Scheduling method and device for RPA service process

By obtaining text data from user voice instructions and generating scheduling expressions using large language models, the problem of lack of flexibility in traditional RPA scheduling strategies is solved, and a deep understanding and flexible response to business scenarios is achieved, and the execution efficiency of automated processes is improved. Accuracy, adapt to different corporate needs.

CN118941226APending Publication Date: 2024-11-12BEIJING HONGLIAN 95 INFORMATION IND CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410888558.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-03
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Traditional RPA scheduling strategies are based on fixed rule arrangements, lack in-depth understanding and flexible response to business scenarios, and are difficult to adapt to the specific needs of different companies.

Method used

By obtaining text data describing RPA business processes from the user's voice instructions, generating embedding vectors, and using a large language model to match similar vectors, generating scheduling expressions for business process scheduling, and scheduling combinations with business process information and scheduling expressions. Achieve deep understanding and flexible response to business scenarios.

Benefits of technology

It improves the execution efficiency and accuracy of automated processes, has good scalability and customizability, can adapt to the specific needs of different industries and enterprises, and brings significant economic and social benefits.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941226A_ABST
    Figure CN118941226A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of data processing, and particularly discloses a scheduling method and device for an RPA service process. The scheduling method is suitable for scheduling a periodically executed business process, and comprises the following steps: acquiring text data for describing an RPA business process from a voice instruction of a user; generating an embedded vector of the text data; a similar vector of the embedded vector is matched from a business process vector set, business process information corresponding to the similar vector is determined, and the business process information at least comprises a business name and process description information; in combination with the text data and the process description information, a scheduling strategy generation model is utilized to generate a scheduling expression, and the scheduling expression utilizes a fixed character string to represent time information for executing the business process; and performing business process scheduling based on the business process information and the scheduling expression. Based on the method and the device, the scheduling strategy better meeting the actual demand can be generated, and the execution efficiency and accuracy of the automatic process are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a scheduling method and device for an RPA business process. Background Art

[0002] As enterprises become more information-based, business process automation has become a key means to improve efficiency and reduce costs. However, traditional RPA (Robotic Process Automation) scheduling strategies are often based on fixed rules and lack in-depth understanding and flexible response to business scenarios.

[0003] Therefore, a new RPA business process scheduling solution is needed, which aims to improve the flexibility of business process scheduling by generating scheduling strategies in a semantic and intelligent way. Summary of the invention

[0004] The present application provides a method and device for scheduling an RPA business process, in an effort to solve or at least alleviate at least one of the above problems.

[0005] According to one aspect of the present application, a scheduling method for an RPA business process is provided, which is suitable for scheduling periodically executed business processes, including: obtaining text data describing the RPA business process from a user's voice command; generating an embedded vector for the text data; matching similar vectors of the embedded vector from a business process vector set, and determining business process information corresponding to the similar vector, wherein the business process information includes at least a business name and process description information; combining the text data and the process description information, using a scheduling strategy generation model, and generating a scheduling expression, wherein the scheduling expression uses a fixed character string to represent the time information for executing the business process; and scheduling the business process based on the business process information and the scheduling expression.

[0006] Optionally, the method according to the present application also includes constructing the business process vector set, including: obtaining multiple business process information, and for each business process information: converting the business process information into multiple text blocks; based on each text block, generating a business process vector corresponding to the business process information; and generating a business process vector set by associating the business process information with the business process vector.

[0007] Optionally, in the method according to the present application, the scheduling policy generation model is based on a large language model, and generates a scheduling expression by parsing descriptive information related to the scheduling policy.

[0008] Optionally, the method according to the present application further includes: constructing training samples to train a large language model to generate a scheduling policy parsing model.

[0009] Optionally, the method according to the present application also includes constructing training samples, including: determining multiple business process information based on real data of scheduling scenarios in the business system, wherein the scheduling scenarios are periodic scheduling scenarios; generating scheduling expressions for each business process information; processing the business process information and scheduling expressions according to a preset data format to generate training samples, wherein the preset data format includes a prompt part and a response part.

[0010] Optionally, in the method according to the present application, generating a scheduling expression for each business process information includes: generating an erroneous and non-compliant scheduling expression for part of the business process information as a negative sample.

[0011] Optionally, in the method according to the present application, matching a similar vector to the embedded vector from the business process vector set includes: using cosine similarity to determine the similarity between the embedded vector and the business process vector.

[0012] According to another aspect of the present application, a scheduling device for an RPA business process is provided, which is suitable for scheduling periodically executed business processes, and includes: a description information acquisition module, suitable for acquiring text data describing the RPA business process from a user's voice instruction; a text data vectorization module, suitable for generating an embedded vector of the text data, and determining the business process information corresponding to the similar vector by matching a similar vector of the embedded vector from a business process vector set, wherein the business process information includes at least a business name and process description information; an expression generation module, suitable for using a scheduling strategy generation model to process the text data and generate a scheduling expression; a scheduling execution module, suitable for scheduling the business process based on the business process information and the scheduling expression.

[0013] Optionally, the scheduling device according to the present application also includes: a training module, suitable for constructing training samples to train a large language model to generate a scheduling policy parsing model, wherein the scheduling policy parsing model generates a scheduling expression by parsing the descriptive information related to the scheduling policy, and the scheduling expression uses fixed fields to represent the time information of executing the business process.

[0014] According to another aspect of the present application, a computing device is provided, comprising: one or more processor memories; one or more programs, wherein the one or more programs are stored in the memories and configured to be executed by one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0015] According to another aspect of the present application, a computer-readable storage medium storing one or more programs is provided. The one or more programs include instructions. When the instructions are executed by a computing device, the computing device executes any of the methods described above.

[0016] According to another aspect of the present application, a computer program product is provided, comprising a computer program / instruction, wherein the computer program / instruction implements the steps of the above-mentioned method when executed by a processor.

[0017] In summary, according to the solution of this application, it is possible to achieve a deep understanding and flexible response to business scenarios, and the generated scheduling strategies are more in line with actual needs, improving the execution efficiency and accuracy of the automated process. At the same time, this solution also has good scalability and customizability, and can adapt to the specific needs of different industries and enterprises. Whether it is financial services, customer service or other business processes, the solution of this application can be used to perform automated business process scheduling, bringing significant economic and social benefits to the enterprise.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to achieve the above and related purposes, certain illustrative aspects are described herein in conjunction with the following description and accompanying drawings, which indicate various ways in which the principles applied herein can be practiced, and all aspects and their equivalents are intended to fall within the scope of the claimed subject matter. By reading the following detailed description in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. Throughout the application, the same reference numerals generally refer to the same parts or elements.

[0020] Figure 1 A schematic diagram of a scheduling device 100 for an RPA business process according to some embodiments of the present application is shown;

[0021] Figure 2 A schematic diagram of a computing device 200 according to some embodiments of the present application is shown;

[0022] Figure 3 A flow chart of a method 300 for scheduling an RPA business process according to some embodiments of the present application is shown;

[0023] Figure 4 A flow chart of a method for scheduling an RPA business process according to other embodiments of the present application is shown. DETAILED DESCRIPTION

[0024] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0025] This application introduces a large language model (LLM) and uses its deep understanding and generation capabilities of natural language to achieve semantic parsing of specific business processes, thereby generating a scheduling strategy that conforms to business logic to schedule business processes. According to the implementation method of this application, a specific business process refers to a business process that is executed periodically, such as: "verify invoices every Friday", "execute once every half hour", etc.

[0026] Figure 1 A schematic diagram of a scheduling device 100 for an RPA business process according to some embodiments of the present application is shown. The scheduling device 100 can be arranged in a specific business process processing system (such as an RPA digital employee platform) to schedule periodically executed business processes. In some embodiments, the scheduling device 100 can also communicate with the user through a voice processing module (not shown) to obtain the user's voice instructions. The user's voice instructions will include description information of the RPA business process, and the description information of the business process may include, for example, the business name, execution cycle, other process information, etc.

[0027] like Figure 1 As shown, the scheduling device 100 includes: a description information acquisition module 110, a text data vectorization module 120, an expression generation module 130 and a scheduling execution module 140.

[0028] The description information acquisition module 110 first obtains text data describing the RPA business process from the user's voice command. The user's voice command can be a single voice command or a voice command extracted from multiple rounds of conversations. The embodiment of the present application does not limit the specific algorithm used to convert voice data into text data. The following shows a conversation:

[0029] User: I need a process for batch verification of invoices.

[0030] Robot: OK, do you want to execute it now?

[0031] User: No.

[0032] Robot: What time will it be executed?

[0033] User: Execute at 8 am every day.

[0034] From the above conversation, the description information acquisition module 110 obtains text data describing the RPA business process: the batch invoice verification process is executed at 8 am every day.

[0035] Afterwards, the text data vectorization module 120 generates an embedding vector for the text data. In some embodiments, an embedding vector is generated using an Embeddings vector model, such as acge_text_embedding, baichuan-text-embedding, BAAI / bge-m3, and the like.

[0036] Afterwards, the text data vectorization module 120 determines the business process information corresponding to the similar vector by matching the similar vector of the embedded vector from the business process vector set. In some embodiments, the business process information includes at least a business name and process description information, and the process description information is usually a description of the execution cycle.

[0037] On the other hand, the expression generation module 130 uses the scheduling strategy generation model to process the text data and generate a scheduling expression. The scheduling expression uses a fixed field to represent the time information of executing the business process, and the time information is usually periodic. In some embodiments, the fixed field is a string of fixed length.

[0038] In some embodiments, the scheduling device 100 further includes a training module 150 for constructing training samples to train the large language model to generate a scheduling policy parsing model, wherein the scheduling policy parsing model generates a scheduling expression by parsing the descriptive information related to the scheduling policy. The scheduling policy generation model will be described in detail below and will not be elaborated here.

[0039] Afterwards, the scheduling execution module 140 performs business process scheduling based on the business process information and the scheduling expression.

[0040] Based on the scheduling device 100 of the present application, based on the user's voice instructions, first undergo vectorization processing and vector matching to determine the business process information that the user wants to execute. Afterwards, based on the business process information and the text data obtained from the voice instructions, the scheduling strategy generation model generated by training is used to determine the scheduling expression. Finally, a scheduling strategy is generated according to the business process information and the scheduling expression to perform business process scheduling. The scheduling device 100 can be used in business scenarios that are executed periodically to achieve a deep understanding of and flexible response to business scenarios. The generated scheduling strategy is more in line with actual needs, improving the execution efficiency and accuracy of automated processes.

[0041] The scheduling device 100 of the present application can be implemented by one or more computing devices. The computing device can be any device with storage and computing capabilities, such as a server, a workstation, etc., or a personal computer such as a desktop computer, a notebook computer, or a terminal device such as a mobile phone, a tablet computer, a smart wearable device, an Internet of Things device, etc., but is not limited thereto.

[0042] Figure 2 FIG. 2 shows a schematic diagram of a computing device 200 according to some embodiments of the present application. It should be noted that: Figure 2 The computing device 200 shown is only an example. In practice, the computing device used to implement the method of the present application can be any type of device, and its hardware configuration can be different from the above. Figure 2 The computing device 200 shown is the same as Figure 2 The computing device 200 shown is different. In practice, the computing device used to implement the embodiment of the present application can be Figure 2 The hardware components of the computing device 200 shown are increased or deleted, and the present application does not limit the specific hardware configuration of the computing device.

[0043] like Figure 2 In a basic configuration, computing device 200 includes at least one processing unit 202 and system memory 204. According to one aspect, depending on the configuration and type of computing device, processing unit 202 can be implemented as a processor. System memory 204 includes, but is not limited to, volatile storage (e.g., random access memory), non-volatile storage (e.g., read-only memory), flash memory, or any combination of such memories. According to one aspect, system memory 204 includes operating system 205 and program modules 206.

[0044] According to one aspect, operating system 205 is suitable for controlling the operation of computing device 200, for example. Furthermore, examples are practiced in conjunction with graphics libraries, other operating systems, or any other application programs, and are not limited to any particular application or system. Figure 2 This basic configuration is illustrated in FIG. 2 by those components within dashed line 208. According to one aspect, computing device 200 has additional features or functionality. For example, according to one aspect, computing device 200 includes additional data storage devices (removable and / or non-removable), such as magnetic disks, optical disks, or tapes. Such additional storage Figure 2 2 is illustrated by removable storage 209 and non-removable storage 210.

[0045] As stated above, according to one aspect, program modules are stored in the system memory 204. According to one aspect, the program modules may include one or more application programs, and the present application does not limit the types of application programs, for example, the application programs may include: email and contact applications, word processing applications, spreadsheet applications, database applications, slide show applications, drawing or computer-aided applications, web browser applications, etc.

[0046] According to one aspect, the examples may be practiced on a circuit comprising discrete electronic components, a packaged or integrated electronic chip containing logic gates, a circuit utilizing a microprocessor, or a single chip containing electronic components or a microprocessor. Figure 2 Each or many components shown in can be integrated in a system on chip (SOC) on a single integrated circuit to practice examples. According to one aspect, such a SOC device may include one or more processing units, a graphics unit, a communication unit, a system virtualization unit, and various application functions, all of which are integrated (or "burned") into a chip substrate as a single integrated circuit. When operated via SOC, the functions described in the present application can be operated via a dedicated logic integrated with other components of a computing device 200 on a single integrated circuit (chip). Other technologies that can perform logical operations (such as AND, OR, and NOT) can also be used to practice embodiments of the present application, including but not limited to mechanical, optical, fluid, and quantum technologies. In addition, embodiments of the present application can be practiced in a general-purpose computer or in any other circuit or system.

[0047] According to one aspect, the computing device 200 may also have one or more input devices 212, such as a keyboard, a mouse, a pen, a voice input device, a touch input device, a VR motion capture input device, etc. Output devices 214 may also be included, such as a display, a speaker, a printer, etc. The aforementioned devices are examples and other devices may also be used. The computing device 200 may include one or more communication connections 216 that allow communication with other computing devices 218. Examples of suitable communication connections 216 include, but are not limited to: RF transmitters, receivers and / or transceiver circuits; Universal Serial Bus (USB), parallel and / or serial ports.

[0048] The term computer-readable media as used in this application includes computer storage media. Computer storage media may include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (e.g., computer-readable instructions, data structures, or program modules). System memory 204, removable storage 209, and non-removable storage 210 are all examples of computer storage media (i.e., memory storage). Computer storage media may include random access memory (RAM), read-only memory (ROM), electrically erasable read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical storage, cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other products that can be used to store information and can be accessed by computing device 200. According to one aspect, any such computer storage media may be part of computing device 200. Computer storage media do not include carrier waves or other propagated data signals.

[0049] According to one aspect, communication media is implemented by computer readable instructions, data structures, program modules, or other data in a modulated data signal (e.g., a carrier wave or other transport mechanism), and includes any information delivery media. According to one aspect, the term "modulated data signal" describes a signal that has one or more characteristics set or changed in a manner that encodes information in the signal. By way of example and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, radio frequency (RF), infrared, and other wireless media.

[0050] In an embodiment according to the present application, the computing device 200 is configured to execute the scheduling method 300 of the RPA business process according to the present application. The program module 206 includes a plurality of program instructions for executing the method 300, and these program instructions can instruct the processing unit 202 to execute the method 300 of the present application.

[0051] Figure 3 A flow chart of a method 300 for scheduling an RPA business process according to some embodiments of the present application is shown.

[0052] According to some embodiments, the RPA business process scheduling method 300 is suitable for execution in the scheduling device 100 as described above. For a periodically executed business process, on the one hand, the embedded vector is used to retrieve matching business process information from the business process vector set; on the other hand, the scheduling strategy generation model is used to generate a scheduling expression. In this way, by combining the business process information with the scheduling expression, the business process can be automatically scheduled. It should be understood that the description of this method 300 and the scheduling device 100 complement each other, and the relevant parts will not be repeated.

[0053] like Figure 3 As shown, the method 300 starts at step S310 .

[0054] In S310 , text data describing the RPA business process is obtained from the user's voice instruction.

[0055] As mentioned above, this application does not impose too many restrictions on the user's voice instructions. The description information acquisition module 110 can determine the user's intention indicated in the voice instruction through natural language understanding (NLU), and the user's intention contains information describing the RPA business process. At the same time, it is processed into text data (i.e., the description text of the business process) for subsequent processing.

[0056] The text data does not need to describe the business process in detail, but only needs to include some key description information, such as a simple description of the business and the execution process, where the business can be represented by the business name and the execution process can be represented by the process ID. In some embodiments, the scheduling device 100 identifies various processes involved in the business scenario by the process ID.

[0057] Then in S320 , an embedding vector of the text data is generated.

[0058] According to an embodiment of the present application, text data is vectorized so that structured data can be used to manage the description text of the business process. Specifically, based on the Embedding algorithm, an embedding vector of the text data is generated. This embodiment does not limit the specific Embedding vector model used, such as acge_text_embeding, baichuan-text-embeding, BAAI / bge-m3.

[0059] For example, the text data is: batch verification of invoices, execution process 101. Using Embedding, the generated embedding vector is 0101010101.

[0060] Then, in S330, a similar vector of the embedded vector is matched from the business process vector set, and the business process information corresponding to the similar vector is determined. According to the implementation of the present application, the business process information at least includes a business name and process description information.

[0061] The method 300 further includes constructing a business process vector set. Specifically, the process of constructing a business process vector set includes the following two steps.

[0062] First, obtain multiple business process information, and for each business process information: first convert the business process information into multiple text blocks; then generate a business process vector corresponding to the business process information based on each text block. Refer to the relevant description in S320, and generate the business process vector through the Embedding vector model, which will not be repeated here. According to the implementation mode of the present application, the business process information is a detailed description of the business process, and the text data obtained from the user's voice command can be a simple expression of the business process.

[0063] Afterwards, a business process vector set is generated by associating the business process information with the corresponding business process vector.

[0064] In this way, by calculating the similarity between the embedded vector and each business process vector in S320, the business process vector with the highest similarity is used as the matching vector of the embedded vector, and the business process information associated with the matched business process vector can be used as the business process information of the embedded vector. In some embodiments, the cosine similarity is used to determine the similarity between the embedded vector and the business process vector.

[0065] In S340, the text data and the process description information are combined and the scheduling strategy generation model is used to generate a scheduling expression.

[0066] Among them, the scheduling expression uses a fixed string to represent the time information of executing the business process. In some embodiments, the scheduling expression is implemented through Crontab. Crontab is a tool for setting periodically executed tasks in Unix and Unix-like operating systems (such as Linux). Through the Crontab command, a specified system instruction or shell script can be executed at a fixed interval. The unit of the time interval can be any combination of minutes, hours, days, months, weeks and above. The scheduling expression is usually expressed in the form of "*****", where each "*" represents a time interval. Table 1 lists the time intervals and value ranges indicated by each "*" in order from left to right. It should be noted that Table 1 is only an exemplary description of the scheduling expression. In actual applications, English abbreviations can also be used to represent months, weeks, etc., and this application does not limit this.

[0067] Table 1

[0068] Time interval Value range * minute 0-59 * Hour 0-23 * date 1-31 * month 1-12 * Week 0-6

[0069] According to the implementation mode of the present application, the scheduling strategy generation model is based on a large language model (LLM), and generates a scheduling expression by parsing the descriptive information related to the scheduling strategy. According to some embodiments, the descriptive information related to the scheduling strategy is, for example, text data obtained from a user's voice instruction, or it can be process description information determined by S330, or a combination of the two, and the present application does not impose too many restrictions on this. In some preferred embodiments, the text data and process description information can be combined to obtain relatively complete process description information.

[0070] The method 300 further includes: constructing training samples to train the large language model to generate a scheduling policy parsing model.

[0071] The process of constructing training samples includes the following three steps.

[0072] In the first step, based on the real data of the scheduling scenario in the business system, multiple business process information is determined, wherein the scheduling scenario is a periodic scheduling scenario.

[0073] Specifically, based on the real data of scheduling scenarios in the business system, periodic business processes are collected and business process information is generated. Among them, periodic business processes contain multiple time intervals. For example, common Cron expression usage scenarios are collected in multiple ways, and various time interval combinations are covered.

[0074] The second step is to generate a scheduling expression for each business process information. The relevant content of the scheduling expression can refer to the relevant description content of Table 1 above.

[0075] At the same time, in order to enhance the generalization ability of the model, consider adding incorrect and non-compliant scheduling expressions to learn and distinguish correct and incorrect scheduling expressions during model training. Specifically, for most or all of the business process information, generate correct scheduling expressions as positive samples; for part of the business process information, generate incorrect and non-compliant scheduling expressions as negative samples. Table 2 shows some sample data.

[0076] Table 2

[0077]

[0078]

[0079] The third step is to process the business process information and scheduling expressions according to the preset data format to generate training samples.

[0080] Since the real data obtained comes from tables, databases, text files, etc., it is necessary to unify them according to the preset data format. The preset data format includes a prompt part and a response part. The following is an example, which contains 3 business process information and scheduling expressions. The specific example is as follows:

[0081] [{"prompt":"Based on the following description, generate a Cron expression as follows: execute every two weeks","response":"0 0* / 14**"},{"prompt":"Based on the following description, generate a Cron expression as follows: execute at 00:00 every morning","response":"0 0**?"},{"prompt":"Based on the following description, generate a Cron expression as follows: execute every 5 minutes","response":"* / 5****"}].

[0082] Among them, the "prompt" part is the prompt part, and the "response" part is the response part. Through this prompt+response method, the previously collected data is uniformly processed and stored in jsonl format as training samples.

[0083] At this point, a training sample for generating a scheduling policy parsing model is obtained. The embodiments of the present application do not limit the specific method used for model training. In some embodiments, ERNIE_Lite-8K is selected as the basic model (i.e., the large language model described above), and the SFT (Supervised Fine-Tuning) supervised fine-tuning method is used to train the full data set.

[0084] At the same time, in some embodiments, multiple rounds of parameter configuration tuning and training sample supplementation are performed, and model training and evaluation are repeated many times, and finally a relatively satisfactory result is achieved, and the corresponding model at this time is used as the final generated scheduling strategy analysis model.

[0085] Specifically, the training samples are expanded twice to add data that did not appear in the training process as validation set data for model evaluation. The model is evaluated using the validation set to check its performance in generating effective scheduling expressions. The model is tuned according to the evaluation results, including adjusting the model architecture, hyperparameters, training strategies, etc. The model evaluation is carried out using a manual evaluation model, which scores the model responses from different evaluation dimensions based on the subjective opinions and experience of human experts to evaluate the effectiveness of the model responses. The following three indicators are used for model evaluation: Average score: the sum of all evaluation dimension scores / the number of data volume evaluation dimensions, with a value of [0,100]; Goodcase ratio: the number of all evaluation dimensions with a score of 2 / the number of data volume evaluation dimensions, with a value of [0,1]; Satisfaction: manual evaluation, with a value of [0,5], where 0-bad, 1-defective, 2-available, 3-qualified, 4-good, and 5-satisfactory.

[0086] Table 3 shows some of the results after multiple rounds of tuning, training, and model evaluation.

[0087] Table 3

[0088]

[0089] Table 4 shows one of the training parameter configuration tuning use cases.

[0090] Table 4

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] Afterwards, in S350 , the business process is scheduled based on the business process information and the scheduling expression.

[0097] According to an embodiment of the present application, a scheduling strategy is generated based on the results of S330 and S340. The scheduling strategy is usually composed of "business process information + scheduling expression". The scheduling device 100 performs business process scheduling according to the scheduling strategy.

[0098] According to the scheduling method 300 of the present application, it is possible to achieve a deep understanding of the business scenario and flexible response, and the generated scheduling strategy is more in line with actual needs, improving the execution efficiency and accuracy of the automated process. At the same time, this solution also has good scalability and customizability, and can adapt to the specific needs of different industries and enterprises. Whether it is financial services, customer service or other business processes, the method 300 of the present application can be used to perform automated business process scheduling, bringing significant economic and social benefits to the enterprise.

[0099] Figure 4 The flowchart of the scheduling method of the RPA business process according to some other embodiments of the present application is shown. Among them, three types of arrows are used to represent the "vectorization" process, the "vector matching (i.e., vector retrieval)" process, and the "using the model to generate scheduling expressions and scheduling strategies (and, large model generation)" process. Figure 4 As shown in the figure, the user initiates a conversation and obtains the descriptive text related to the business process (for example, the process of batch verification of invoices at 8 am every day) as input information. The input information is submitted to the vector search program, and the corresponding business process information is obtained through Embedding → Similarity Search → Return Related Data; the input information and business process information are submitted to the large language model together to obtain the specific scheduling expression, and combined with the business process information, the business process scheduling strategy, that is, the answer, is output in the agreed output format.

[0100] In addition, for the vectorization program, the database of business process information is first loaded to convert the business process information into text content (Text). Usually, the business process information contains the business name, description, process information, etc. Then, the text content is divided into blocks (i.e., multiple chunks are obtained), and the block text is embedded to generate the business process vector and store it in the vector database (i.e., the business process vector set). Text vectorization uses the Embeddings vector model, such as BAAI / bge-m3.

[0101] In addition, embodiments of the present invention further include: A11. A computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by a computing device, the computing device executes a method as described in any one of A1-A7. A12. A computer program product, including a computer program / instruction, wherein the computer program / instruction implements any one of A1-A7 when executed by a processor.

[0102] The various techniques described herein may be implemented in combination with hardware or software, or a combination thereof. Thus, the method and apparatus of the present application, or certain aspects or portions of the method and apparatus of the present application may be in the form of program codes (i.e., instructions) embedded in a tangible medium, such as a removable hard disk, a USB flash drive, a floppy disk, a CD-ROM, or any other machine-readable storage medium, wherein when the program is loaded into a machine such as a computer and executed by the machine, the machine becomes a device for practicing the present application.

[0103] In the case where the program code is executed on a programmable computer, the computing device generally includes a processor, a storage medium readable by the processor (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device. The memory is configured to store the program code; the processor is configured to execute the method of the present application according to the instructions in the program code stored in the memory.

[0104] By way of example and not limitation, readable media include readable storage media and communication media. Readable storage media stores information such as computer readable instructions, data structures, program modules or other data. Communication media generally embody computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and include any information transfer medium. Any combination of the above is also included within the scope of readable media.

[0105] In the description provided herein, algorithms and displays are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the examples of the present application. According to the above description, it is obvious to construct the structure required for such systems. In addition, the present application is not directed to any specific programming language either. It should be understood that the content of the present application described herein can be realized by various programming languages, and the description of the specific language above is to disclose the preferred implementation of the present application.

[0106] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0107] Those skilled in the art will appreciate that the modules or units or components of the devices in the examples disclosed herein may be arranged in the devices described in the embodiment, or alternatively may be located in one or more devices different from the devices in the examples. The modules in the foregoing examples may be combined into one module or may be divided into multiple submodules.

[0108] Those skilled in the art will appreciate that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition they can be divided into multiple submodules or subunits or subassemblies. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification and all processes or units of any method or device disclosed in this manner can be combined in any combination. Unless otherwise clearly stated, each feature disclosed in this specification can be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0109] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present application and to form different embodiments.

[0110] In addition, some of the embodiments are described herein as methods or combinations of method elements that can be implemented by a processor of a computer system or by other devices that perform the functions. Therefore, a processor with necessary instructions for implementing the method or method elements forms a device for implementing the method or method elements. In addition, the elements described herein of the device embodiments are examples of devices that are used to implement the functions performed by the elements for the purpose of implementing the disclosure.

[0111] As used herein, unless otherwise specified, the use of ordinal numbers "first," "second," "third," etc. to describe common objects merely refers to different instances of similar objects and is not intended to imply that the objects so described must have a given order in time, space, order, or in any other manner. In addition, the quantifier "plurality" means "two" and / or "more than two."

[0112] Although the present application is described according to a limited number of embodiments, it will be appreciated by those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the present application thus described. In addition, it should be noted that the language used in this specification is selected primarily for the purpose of readability and teaching, rather than for the purpose of explaining or limiting the subject matter of the present application. Therefore, many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the appended claims. For the scope of the present application, the disclosure made to the present application is illustrative, not restrictive, and the scope of the present application is defined by the appended claims.

Claims

1. A method for scheduling an RPA business process, the method being suitable for scheduling a periodically executed business process, the method comprising: Obtain text data describing the RPA business process from the user's voice command; Generate an embedding vector for the text data; Matching a similar vector of the embedded vector from the business process vector set, and determining business process information corresponding to the similar vector, wherein the business process information at least includes a business name and process description information; Combining the text data and the process description information, using a scheduling strategy generation model, generating a scheduling expression, wherein the scheduling expression uses a fixed character string to represent time information for executing a business process; Business process scheduling is performed based on business process information and scheduling expressions.

2. The method according to claim 1, further comprising constructing the business process vector set, comprising: Obtain multiple business process information, and for each of the business process information: Converting the business process information into a plurality of text blocks; Based on each text block, generating a business process vector corresponding to the business process information; A business process vector set is generated by associating the business process information with the business process vector.

3. The method according to claim 1 or 2, wherein: The scheduling strategy generation model is based on a large language model and generates a scheduling expression by parsing descriptive information related to the scheduling strategy.

4. The method according to any one of claims 1 to 3, further comprising: Construct training samples to train the large language model to generate a scheduling policy parsing model.

5. The method of claim 4, wherein: It also includes building training samples, including: Based on real data of scheduling scenarios in the business system, multiple business process information is determined, wherein the scheduling scenario is a periodic scheduling scenario; Generate a scheduling expression for each of the business process information; The business process information and the scheduling expression are processed according to a preset data format to generate a training sample, wherein the preset data format includes a prompt part and a response part.

6. The method of claim 5, wherein: The generating of the scheduling expression of each of the business process information comprises: For some business process information, incorrect and non-compliant scheduling expressions are generated as negative samples.

7. The method of claim 2, wherein: The similar vector matched to the embedded vector from the business process vector set includes: The cosine similarity is used to determine the similarity between the embedding vector and the business process vector.

8. A RPA business process scheduling device, suitable for scheduling a periodically executed business process, comprising: A description information acquisition module, adapted to acquire text data describing the RPA business process from the user's voice command; A text data vectorization module, adapted to generate an embedding vector of the text data, and determine the business process information corresponding to the similar vector by matching a similar vector to the embedded vector from a business process vector set, wherein the business process information at least includes a business name and process description information; An expression generation module, adapted to utilize a scheduling strategy generation model to process the text data and generate a scheduling expression; The scheduling execution module is suitable for scheduling business processes based on business process information and scheduling expressions.

9. The scheduling device according to claim 8, further comprising: A training module is suitable for constructing training samples to train a large language model to generate a scheduling policy parsing model, wherein the scheduling policy parsing model generates a scheduling expression by parsing descriptive information related to the scheduling policy, and the scheduling expression uses fixed fields to represent the time information of executing the business process.

10. A computing device comprising: one or more processors; Memory; One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for executing the method according to any one of claims 1 to 7.