stores the target information and sends the target information to the user at a preset period.
[0024] In some possible implementations, the characteristic information includes at least one of the subject, publisher, keyword, or release time of the information to be obtained.
[0025] In a second aspect, the present application provides an information processing system, comprising a robotic process automation (RPA) subsystem and a large language model (LLM);
[0026] The RPA subsystem is configured to receive instruction information input by a user, where the instruction information is natural language information indicating information to be acquired;
[0027] The large language model is used to determine the information to be acquired, the first target webpage, and feature information of the information to be acquired based on the instruction information sent by the RPA subsystem;
[0028] The RPA subsystem is further configured to open the first target webpage and send all elements on the first target webpage to the large language model, where the elements include at least one of text information or image information of the first target webpage;
[0029] The large language model is further used to determine a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generate an action instruction set executable by the RPA subsystem;
[0030] The RPA subsystem is further configured to operate the target operation object according to the action instruction set to obtain the information to be obtained.
[0031] The system can also be used to execute the information processing method as described in any implementation of the first aspect.
[0032] In a third aspect, the present application provides a computer device, comprising a processor and a memory.
[0033] The processor is configured to execute instructions stored in the memory, so that the computer device executes the information processing method as described in the first aspect or any one of the implementations of the first aspect.
[0034] In a fourth aspect, the present application provides a computer-readable storage medium comprising instructions that, when executed on a computer device, cause the computer device to execute the information processing method described in the first aspect or any one of the implementations of the first aspect.
[0035] In a fifth aspect, the present application provides a computer program product comprising instructions, which, when executed on a computer device, enables a server to execute the information processing method described in the first aspect or any one of the implementations of the first aspect.
[0036] Based on the implementation methods provided in the above aspects, this application can also be further combined to provide more implementation methods.
[0037] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:
[0038] The present application provides an information processing method. This method uses LLM to analyze the natural language information input by the user to obtain the information to be obtained, the corresponding web page, and the characteristic information of the information to be obtained, thereby eliminating the need for users to write computer instructions themselves and lowering the technical threshold. Furthermore, it can analyze the inherent logic of the web page to accurately generate corresponding action instructions for the target operation object, thereby improving the accuracy of the method. Furthermore, this method uses an RPA subsystem to introduce a digital workforce, achieving efficient work around the clock and avoiding human operational errors, thereby improving the efficiency of information processing. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0040] Figure 1 A flowchart of an information processing method provided in an embodiment of the present application;
[0041] Figure 2 A schematic diagram of the structure of an information processing system provided in an embodiment of the present application. DETAILED DESCRIPTION
[0042] The following will describe the solutions in the embodiments provided in this application in conjunction with the drawings in this application.
[0043] The terms used in the following embodiments are for the purpose of describing specific embodiments only and are not intended to limit the present application. The terms "first" and "second" in the embodiments of the present application are used for descriptive purposes only and should not be understood to indicate or imply relative importance, the temporal order of operation, or implicitly indicate the number of technical features indicated. Therefore, features defined as "first" or "second" may explicitly or implicitly include one or more of such features.
[0044] In order to facilitate understanding of the technical solution of this application, some technical terms involved in this application are first introduced.
[0045] Robotic Process Automation (RPA) is a business process automation technology based on software robots and artificial intelligence (AI). It can simulate manual operations through simple configuration without changing the original system, helping enterprises or employees complete repetitive and monotonous procedural tasks.
[0046] Large Language Models (LLMs) are deep learning models trained using large amounts of text data. LLMs rely on large-scale data training, for example, by training language models (such as BERT and the GPT series) based on neural networks (e.g., the Transformer architecture), enabling semantic parsing, reasoning, and generation of text and speech.
[0047] With the rapid development of the information age, users are required to process vast amounts of information daily. For example, businesses need to quickly retrieve relevant information for industry analysis, while employees face the daily burden of processing vast amounts of data in office automation (OA) systems. Traditional information processing relies primarily on manual search and processing, which is not only inefficient but also difficult to ensure the comprehensiveness and accuracy of information. To address this, the industry has developed tools such as Robotic Process Automation (RPA) systems to automate some information processing. However, these tools struggle to understand the inherent logic of information sources (such as web pages), resulting in inaccurate operations. Furthermore, these operations require triggering computer instructions, posing a high technical barrier to entry.
[0048] In view of this, an embodiment of the present application provides an information processing method, which is applied to an information processing system, wherein the information processing system includes an RPA subsystem and an LLM. Specifically, the method includes: the RPA subsystem receives instruction information input by a user, where the instruction information is natural language information used to indicate the information to be obtained; the LLM determines the information to be obtained, the first target web page, and the feature information of the information to be obtained based on the instruction information; then, the RPA subsystem opens the first target web page and sends all elements on the web page to the LLM; the LLM determines the target operation object on the first target web page based on the elements, and generates an action instruction set executable by the RPA subsystem; finally, the RPA subsystem operates the target operation object according to the action instruction set to obtain the information to be obtained.
[0049] This method uses LLM to analyze natural language input from users to obtain the information to be retrieved, the corresponding webpage, and the characteristics of the information to be retrieved. This eliminates the need for users to write computer instructions, lowering the technical threshold. Furthermore, it analyzes the inherent logic of webpages to accurately generate corresponding action instructions for target operations, improving the accuracy of the method. Furthermore, this method utilizes an RPA subsystem, introducing a digital workforce to achieve efficient 24 / 7 operation while avoiding human error and improving information processing efficiency.
[0050] Next, the information processing method provided in the embodiments of the present application will be introduced with reference to the accompanying drawings.
[0051] See also Figure 1 The flowchart of the information processing method shown is applied to an information processing system that includes an RPA subsystem and LLM. RPA is a business process automation technology based on software robots and artificial intelligence. It can simulate manual operations through simple configuration without modifying the original system, helping companies or employees complete repetitive and monotonous procedural tasks. LLM is a deep learning model trained using large amounts of text data that supports semantic parsing, reasoning, and generation in both text and speech. The method specifically includes the following steps:
[0052] S102: The RPA subsystem receives instruction information input by the user.
[0053] The RPA subsystem can obtain information in response to the indication information input by the user. The indication information is natural language information used to indicate the information to be obtained, and the information to be obtained refers to text information obtained based on at least one information source. The user can input the indication information through any input device, for example, the input device can be at least one of a mouse, keyboard, stylus, finger, and microphone. It is understandable that the indication information can be text information directly input by the user, or it can be text information converted from the voice information input by the user through the automatic speech recognition (ASR) technology, or it can be text information converted from the picture information input by the user through the optical character recognition (OCR) technology.
[0054] In some possible implementations, the RPA subsystem can also store historical user input, allowing it to periodically acquire information at preset intervals. The RPA subsystem can automatically acquire information at preset intervals, such as every 10 minutes, an hour, a day, or a fixed time period each day. Of course, the RPA system can also acquire information at varying intervals. For example, when determining team-building options for a statistics department, it can be assumed that the majority of colleagues will select an option within a day of its release, allowing for shorter information acquisition intervals. Once the vote is nearly finalized, information can be acquired at longer intervals, allowing for more efficient acquisition of valuable information.
[0055] S104: The LLM determines characteristic information of the first target webpage and the information to be obtained according to the instruction information.
[0056] Instructions can be entered in natural language, and the LLM interprets and analyzes them to obtain the user's desired information and its characteristic information. The LLM can be deployed on any server or terminal with the required computing power and interact with the RPA subsystem through an interface. The RPA subsystem can input the collected instructions into the LLM, which then performs content recognition and extraction to obtain the characteristic information of the information to be obtained. The characteristic information of the information to be obtained can include at least one of the following: subject, publisher, keywords, or publication time. This allows the extraction of characteristic information of the information to be obtained, facilitating accurate search using this characteristic information.
[0057] For example, a user can input "I want to obtain the attendance records of Department A last month" through the input device. At this time, LLM can analyze the instruction information input by the user and obtain that the information to be obtained by the user is "attendance records", and the characteristic information of the information to be obtained is the release / recording time and the publisher / recorder, specifically "last month" and "Department A". For the release time, the calendar of the computing device itself can be combined to determine the accurate release time range.
[0058] Furthermore, the LLM can analyze the target webpage for the information the user needs to obtain. For example, if the information to be obtained is "attendance records," the LLM can further analyze the specific attendance page in the company's office automation (OA) system from which the information should be obtained, thereby determining that the first target webpage is the attendance page.
[0059] In some possible implementations, the LLM may store preset correspondences between web page keywords and web pages. Thus, when the information to be acquired, as analyzed by the LLM, matches the preset keywords, the LLM can directly determine the first target web page based on the preset correspondences between the preset keywords. For example, if the information processing system is used for daily office work at a specific enterprise, the LLM may pre-store keywords and corresponding web page URLs for different modules in the enterprise's OA system.
[0060] In some possible implementations, when the LLM does not store preset web page keywords and corresponding relationships, the LLM may also default the first target web page to a specific web page, such as a default jump search engine.
[0061] It should be noted that the embodiments of the present application do not impose any restrictions on the specific model structure and training method of LLM, that is, the LLM used in the embodiments of the present application is any artificial intelligence model that can realize text parsing and information extraction.
[0062] S106: The RPA subsystem opens the first target webpage and sends all elements on the first target webpage to the LLM.
[0063] All elements on the first target webpage include at least one of all text information or all image information on the first target webpage. In this way, the RPA subsystem can send all information on the first target webpage to the LLM, which analyzes and determines the specific logic of the webpage and instructs the RPA subsystem to perform subsequent operations.
[0064] S108: The LLM determines a target operation object on the first target webpage based on the elements sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem.
[0065] The LLM may analyze the logic of the first target webpage according to the elements sent by the RPA subsystem, and further determine the target operation object on the first target webpage.
[0066] In some possible implementations, the information to be obtained needs to be searched and obtained on the first target web page according to the characteristic information of the information to be obtained. For example, when the first target web page is the search page of an academic website, the information to be obtained is an academic paper, and the characteristic information of the information to be obtained is the paper topic and / or the publication time range, it is obvious that the RPA subsystem needs to enter the paper topic and / or the publication time range in the academic website for searching. At this time, LLM can determine that the target operation object is an input box in the element returned by the RPA subsystem, realize the positioning of the input box, and generate characteristic information input instructions, so that the RPA subsystem can enter the paper topic and / or the publication time range in the input box of the academic website according to the instructions.
[0067] In some possible implementations, the characteristic information of the information to be retrieved can match the text information of an index control on the first webpage. For example, a webpage link with a title matching the characteristic information appears on the first target webpage, or a selection control matching the characteristic information (such as a specific publication year) appears on the first target webpage. In this case, the LLM can determine that the target operation object is the index control whose text information matches the characteristic information, and generate a click instruction to cause the RPA subsystem to click the index control and load the second target webpage containing the information to be retrieved. The second target webpage can be different from the first target webpage. For example, the first target webpage can be a search page on an academic website, and the second target webpage can be a page specifically displaying academic papers. Alternatively, the second target webpage can be refreshed based on the first target webpage and the characteristic information. For example, the first target webpage can be the attendance page in an OA system, and the second target webpage can be the attendance page displayed after selecting a department and date.
[0068] In some possible implementations, the target operation object may also be at least one page-turning control. For example, if a feature search yields a large number of results, requiring multiple web pages to be displayed, the LLM can also generate click and traverse instructions to instruct the RPA subsystem to click the page-turning control and traverse each web page indicated by the page-turning control.
[0069] In some possible implementations, as the RPA subsystem traverses each webpage, it can also send all elements on each webpage to the LLM. Based on this, the LLM can also analyze the logic of each webpage and match it with the feature information of the information to be retrieved. If a match is successful, it will generate a new click instruction, causing the RPA subsystem to load a third target page containing the information to be retrieved.
[0070] It should be noted that the "action instruction set" and "action instructions" in the embodiments of the present application are computer instructions that can be executed directly or simply compiled by the RPA subsystem. That is, the embodiments of the present application do not specifically limit how the LLM writes instructions and the type of computer language used to write instructions. Relevant technical personnel training LLM should choose an appropriate method to determine the generation method of the instruction set based on actual conditions.
[0071] S110: The RPA subsystem operates the target operation object according to the action instruction set to obtain the information to be obtained.
[0072] The information to be processed may be directly acquired text information, or may be text information converted from video, audio, or images using Automatic Speech Recognition (ASR) technology or Optical Character Recognition (OCR) technology.
[0073] In some possible implementations, the RPA subsystem can also send the acquired information to the LLM, which then generates the target information based on the acquired information. For example, the LLM can perform format conversion, converting the information to PDF to reduce the possibility of information tampering. Another example is that the LLM can perform information integration, generating a list of all eligible information to be processed and saving it.
[0074] In some possible implementations, the RPA subsystem can also send the target information returned by the LLM to the user, for example, via email. The RPA subsystem can also save the target information and automatically send it to the user at a preset time interval, such as every 10 minutes, one hour, one day, or a fixed time period each day.
[0075] In some possible implementations, the RPA subsystem may also send target information to authorized specific users based on user range information when sending target information.
[0076] It should be noted that the embodiments of this application do not limit the specific hardware deployment relationship between the RPA subsystem and the LLM. That is, the RPA subsystem and LLM can run independently on different terminals or servers with network connectivity, or they can be deployed in whole or in part on a single terminal or server if conditions permit.
[0077] Based on the above description, the present application provides an information processing method. On the one hand, this method uses LLM to analyze the natural language information input by the user to obtain the information to be obtained, the corresponding web page, and the characteristic information of the information to be obtained, thereby eliminating the need for users to write computer instructions themselves and lowering the technical threshold. On the other hand, it can analyze the inherent logic of the web page to accurately generate corresponding action instructions for the target operation object, thereby improving the accuracy of the method. In addition, this method adopts the RPA subsystem and introduces a digital workforce to achieve efficient work around the clock and avoid human operational errors, thereby improving the efficiency of information processing.
[0078] The present application also provides an information processing system. The information processing system of the present application is described in detail below with reference to the accompanying drawings.
[0079] See also Figure 2 A structural diagram of an information processing system 200 is shown in FIG. Figure 2 As shown, the RPA system includes:
[0080] The RPA subsystem 202 is configured to receive instruction information input by a user, where the instruction information is natural language information indicating information to be obtained;
[0081] LLM 204, configured to determine the information to be acquired, the first target webpage, and characteristic information of the information to be acquired according to the instruction information sent by the RPA subsystem;
[0082] The RPA subsystem 202 is further configured to open the first target webpage and send all elements on the first target webpage to the LLM 204, where the elements include at least one of text information or image information of the first target webpage;
[0083] LLM 204 is further configured to determine a target operation object on the first target webpage based on the elements sent by the RPA subsystem 202 and generate an action instruction set executable by the RPA subsystem 202;
[0084] The RPA subsystem 202 is further configured to operate the target operation object according to the action instruction set to obtain the information to be obtained.
[0085] In some possible implementations, LLM 204 determines a target operation object on the first target webpage based on the elements sent by RPA subsystem 202 and generates an action instruction set executable by RPA subsystem 202, specifically for:
[0086] The target operation object is determined to be an input box, and a characteristic information input instruction is generated so that the RPA subsystem 202 inputs characteristic information of the information to be obtained in the input box.
[0087] In some possible implementations, LLM 204 determines a target operation object on the first target webpage based on the elements sent by RPA subsystem 202 and generates an action instruction set executable by RPA subsystem 202, specifically for:
[0088] The target operation object is determined to be an index control whose text information matches the feature information, and a click instruction is generated to cause the RPA subsystem 202 to click the index control to load a second target webpage including the information to be obtained.
[0089] In some possible implementations, LLM 204 determines a target operation object on the first target webpage based on the elements sent by RPA subsystem 202 and generates an action instruction set executable by RPA subsystem 202, specifically for:
[0090] Determine that the target operation object is at least one page turning control, generate a click instruction and a traversal instruction, so that the RPA subsystem 202 traverses each web page indicated by the at least one page turning control to obtain the information to be obtained.
[0091] In some possible implementations, when the LLM 204 determines the first target webpage based on the instruction information sent by the RPA subsystem 202, it is specifically configured to:
[0092] The information to be obtained is obtained according to the instruction information. When the information to be obtained matches the preset keywords, the first target webpage is determined according to the preset corresponding relationship of the preset keywords.
[0093] In some possible implementations, the RPA subsystem 202 is also used to send the information to be acquired to the LLM 204. At this time, the LLM 204 is also used to generate target information based on the information to be acquired and send the target information to the RPA subsystem 202. The RPA subsystem 202 is also used to save the target information and send the target information to the user at a preset period.
[0094] The present application provides a computer device for implementing an information processing method. The computer device includes a processor and a memory. The processor and the memory communicate with each other. The processor is configured to execute instructions stored in the memory, causing the computer device to perform the information processing method.
[0095] The present application provides a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium. When the computer-readable storage medium is executed on a computer device, the computer device executes the above-mentioned information processing method.
[0096] The present application provides a computer program product comprising instructions, which, when executed on a computer device, enables the computer device to execute the above-mentioned information processing method.
[0097] It should also be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided in this application, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines.
[0098] Through the above description of the embodiments, those skilled in the art will clearly understand that the present application can be implemented using software plus necessary general-purpose hardware. Of course, it can also be implemented using dedicated hardware, including application-specific integrated circuits, dedicated CPUs, dedicated memories, and dedicated components. Generally speaking, any function performed by a computer program can be easily implemented using corresponding hardware. Moreover, the specific hardware structures used to implement the same function can also be diverse, such as analog circuits, digital circuits, or dedicated circuits. However, for the present application, software program implementation is often the preferred implementation method. Based on this understanding, the technical solution of the present application, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored on a readable storage medium, such as a computer floppy disk, USB flash drive, removable hard disk, ROM, RAM, magnetic disk, or optical disk, and includes a number of instructions for causing a computer device (which can be a personal computer, training equipment, or network equipment, etc.) to execute the methods described in the various embodiments of the present application.
[0099] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.
[0100] The computer program product includes one or more computer instructions. When loaded and executed on a computer, the computer program instructions fully or partially perform the processes or functions described in the embodiments of the present application. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium capable of computer storage, or a data storage device such as a training device or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).
[0101] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. An information processing method, characterized in that: Applied to an information processing system, the information processing system including a robotic process automation (RPA) subsystem and a large language model (LLM), the method includes: The RPA subsystem receives instruction information input by a user, where the instruction information is natural language information used to indicate information to be obtained; The large language model determines the information to be acquired, the first target webpage, and feature information of the information to be acquired according to the instruction information sent by the RPA subsystem; The RPA subsystem opens the first target webpage and sends all elements on the first target webpage to the large language model, where the elements include at least one of text information or image information of the first target webpage; The large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem; The RPA subsystem operates the target operation object according to the action instruction set to obtain the information to be obtained.
2. The method according to claim 1, characterized in that The large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including: The large language model determines that the target operation object is an input box, and generates a feature information input instruction so that the RPA subsystem inputs the feature information of the information to be acquired into the input box.
3. The method according to claim 1, characterized in that The large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including: The large language model determines that the target operation object is an index control whose text information matches the feature information, and generates a click instruction to enable the RPA subsystem to click the index control to load a second target web page including the information to be obtained.
4. The method according to claim 1, wherein The large language model determines a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generates an action instruction set executable by the RPA subsystem, including: The large language model determines that the target operation object is at least one page turning control, and generates a click instruction and a traversal instruction to enable the RPA subsystem to traverse each web page indicated by the at least one page turning control to obtain the information to be obtained.
5. The method according to claim 1, characterized in that The large language model determines a first target webpage according to the instruction information sent by the RPA subsystem, including: The large language model obtains the information to be obtained according to the indication information, and when the information to be obtained matches a preset keyword, determines the first target webpage according to a preset correspondence relationship between the preset keywords.
6. The method according to claim 1, characterized in that The method further comprises: The RPA subsystem sends the information to be acquired to the large language model; The large language model generates target information according to the information to be acquired, and sends the target information to the RPA subsystem; The RPA subsystem stores the target information and sends the target information to the user at a preset period.
7. The method according to any one of claims 1 to 6, characterized in that The characteristic information includes at least one of the subject, publisher, keyword or publishing time of the information to be obtained.
8. An information processing system, characterized in that: The information processing system includes a robotic process automation (RPA) subsystem and a large language model (LLM); The RPA subsystem is configured to receive instruction information input by a user, where the instruction information is natural language information indicating information to be obtained; The large language model is used to determine the information to be acquired, the first target webpage, and feature information of the information to be acquired based on the instruction information sent by the RPA subsystem; The RPA subsystem is further configured to open the first target webpage and send all elements on the first target webpage to the large language model, where the elements include at least one of text information or image information of the first target webpage; The large language model is further used to determine a target operation object on the first target webpage based on the element sent by the RPA subsystem, and generate an action instruction set executable by the RPA subsystem; The RPA subsystem is further configured to operate the target operation object according to the action instruction set to obtain the information to be obtained.
9. A computer device, characterized in that: The computer equipment includes: Memory for storing computer programs or computer instructions; A processor, configured to execute the computer program or computer instructions stored in the memory, so that the computer device performs the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store a computer program, and when the computer program is executed, it is used to implement the method according to any one of claims 1 to 7.