Multi-dimensional service control method based on artificial intelligence technology

Abstract

The invention discloses a multi-dimensional business control method based on an artificial intelligence technology, and the method comprises the steps: constructing an automatic business process through an RPA technology, and enabling the automatic business process to be used for processing a to-be-processed business. And when the original data of the to-be-processed service is the to-be-identified image, processing the original data in combination with an OCR technology. And when the original data are characters expressed by using a human natural language, processing the to-be-processed service in combination with an NLP technology. And establishing a business rule engine and a knowledge base, and processing the to-be-processed business in combination with the business rule engine and the knowledge base. According to the method, related business data can be efficiently collected and analyzed, a large amount of repeated, tedious, highly standardized and regular business work is automatically processed, the burden of employees is relieved, and the overall digitization and intelligentization of the business processing flow are improved from multiple dimensions.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a multi-dimensional business control method based on artificial intelligence technology. Background Art

[0002] RPA (Robotic Process Automation) technology, that is, robotic process automation technology, is a technology that uses software robots or artificial intelligence assistants to simulate and execute repetitive, rule-defined and large-scale business process operations performed by humans in a digital environment.

[0003] The robots used in RPA technology are generally software robots that can handle a large number of repetitive processes with certain logical rules. As a new generation of digital labor, RPA robots can imitate human operation methods, break down data barriers between systems, and automatically execute a series of business processes. Compared with manual operations, the speed of processing business processes is faster, which greatly improves work efficiency.

[0004] RPA technology has the following characteristics and development trends:

[0005] (1) The scope of automation is constantly expanding: As an automation technology, RPA technology is widely used in logistics management, supply chain management, customer service and other fields in addition to the common financial management and human resources automation. Moreover, with the continuous development of RPA technology, its automation scope is constantly expanding.

[0006] (2) Diversified application scenarios: RPA technology can be used for various types of automated business processes, including complex processes and cross-system processes.

[0007] (3) Low-code / no-code platforms: To lower the technical threshold, RPA vendors have launched low-code or no-code platforms that enable people with non-technical backgrounds to create and manage automated processes.

[0008] (4) Human-machine collaboration: RPA technology is not a technology that replaces human work, but a technology that collaborates with human work. In many cases, RPA technology is mainly used to handle a large number of tedious and repetitive tasks, allowing human employees to focus on more valuable work, such as innovation and decision-making.

[0009] In today's power grid industry, the overall management of business data is relatively independent and scattered, and there is no unified data storage management platform. In the process of collecting and processing business data, there are problems such as inefficient data analysis, low utilization, and repeated collection.

[0010] At the same time, due to the single means of business data collection and low degree of automation, there are a lot of repetitive, cumbersome, highly standardized, and rule-defined transactional tasks in the process of business data collection and processing that need to be manually processed by employees, such as the inspection plan preparation process, the missing process of the bottom of the line loss table in the same period, and the entry of ledger information. For employees, these large amounts of repetitive and cumbersome transactional tasks require a lot of time and energy to handle, resulting in a large number of front-line employees being unable to focus their main energy on high-value-added innovation work that is conducive to business development, such as core work such as safe production and high-quality services.

[0011] Therefore, in order to solve the above problems, combining RPA technology with other artificial intelligence technologies to improve the overall digitization and intelligence of business processing processes from multiple dimensions is of great significance to the current power grid industry. Summary of the invention

[0012] The purpose of the present invention is to provide a multi-dimensional business control method based on artificial intelligence technology, which is used to efficiently collect and analyze relevant business data, and to automate a large number of repetitive, tedious, highly standardized, and rule-clear business tasks, thereby reducing the burden on employees and improving the overall digitization and intelligence of business processing flows from multiple dimensions.

[0013] In order to achieve the above objectives, the present invention is implemented by the following technical solutions:

[0014] A multi-dimensional business control method based on artificial intelligence technology, comprising:

[0015] Build automated business processing flows through RPA technology to handle pending business;

[0016] When the original data of the service to be processed is an image to be recognized, the image to be recognized is processed in combination with OCR (Optical Character Recognition) technology to convert the image to be recognized into text information;

[0017] When the original data is text expressed in human natural language, the business to be processed is processed in combination with NLP (Natural Language Processing) technology;

[0018] A business rule engine and a knowledge base are established, and the business to be processed is processed in combination with the business rule engine and the knowledge base.

[0019] Optionally, the processing of the image to be recognized by combining OCR technology comprises the following steps:

[0020] Scanning the image to be recognized, and determining the shape of the characters in the image to be recognized by detecting the dark and bright areas of the scan result;

[0021] According to the shape of the characters, the characters and layout information are obtained, and the unstructured data in the characters and layout information are converted into structured data to form a template file to be identified;

[0022] A requirement document is generated according to user requirements, and the template file to be identified is searched according to the requirement document to obtain a file path, thereby obtaining an object to be identified.

[0023] Optionally, in combination with ICR (Intelligent Character Recognition) technology, a neural network model for identifying and transforming the object to be identified is formed through self-learning of the ICR system.

[0024] Optionally, the ICR system is configured with a plurality of different recognition engines, and the recognition results are mutually verified by the plurality of different recognition engines;

[0025] The multiple different recognition engines are each given selective voting rights.

[0026] Optionally, the image to be identified is scanned by a unified high-definition camera or scanner.

[0027] Optionally, the processing of the to-be-processed business in combination with NLP technology includes:

[0028] The original data is understood through NLP technology, and the processing results of the business to be processed are converted into text expressed in human natural language.

[0029] Optionally, the text information obtained by converting the image to be recognized is optimized in terms of text logic through NLP technology.

[0030] Optionally, the business rule engine supports real-time updating and execution of business rules, and allows the business rules to be defined in a visual manner.

[0031] Optionally, the business rule engine is decoupled from the automated business processing flow.

[0032] Optionally, in combination with the knowledge base, credibility analysis is performed to determine whether the business to be processed requires manual assistance.

[0033] The present invention has at least one of the following technical effects:

[0034] By combining RPA technology with OCR technology, relevant business data can be efficiently collected and analyzed, so that RPA robots can be responsible for the recognition and processing of work order scans or image business, effectively improving work efficiency and reducing the burden on employees.

[0035] By combining RPA technology with NLP technology, RPA robots have the ability to understand text expressed in human natural language and the ability to feedback processing information to users in human natural language after completing business operations, making them more intelligent.

[0036] By combining RPA technology with decision-making technology, a business rule engine and knowledge base are established to enable RPA robots to make more intelligent judgments and decisions, further improving the user experience.

[0037] Through the organic combination of RPA technology, OCR technology, NLP technology and decision-making judgment technology, the overall digitization and intelligence of business processing procedures have been improved from multiple dimensions. By making full use of data support technology, RPA robots can complete a large amount of repetitive, tedious, highly standardized and rule-clear business tasks, allowing front-line employees to focus more on high-value-added innovation work that is conducive to business development, which is beneficial to the long-term development of the industry and has significant beneficial effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A flowchart of a multi-dimensional service control method based on artificial intelligence technology provided by an embodiment of the present invention;

[0039] Figure 2 for Figure 1 Schematic diagram of the specific steps of combining OCR technology to process the image to be recognized. DETAILED DESCRIPTION

[0040] The following is a further detailed description of a satellite-borne radar measurement data timestamp synchronization method and system proposed by the present invention in conjunction with the accompanying drawings and specific embodiments. According to the following description, the advantages and features of the present invention will be clearer. It should be noted that the accompanying drawings are in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. In order to make the purposes, features and advantages of the present invention more obvious and easy to understand, please refer to the accompanying drawings. It should be noted that the structures, proportions, sizes, etc. illustrated in the accompanying drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention, so they have no technical substantive significance. Any structural modification, change in proportional relationship or adjustment of size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope of the technical content disclosed by the present invention.

[0041] like Figure 1As shown, the present invention provides a multi-dimensional business control method based on artificial intelligence technology, including: constructing an automated business processing flow through RPA technology to process the business to be processed; when the original data of the business to be processed is an image to be recognized, combining OCR technology to process the image to be recognized to convert the image to be recognized into text information; when the original data is text expressed in human natural language, combining NLP technology to process the business to be processed. Establish a business rule engine and a knowledge base, and combine the business rule engine and the knowledge base to process the business to be processed.

[0042] The present invention combines RPA technology with OCR technology to efficiently collect and analyze relevant business data, so that RPA robots can undertake the recognition and processing of work order scans or image services, effectively improve work efficiency, and reduce the burden on employees. By combining RPA technology with NLP technology, RPA robots have the ability to understand text expressed in human natural language and the ability to feed back processing information to users in human natural language after completing the business, with a higher degree of intelligence. By combining RPA technology with decision-making judgment technology, a business rule engine and knowledge base are established to achieve more intelligent judgment and decision-making of RPA robots, further improving the user experience. By combining RPA technology with a variety of artificial intelligence technologies, the present invention enables RPA robots to undertake a large number of daily businesses, improving the overall digitization and intelligence of business processing flows from multiple dimensions.

[0043] Specifically, in order to build an automated business process, it is necessary to model and program the business process, establish a flowchart or process model, automate the business process, and further execute the business process through the RPA robot. The control software of the RPA robot has a large number of components built in. The modeling and programming process does not require writing complex codes. Usually, you only need to drag, adjust, and configure various components according to the flowchart according to the actual situation. This method is simple and easy, and can greatly shorten the deployment time of automation.

[0044] RPA robots can achieve data transmission and integration of automated processes by connecting different systems. RPA robots can also be integrated with a variety of software, such as ERP, CRM, spreadsheets, databases, etc. RPA robots can complete automated operations through open APIs or other technical interfaces, thereby building a tight automated process and automatically executing it between multiple systems.

[0045] Although RPA robots can automatically process a large number of repetitive, tedious, and rule-based power supply station businesses by simulating human interaction. However, since there are a large number of identification and processing of work order scans or image businesses in the business of grassroots power supply stations, it is necessary to further combine OCR technology to process related businesses. The so-called OCR technology, or optical character recognition technology, refers to the computer input technology that converts the text of various bills, newspapers, books, manuscripts, and other printed materials into image information through optical input methods such as scanning, and then uses text recognition technology to convert the image information into usable text. It can be applied to the entry and processing of bank bills, large amounts of text materials, archives, and documents.

[0046] like Figure 2 As shown, processing the image to be recognized by combining RPA technology with OCR technology specifically includes steps S1, S2 and S3.

[0047] First, in step S1, the image to be recognized needs to be scanned, and the shape of the characters in the image to be recognized is determined by detecting the dark and bright areas of the scanning result.

[0048] In order to improve the accuracy and efficiency of image recognition, when scanning the image to be recognized, a unified high-speed scanner or scanner can be used and operated according to pre-established specifications. In this way, a large number of high-quality images with similar specifications can be obtained, so as to analyze the dark and bright areas on the scan results to determine the shape of the text. By using a unified high-speed scanner or scanner and avoiding the use of personal devices, such as the camera function of a personal mobile phone for shooting, it can effectively reduce the situation where the analysis and recognition of the image are affected by differences in equipment, shooting angles and light.

[0049] In order to further improve work efficiency, a pre-verification function can be added to step S1, that is, scans with low recognition rates are excluded in advance and directly transferred to manual processing, thereby avoiding the process entering large-scale processing and then being manually processed, which affects the work efficiency of business personnel. It is also possible to directly attach the cut image slices to the system user interface that needs to be compared, avoiding the need for users to switch back and forth between two screens to find comparison elements. Through similar business adjustments and management methods, the workload of business personnel can be further reduced and the work quality and efficiency can be improved.

[0050] Secondly, in step S2, it is necessary to obtain text and layout information according to the shape of the text, and convert the unstructured data in the text and layout information into structured data to form a template file to be identified.

[0051] Finally, in step S3, a requirement document needs to be generated according to user requirements, and the template file to be identified is searched according to the requirement document to obtain the file path, thereby obtaining the object to be identified.

[0052] The requirements document is a guiding document for system development, which describes in detail the functions and goals that the system should achieve. For example, the requirements document may include the following: "The system needs to be able to automatically extract the signing date, contract amount, and names of both parties to the contract from the contract and store this information in the database." The requirements document can serve as the basis for all work and decisions during the entire development process, and is used to explain what information needs to be extracted from the contract.

[0053] The template file to be identified is equivalent to a "template" that defines how to identify specific information in the contract. For example, the template file to be identified may have the following provisions: "The contract date is usually on the first page of the contract, usually near the words 'Signing Date'; the contract amount is usually after the word 'Total Amount'." The template file to be identified provides an identification rule that defines how to identify the specific information that needs to be obtained (for example, through rules such as keywords and locations).

[0054] Finally, the object to be identified is the specific information that needs to be extracted from the contract document. The object to be identified includes, for example, the following information: contract signing date, contract amount, names of both parties to the contract. The system will search for relevant content in the contract text according to the rules in the template file to be identified, and then extract this information.

[0055] The present invention can further combine ICR technology based on OCR technology. The so-called ICR technology is intelligent character recognition technology, which is an artificial intelligence technology based on OCR technology and implanted with computer deep learning function.

[0056] Since the ICR system used by the ICR technology itself has a self-learning system, in order to improve the recognition and conversion effect of the object to be recognized, a neural network model for recognizing and converting the object to be recognized can be formed through the self-learning of the ICR system. Specifically, the character recognition library can be automatically updated with the help of machine learning (ML) technology and convolutional neural network (CNN) technology. By annotating and training a large number of character sets in the early stage, the neural network model required for recognizing and converting the object to be recognized can be gradually constructed. ICR technology can accurately identify different types of characters through machine learning and neural network models. This may be necessary for business control methods that require high accuracy. For example, when the system needs to recognize information such as handwritten signatures and handwritten dates.

[0057] The intelligent character recognition system can be configured with multiple different recognition engines at the same time, and the recognition results can be mutually verified by the multiple different recognition engines, thereby improving the accuracy of the recognition results. The multiple different recognition engines are all given selective voting rights to determine the credibility of the characters. Since the expertise of multiple recognition engines is different, for example, some are good at recognizing numbers, some are good at recognizing English, and some are good at recognizing Chinese. Therefore, the user needs to automatically select the recognition engine or configure the voting weights of different engines according to the type of content to be recognized, so as to further improve the accuracy of the recognition results.

[0058] By configuring the multiple different recognition engines, the robustness and reliability of the system can be improved. Each recognition engine may have different advantages and disadvantages in different scenarios. Therefore, by configuring multiple engines and verifying each other, the risk of recognition errors can be effectively reduced. For example, one engine is good at recognizing printed text, while the other is more sensitive to handwriting. Suppose a contract document contains both machine-printed text and handwritten signatures. If the system is only configured with one recognition engine that is good at recognizing printed text, the handwritten signature information may be missed. If multiple recognition engines are configured, the system can compare and verify the recognition results of different engines to ensure the accuracy of recognition. Therefore, combining ICR technology and configuring multiple recognition engines can help achieve more efficient and accurate multi-dimensional business control, especially when faced with complex and diverse document content.

[0059] In the process of simulating the power station staff to handle daily business through RPA robots, the information received by the RPA robots to be processed may not be structured fields or scanned images to be processed, but a piece of text expressed in human natural language. In order to handle such business, the RPA robot not only needs to have the ability to understand text expressed in human natural language, but also needs to have the ability to feedback the processing information to the user in the form of human natural language after completing the business.

[0060] Therefore, it is necessary to further combine NLP technology on the basis of RPA technology so that RPA robots have the ability to handle the above business. The so-called NLP technology, namely natural language processing technology, is a technology that studies how to make computers understand and generate human natural language. Specifically, NLP technology consists of two main technologies: natural language understanding and natural language generation. Among them, the main goal of natural language understanding is to help machines better understand human language; while the main goal of natural language generation is to help machines generate language that humans can understand.

[0061] By integrating NLP technology into the application scenarios of RPA robots, RPA robots can be helped to analyze structured, unstructured and semi-structured data. Especially in the automated end-to-end process, NLP technology can enable RPA robots to understand language and make judgments and decisions like humans, and process specific business to further improve work efficiency and user experience.

[0062] Specifically, when the raw data of the business to be processed received by the RPA robot is not a structured field, nor a scanned image that needs to be processed by OCR technology, but a text expressed in human natural language, the natural language understanding technology in NLP technology can be used to help the RPA robot understand the raw data. After the RPA robot processes the business to be processed, the natural language generation technology in NLP technology can be used to convert the processing result of the business to be processed into text expressed in human natural language.

[0063] In addition to the above application scenarios, after the OCR technology converts the image to be recognized into the text information, the NLP technology can be further used to optimize the text logic of the converted text information. For example, the NLP technology can be used to perform word segmentation, part-of-speech tagging, and syntactic analysis on the text information, and identify the most critical logical words to achieve optimization processing.

[0064] As described above, the present invention can solve most of the problems encountered in the data acquisition and processing links, improve work efficiency, and reduce the burden on employees through the organic combination of RPA technology, OCR technology, and NLP technology. However, when handling business at the grassroots power supply, in addition to acquiring and processing data, it is often necessary to make subjective judgments and decisions on the processing methods of specific businesses based on the work experience of employees. For these judgments and decisions, it is necessary to further combine RPA technology and decision-making judgment technology for control and processing, so as to achieve more intelligent judgment and decision-making of RPA robots and further improve the user experience.

[0065] Specifically, the decision-making judgment technology may include establishing a business rule engine and establishing a knowledge base. Among them, the business rule engine (Business Rule Engine, BRE) refers to a program or software that can execute one or more business rules. It is a component embedded in an application that can separate business decisions from application code and write business decisions using predefined semantic modules. The business rule engine can be used to enable the RPA robot to accept data input and explain business rules to the RPA robot, so that the RPA robot can make business decisions and process the business according to the business rules.

[0066] The business rules executed by the business rule engine are related to the business norms, regulations, and logical judgments in the enterprise, but have nothing to do with the processing of the program. The business rule engine supports real-time updating and execution of business rules, which can take effect in a timely manner during operation. In addition, the business rule engine allows users to define the business rules in a visual way, that is, to define the business rules in a way that is close to natural language rather than programming language. This can lower the operating threshold and make it easy for business personnel to define and change the business rules.

[0067] When designing the business rule engine, the business rules are not written directly in the automation script of the RPA robot, but an independent engine is used. That is, the business rule engine is decoupled from the automated business process flow, so that the rights and responsibilities of developers and business personnel can be separated. Developers focus on the implementation of automated business process flows, while business personnel focus on the definition and maintenance of business rules. By clarifying the rights and responsibilities of both developers and business personnel, it not only has a positive impact on the design process of automated business process flows, but also has a positive impact on the definition and recognition of future production and operation problems.

[0068] The knowledge base is an information system that collects, processes and shares all the knowledge in an organization. It can classify, store and manage a large amount of valuable plans, plans, results, experiences and other knowledge in the organization, accumulate knowledge assets, avoid loss, and promote knowledge learning, sharing, training, reuse and innovation.

[0069] Generally speaking, RPA technology combined with the knowledge base is mainly used in call center scenarios for customer service or employee services. In order to make better use of the knowledge base, on the one hand, it is necessary to consider the accumulation method and representation mode of knowledge in the knowledge base. Most of the knowledge bases provided for human use are expressed in natural language, and the knowledge base in the present invention is provided for RPA robots, so it is more necessary to adopt regularized and knowledge graph storage methods. On the other hand, by adding credibility analysis to the knowledge base, it can be determined whether the business to be processed requires manual assistance, thereby ensuring that knowledge with high credibility can be directly used by the RPA robot, while knowledge with low credibility still requires manual assistance verification.

[0070] The multi-dimensional business control method based on artificial intelligence technology of the present invention improves the overall digitization and intelligence of the business processing flow from multiple dimensions through the organic combination of RPA technology, OCR technology, NLP technology and decision-making judgment technology, and makes full use of data support technology to complete a large number of repetitive, tedious, highly standardized and rule-clear business tasks, allowing front-line employees to focus more on high-value-added innovation work that is conducive to business development, which is beneficial to the long-term development of the industry and has significant beneficial effects.

[0071] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0072] It should be noted that the devices and methods disclosed in the embodiments of this article can also be implemented in other ways. The device implementation described above is only schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of this article. In this regard, each box in the flowchart or block diagram can represent a part of a module, program or code, and the module, program segment or part of the code contains one or more executable instructions for implementing the specified logical function, and the module, program segment or part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented by a dedicated hardware-based system for performing a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0073] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the present invention. After reading the above content, it will be apparent to those skilled in the art that various modifications and substitutions of the present invention will occur. Therefore, the protection scope of the present invention should be limited by the appended claims.

Claims

1. A multi-dimensional business control method based on artificial intelligence technology, characterized in that: include: Build automated business processing flows through RPA technology to handle pending business; When the original data of the service to be processed is an image to be recognized, the image to be recognized is processed in combination with OCR technology to convert the image to be recognized into text information; When the original data is text expressed in human natural language, the business to be processed is processed in combination with NLP technology; A business rule engine and a knowledge base are established, and the business to be processed is processed in combination with the business rule engine and the knowledge base.

2. The multi-dimensional business control method based on artificial intelligence technology according to claim 1 is characterized in that: The processing of the image to be recognized by combining OCR technology comprises the following steps: Scanning the image to be recognized, and determining the shape of the characters in the image to be recognized by detecting the dark and bright areas of the scan result; According to the shape of the text, the text and layout information are obtained, and the unstructured data in the text and layout information are converted into structured data to form a template file to be identified; A requirement document is generated according to user requirements, and the template file to be identified is searched according to the requirement document to obtain a file path, thereby obtaining an object to be identified.

3. The multi-dimensional business control method based on artificial intelligence technology according to claim 2 is characterized in that: Combined with ICR technology, a neural network model for identifying and transforming the object to be identified is formed through self-learning of the ICR system.

4. The multi-dimensional business control method based on artificial intelligence technology according to claim 3 is characterized in that: The ICR system is configured with a plurality of different recognition engines, and the recognition results are mutually verified by the plurality of different recognition engines; The multiple different recognition engines are each given selective voting rights.

5. The multi-dimensional business control method based on artificial intelligence technology according to claim 2 is characterized in that: The image to be identified is scanned by a unified high-definition camera or scanner.

6. The multi-dimensional business control method based on artificial intelligence technology according to claim 1 is characterized in that: The processing of the pending business in combination with NLP technology includes: The original data is understood through NLP technology, and the processing results of the business to be processed are converted into text expressed in human natural language.

7. The multi-dimensional business control method based on artificial intelligence technology according to claim 1 is characterized in that: The text information obtained by converting the image to be recognized is optimized in terms of text logic through NLP technology.

8. The multi-dimensional business control method based on artificial intelligence technology according to claim 1 is characterized in that: The business rule engine is decoupled from the automated business processing flow.

9. The multi-dimensional business control method based on artificial intelligence technology according to claim 1, characterized in that: The business rule engine supports real-time updating and execution of business rules and allows the business rules to be defined in a visual manner.

10. The multi-dimensional business control method based on artificial intelligence technology according to claim 1, characterized in that: In combination with the knowledge base, credibility analysis is performed to determine whether the business to be processed requires manual assistance.

Images