WO2024055862A1

WO2024055862A1 - Document review method and apparatus for implementing ia by combining rpa and ai, and electronic device

Info

Publication number: WO2024055862A1
Application number: PCT/CN2023/116767
Authority: WO
Inventors: 黄伟
Original assignee: 北京来也网络科技有限公司; 来也科技(北京)有限公司
Priority date: 2022-09-13
Filing date: 2023-09-04
Publication date: 2024-03-21
Also published as: CN115511441A

Abstract

The present application relates to the technical field of robotic process automation (RPA) and artificial intelligence (AI), and relates to a document review method and apparatus for implementing IA by combining RPA and AI, and an electronic device. The method comprises: acquiring at least one document to be reviewed corresponding to a target service item; reviewing each said document on the basis of an AI technology, to determine whether a plurality of preset types of problems exist in each said document; and when it is determined that at least one preset type of problem exist in each said document, generating modification suggestion information corresponding to said document. Thus, documents to be reviewed corresponding to the target service item are automatically reviewed on the basis of the AI technology, thereby reducing the labor costs required for document review, and improving the efficiency of document review. According to the present application, contact information of a provider can be acquired by the IA implemented by combining the RPA and the AI, and a review result of each document to be reviewed is automatically fed back to the corresponding provider, thereby further reducing the labor costs required for feedback of the review result.

Description

Combining RPA and AI to implement IA document review methods, devices and electronic equipment

Cross-references to related applications

This disclosure claims priority from Chinese Patent Application No. 2022111101693 filed in China on September 13, 2022, the entire content of which is incorporated herein by reference.

Technical field

The present disclosure relates to the technical fields of robotic process automation and artificial intelligence, and specifically relates to a document review method and device, electronic equipment, computer-readable storage media, computer program products and computer programs that combine RPA and AI to implement IA.

Background technique

Robotic Process Automation (RPA) uses specific "robot software" to simulate human operations on a computer and automatically execute process tasks according to rules.

Artificial Intelligence (AI for short) is a technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence.

Intelligent Automation (IA) is a general term for a series of technologies from robotic process automation to artificial intelligence. It combines RPA with Optical Character Recognition (OCR), Intelligent Character Recognition (ICR), and process mining. (Process Mining), Deep Learning (DL), Machine Learning (ML), Natural Language Processing (NLP), Speech Recognition (Automatic Speech Recognition, ASR), Speech Synthesis (Text To Speech) , TTS), Computer Vision (CV) and other AI technologies are combined to create end-to-end business processes that can think, learn and adapt, covering from process discovery, process automation, to automatic and continuous The entire process of data collection, understanding the meaning of data, and using data to manage and optimize business processes.

In many business scenarios, documents submitted by users need to be reviewed. For example, in order to handle medical device registration, drug registration and other pharmaceutical registration matters, pharmaceutical companies can submit relevant application documents to the Food and Drug Administration (hereinafter referred to as the Food and Drug Administration). The approval department of the Food and Drug Administration will review the documents submitted by the pharmaceutical company. , and if the review passes, the corresponding certificate will be issued. If the review fails, the pharmaceutical company will be notified to modify the document. In related technologies, document review is usually performed manually, which not only has high labor costs but also has low efficiency. How to efficiently review documents with lower labor costs has become an urgent problem that needs to be solved.

Contents of the invention

Embodiments of the present disclosure provide a document review method and device, electronic equipment, computer-readable storage media, computer program products, and computer programs that combine RPA and AI to implement IA to solve the high labor cost of document review methods in related technologies. and low efficiency technical issues.

The embodiment of the first aspect of the present disclosure provides a document review method that combines RPA and AI to implement IA, including: obtaining at least one document to be reviewed corresponding to the target business matter; reviewing each document to be reviewed based on AI technology to determine each document to be reviewed. Review whether the document has multiple preset types of problems; and based on determining that each document to be reviewed has at least one preset type of problem, generate modification suggestion information corresponding to the document to be reviewed.

In some embodiments, the preset types include information completion types; each document to be reviewed is reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed, including: obtaining the information of each document to be reviewed. identification; based on the target business matter, query the knowledge graph corresponding to the pre-created target business matter to obtain the identification of at least one target document required by the target business matter; and compare the identification of each document to be reviewed with the identification of each target document Yes, to determine whether the documents to be reviewed are complete.

In some embodiments, reviewing each document to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed also includes: text-checking each document to be reviewed based on optical character recognition (OCR) technology. Identify to obtain the text information contained in each document to be reviewed; perform information extraction on the text information contained in each document to be reviewed to obtain the fields to be reviewed and the corresponding field values contained in each text information; based on the target business matters, Query the knowledge graph to obtain the required fields in each target document; and determine whether among the fields to be reviewed contained in each text information, there is a target field that is consistent with the required fields in the corresponding target document, and determine whether the target field is Corresponding field values exist to determine whether the information in each document to be reviewed is complete.

In some embodiments, reviewing each document to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed also includes: obtaining the same fields to be reviewed in all text information; and based on There are corresponding field values for the same fields to be reviewed, and the field values corresponding to the same fields to be reviewed are compared to determine whether the information in each document to be reviewed is consistent.

In some embodiments, the preset types include process specification types; each document to be reviewed is reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed, including: based on OCR technology, each document to be reviewed is reviewed. Review documents for text recognition to obtain the text information contained in each document to be reviewed; based on the target business matters, query the knowledge graph corresponding to the pre-created target business matters to obtain the process specifications corresponding to the target business matters; and based on each to be reviewed Documents and the text information they contain to determine whether each document to be reviewed meets the process specifications.

In some embodiments, the preset types include writing specification types; each document to be reviewed is reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed, including: based on OCR technology, each document to be reviewed is reviewed. Review documents for text recognition to obtain the text information contained in each document to be reviewed; and identify the text information contained in each document to be reviewed The text information is input into the pre-trained language model, so that through the language model, it can be determined whether there are problems with the writing standard type in each document to be reviewed.

In some embodiments, the document review method for implementing IA by combining RPA and AI also includes: sending each document to be reviewed to a manual review platform based on determining that each document to be reviewed does not have multiple preset types of problems.

In some embodiments, the document review method that combines RPA and AI to implement IA also includes: calling a robotic process automation RPA robot to access the business system to obtain the contact information of the provider of each document to be reviewed; and using the RPA robot to use the contact information , feedback the review results of each document to be reviewed to the corresponding provider.

In some embodiments, the target business matter is a pharmaceutical registration matter.

The embodiment of the second aspect of the present disclosure provides a document review device that combines RPA and AI to implement IA, including: an acquisition module for acquiring at least one document to be reviewed corresponding to a target business matter; and an review module for reviewing each document based on AI technology. The documents to be reviewed are reviewed to determine whether each document to be reviewed has multiple preset types of problems; and a generation module is used to generate the documents to be reviewed based on determining that each document to be reviewed has at least one preset type of problem. Corresponding modification suggestion information.

In some embodiments, the preset types include information completion types; an audit module is used to: obtain the identification of each document to be audited; based on the target business matter, query the knowledge graph corresponding to the pre-created target business matter to obtain the target business The identification of at least one target document required by the matter; and compare the identification of each document to be reviewed with the identification of each target document to determine whether each document to be reviewed is complete.

In some embodiments, the review module is also used to: perform text recognition on each document to be reviewed based on optical character recognition OCR technology to obtain the text information contained in each document to be reviewed; Extract information to obtain the fields to be reviewed and the corresponding field values contained in each text information; query the knowledge graph based on the target business matters to obtain the required fields in each target document; and determine the fields contained in each text information. Among the fields to be reviewed, whether there is a target field that is consistent with the required field in the corresponding target document, and whether there is a corresponding field value in the target field, to determine whether the information in each document to be reviewed is complete.

In some embodiments, the audit module is also used to: obtain the same field to be audited in all text information; when there is a corresponding field value for the same field to be audited, compare the field values corresponding to the same field to be audited. Yes, to determine whether the information in each document to be reviewed is consistent.

In some embodiments, the preset type includes a process specification type; the audit module is also used to: perform text recognition on each document to be audited based on OCR technology to obtain the text information contained in each document to be audited; based on the target business matters , query the knowledge graph corresponding to the pre-created target business matter to obtain the process specification corresponding to the target business item; and based on each document to be reviewed and the text information contained, determine whether each document to be reviewed meets the process specification.

In some embodiments, the preset types include writing specification types; the review module is also used to: perform text recognition on each document to be reviewed based on OCR technology to obtain the text information contained in each document to be reviewed; and Review document The text information contained in the document is input into the pre-trained language model, so that through the language model, it can be determined whether there are problems with the standard type of writing in each document to be reviewed.

In some embodiments, the document review device that combines RPA and AI to implement IA also includes: a first sending module, configured to send each document to be reviewed to a manual based on determining that each document to be reviewed does not have multiple preset types of problems. Review platform.

In some embodiments, the document review device that combines RPA and AI to implement IA also includes: a calling module for calling a robotic process automation RPA robot to access the business system to obtain the contact information of the provider of each document to be reviewed; and second The sending module is used to use RPA robots to feedback the review results of each document to be reviewed to the corresponding provider through contact information.

The third embodiment of the present disclosure provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above embodiments of the present disclosure are implemented. the method described.

The fourth embodiment of the present disclosure provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described in the above embodiments of the present disclosure is implemented.

The fifth aspect embodiment of the present disclosure proposes a computer program product, including a computer program that implements the method described in the above embodiments of the present disclosure when executed by a processor.

The sixth aspect embodiment of the present disclosure proposes a computer program. The computer program includes computer program code. When the computer program code is run on a computer, the computer executes the method as described in the above embodiments of the present disclosure.

The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

By obtaining at least one document to be reviewed corresponding to the target business matter, each document to be reviewed is reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed. After determining that each document to be reviewed has at least one preset type of problem, Under the condition of setting types of problems, the modification suggestion information corresponding to the documents to be reviewed is generated, and the automatic review of the documents to be reviewed corresponding to the target business matters is realized based on AI technology, which reduces the labor cost required for document review and improves the efficiency of document review. Review efficiency. In addition, by generating modification suggestion information corresponding to the document to be audited when it is determined that each document to be audited has at least one preset type of problem, modification suggestions can be provided to the provider of the document to be audited, and it is convenient for the provider to modify the document to be audited. Revise. This disclosed embodiment can also combine RPA and AI to obtain the contact information of the IA provider, and automatically feed back the review results of each document to be reviewed to the corresponding provider, thereby further reducing the labor cost required to feedback the review results.

Additional aspects and advantages of embodiments of the disclosure will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the disclosure.

Description of drawings

In the drawings, unless otherwise specified, the same reference numbers refer to the same or similar parts or elements throughout the several figures. The drawings are not necessarily to scale. It should be understood that these drawings depict only some embodiments in accordance with the disclosure and are not to be considered limiting of the scope of the disclosure.

Figure 1 is a schematic flowchart of a document review method for implementing IA by combining RPA and AI according to the first embodiment of the present disclosure;

Figure 2 is a schematic flowchart of a document review method for implementing IA by combining RPA and AI according to the second embodiment of the present disclosure;

Figure 3 is a schematic flowchart of a document review method for implementing IA by combining RPA and AI according to the third embodiment of the present disclosure;

Figure 4 is a schematic flowchart of a document review method for implementing IA by combining RPA and AI according to the fourth embodiment of the present disclosure;

Figure 5 is a schematic flowchart of a document review method for implementing IA by combining RPA and AI according to the fifth embodiment of the present disclosure;

Figure 6 is a schematic structural diagram of a document review device that combines RPA and AI to implement IA according to the sixth embodiment of the present disclosure;

FIG. 7 is a block diagram of an electronic device used to implement a document review method for implementing IA by combining RPA and AI according to an embodiment of the present disclosure.

Detailed ways

Embodiments of the disclosure/disclosure are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present disclosure/disclosure and are not to be construed as limitations of the present disclosure/disclosure.

These and other aspects of the present disclosure/disclosed embodiments will become apparent with reference to the following description and accompanying drawings. In these descriptions and drawings, some specific embodiments of the present disclosure/disclosure are specifically disclosed to represent some ways of implementing the principles of the present disclosure/disclosure embodiments, but it should be understood that the present disclosure/disclosure The scope of the embodiments is not limited in this way. On the contrary, the present disclosure/disclosed embodiments include all changes, modifications and equivalents falling within the spirit and scope of the appended claims.

It should be noted that the acquisition, storage and application of data involved in the technical solution of this disclosure application are in compliance with relevant laws and regulations and do not violate public order and good customs.

Embodiments of the present disclosure provide a document review method and device, electronic equipment, computer-readable storage media, computer program products, and computer programs that combine RPA and AI to implement IA. Among them, the method includes: obtaining at least one document to be reviewed corresponding to the target business matter; reviewing each document to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed; based on determining that each document to be reviewed is There is at least one preset type of problem, and modification suggestion information corresponding to the document to be reviewed is generated. As a result, it is possible to automatically review documents to be reviewed corresponding to target business matters based on AI technology, reducing the labor costs required for document review and improving the efficiency of document review.

The document review method, device, electronic device, and storage medium provided by the embodiments of the present disclosure that combine RPA and AI to implement IA can be applied to any field that requires document review, such as the medical field and the judicial field. The embodiments of the present disclosure do not limit this. . Each embodiment of the present disclosure is explained by taking the medical field as an example.

In order to clearly explain each embodiment of the present invention, technical terms involved in the embodiments of the present invention are first explained.

In the description of the embodiments/disclosure of the present disclosure, the term "plurality" means two or more.

In the description of the embodiments of this disclosure, "RPA robot" refers to a software robot that can combine AI technology and RPA technology to automatically handle online business. RPA robots have two characteristics: "connector" and "non-intrusion". By simulating human operation methods, they can extract, integrate and connect data from different systems in a non-intrusive way without changing the information system.

In the description of the embodiments of the present disclosure, both "field" and "field value" are fragments composed of a single character or multiple consecutive characters. Among them, "field" can be understood as the attribute item key, and "field value" can be understood as the attribute value value, and there is a corresponding relationship between fields and field values. The fields and corresponding field values together form a piece of structured data. For example, "Zhang San" is the field value corresponding to the field "Name", and "Name" and "Zhang San" form a piece of structured data.

In the description of the embodiment of the present disclosure, "documents to be reviewed" refers to documents received by the approval department for handling a certain business. Correspondingly, "target business matters" refers to the business. "Provider" refers to the party that submits documents to be reviewed by the approval department. The provider can be an individual or an enterprise, and the embodiment of the present disclosure does not limit this.

For example, in order to handle pharmaceutical registration matters such as medical device registration, drug registration, dosing method change registration, or drug dosage change registration, pharmaceutical companies can submit applications to the Food and Drug Administration and submit relevant application documents. Assuming that a pharmaceutical company submits a drug registration application to the Food and Drug Administration and submits relevant application documents, the drug registration matters are the target business matters, and the documents submitted by the pharmaceutical company when applying for drug registration matters are the corresponding target business matters. For documents to be reviewed, the pharmaceutical company is the provider of the documents to be reviewed.

In the description of the embodiments of this disclosure, "target document" refers to the document required to successfully handle the target business matter, that is, the document required by the target business matter.

In the description of the embodiments of the present disclosure, "preset type" refers to the type of problems that may exist in the preset documents to be reviewed.

In the description of the embodiment of the present disclosure, "information completion type" means that the document to be reviewed has problems such as incomplete documents, incomplete information in the document to be reviewed, or inconsistent information in the document to be reviewed, and needs to be supplemented or modified. Among them, problems belonging to the information completion type may include, for example: incomplete documents to be reviewed. For example, the target documents required by the target business matter include document A, document B and document C, but the documents to be reviewed only include document A and document B; The information in the document to be reviewed is incomplete. For example, document A required by the target business matter requires field a and the corresponding field value. However, the document to be reviewed includes document A, but document A does not include field a and its corresponding field value. field value, or, The document A includes field a but does not include the corresponding field value; the information in the document to be reviewed is inconsistent. For example, the document to be reviewed includes document A, document B and document C, where both document A and document B include field a and the corresponding field value. field value, but the field value corresponding to field a in document A is different from the field value corresponding to field a in document B.

In the description of the embodiment of the present disclosure, the "process specification type" means that the document to be reviewed does not meet the process specification corresponding to the target business matter. Among them, the process specifications corresponding to the target business matters are the specifications that should be followed in the handling process of the target business matters. For example, taking the target business matter as a medical device registration matter, it is assumed that the process specifications corresponding to the medical device registration matter include: If a new mandatory standard or national standard is released and implemented within the validity period of the medical device registration certificate, the registered product must comply with the new If the changes made to mandatory standards and national standards require change registration, the registrant should first go through the change registration procedures and obtain the change registration (filing) document approved by the original approval department before submitting an application for registration renewal. Then within the validity period of the medical device registration certificate of the registrant (i.e., the provider of the materials to be reviewed in this disclosed embodiment), a new mandatory standard is released and implemented, and the medical device is a symbol of the new mandatory standard that should be changed to register. In this case, if the documents to be reviewed submitted by the registrant do not include the change registration (filing) document approved by the original approval department, the documents to be reviewed will have process specification issues.

In the description of the embodiment of the present disclosure, the "standard writing type" means that there are typographical problems, English translation errors, irregular use of professional terms and other writing format problems in the document to be reviewed.

In the description of the embodiments of the present disclosure, "OCR (Optical Character Recognition, Optical Character Recognition)" specifically refers to the electronic device checking the characters printed on the paper, determining their shape by detecting dark and light patterns, and then using the character recognition method The process of translating shapes into computer text; that is, for printed characters, the text in the paper document is optically converted into a black and white dot matrix image file, and the text in the image is converted into a text format through recognition software. Technology for further editing and processing by word processing software.

In the description of the embodiments of the present disclosure, "information extraction" refers to structuring the information contained in the text into a table-like organizational form. Among them, information extraction can include named entity recognition and relationship extraction. Named entity recognition, that is, identifying various named entities in a piece of text. The named entities that need to be recognized usually include person names, place names, organization names, drugs, time, etc., which can be set according to different application scenarios. For example, for pharmaceutical registration matters, the named entities that need to be identified can include medical nouns, normative nouns, the address of the registrant, the name of the registrant, the name of the agent acting for pharmaceutical registration matters, etc. The purpose of relationship extraction is to identify target relationships in text entities and extract semantic relationships between entities by identifying the relationships between entities. Relation extraction can be achieved through sequence annotation, classification, dependency syntax analysis, semantic dependency analysis and other technologies.

In the description of the embodiments of the present disclosure, a "language model" is any machine model used to determine whether a document to be reviewed has a writing specification type problem, such as a neural network model. Among them, the language model can be obtained by training training samples in advance.

In the description of the embodiments of this disclosure, "manual review platform" refers to a platform that can review documents manually, such as a human-machine collaboration platform.

In the description of the embodiments of this disclosure, "business system" refers to the online system used by the approval department to handle business matters, such as the management system of the Food and Drug Administration.

The document review method and device, electronic device, computer readable storage medium, computer program product and computer program that combines RPA and AI to implement IA according to the disclosure/disclosed embodiments are described below with reference to the accompanying drawings.

First, the document review method for implementing IA by combining RPA and AI in the embodiment of the present disclosure will be described with reference to the accompanying drawings.

Figure 1 is a flow chart of a document review method for implementing IA by combining RPA and AI according to the first embodiment of the present disclosure. As shown in Figure 1, the method may include steps 101 to 103.

Step 101: Obtain at least one document to be reviewed corresponding to the target business matter.

It should be noted that the document review method that combines RPA and AI to implement IA in the embodiment of the present disclosure can be executed by a document review device that combines RPA and AI to implement IA. Hereinafter, the document review device that combines RPA and AI to implement IA will be referred to as a document Audit device. Wherein, the document review device can be implemented by software and/or hardware, and the document review device can be an electronic device, or can also be configured in an electronic device to realize automatic review of documents, thereby reducing the labor costs required for document review, Improve the efficiency of document review. The electronic device may include but is not limited to a terminal device, a server, etc., and this embodiment does not specifically limit the electronic device.

The documents to be reviewed corresponding to the target business matter may include one document or multiple documents, and this disclosure does not limit this.

In some embodiments, the document review device can provide an upload interface, so that the provider can upload the documents required to handle the target business matter through the upload interface. Correspondingly, the document review device can obtain at least one to-be-reviewed document corresponding to the target business matter. document.

Step 102: Review each document to be reviewed based on AI technology to determine whether each document to be reviewed contains multiple preset types of problems.

In some embodiments, multiple preset types corresponding to the target business matter can be determined in advance by: summarizing the problems that frequently occur in documents to be reviewed during the processing of the target business matter, and classifying these problems to obtain multiple preset type. Furthermore, the document review device can review each document to be reviewed one by one, and determine for each preset type whether each document to be reviewed has problems of the preset type.

Among them, multiple preset types may include, for example, information completion type, process specification type, writing specification type, etc.

Among them, different target business matters can correspond to different preset types.

Step 103: When it is determined that each document to be reviewed has at least one preset type of problem, generate modification suggestion information corresponding to the document to be reviewed.

In some embodiments, the document review device may generate modification suggestion information corresponding to the document to be reviewed based on preset types of problems existing in the document to be reviewed.

For example, assuming that the document review device determines that the document to be reviewed has an information completion type problem, where the problem is specifically: among the documents to be reviewed submitted by the provider, document A required by the target business matter is missing, then the document review device Modification suggestion information "Supplementary Document A required" can be generated.

The document review method provided by the embodiment of the present disclosure combines RPA and AI to implement IA, and obtains at least one document to be reviewed corresponding to the target business matter; each document to be reviewed is reviewed based on AI technology to determine whether there are multiple documents to be reviewed. Problems of a preset type; when it is determined that each document to be reviewed has at least one problem of a preset type, modification suggestion information corresponding to the document to be reviewed is generated. As a result, it is possible to automatically review documents to be reviewed corresponding to target business matters based on AI technology, reducing the labor costs required for document review and improving the efficiency of document review. In addition, by generating modification suggestion information corresponding to the document to be audited when it is determined that each document to be audited has at least one preset type of problem, modification suggestions can be provided to the provider of the document to be audited, and it is convenient for the provider to modify the document to be audited. Revise.

In some embodiments, the preset types may include information completion types. In conjunction with Figure 2, in this embodiment of the present disclosure, each document to be reviewed is reviewed based on AI technology to determine whether there is an information completion type in each document to be reviewed. Explain the process of the problem.

Figure 2 is a flow chart of a document review method for implementing IA by combining RPA and AI according to the second embodiment of the present disclosure. As shown in Figure 2, the document review method for implementing IA by combining RPA and AI may include steps 201 to 211.

Step 201: Obtain at least one document to be reviewed corresponding to the target business matter.

For the specific implementation process and principle of step 201, reference can be made to the description of the above embodiments and will not be described again here.

In some embodiments, problems belonging to the information completion type may include: incomplete documents to be reviewed. Correspondingly, after the document review device obtains each document to be reviewed, it can determine whether there is an information completion type problem in each document to be reviewed through the following steps 202-205.

Step 202: Obtain the identification of each document to be reviewed.

Among them, this ID is used to uniquely identify the document to be reviewed. The identification of the document to be reviewed may be the file name of the document to be reviewed, or the number corresponding to the document to be reviewed, etc. This disclosure does not limit this.

Step 203: Based on the target business matter, query the knowledge graph corresponding to the pre-created target business matter to obtain the identification of at least one target document required by the target business matter.

In some embodiments, for any business matter, a knowledge graph corresponding to the business matter can be created in advance based on the documents required to handle the business matter. Among them, the knowledge graph may include, for example, a first node corresponding to the business matter, and a second node corresponding to the documents required to handle the business matter, and the first node and the second node are connected through edges. Therefore, the document review device can query the knowledge corresponding to the pre-created target business matter based on the target business matter. Graph, determine the second node in the knowledge graph that is connected to the first node corresponding to the target business matter through an edge, and determine the identifier of the document corresponding to the second node as the identifier of at least one target document required by the target business matter.

Step 204: Compare the identification of each document to be reviewed with the identification of each target document to determine whether each document to be reviewed is complete.

Step 205: When it is determined that each document to be reviewed is incomplete, it is determined that each document to be reviewed has an information completion type problem.

In some embodiments, if it is determined that for the identification of each target document, there is an identification of the document to be reviewed that is the same as the identification, it can be determined that the documents to be reviewed are complete. When it is determined that for the identification of at least one target document, there is no identification of the document to be reviewed that is the same as the identification, it can be determined that each document to be reviewed is incomplete, and further it can be determined that each document to be reviewed has an information completion type problem.

In some embodiments, problems belonging to the information completion type may also include: incomplete information in the document to be reviewed. Correspondingly, after step 205, the document review device can also determine whether there are information completion type problems in each document to be reviewed in the manner shown in the following steps 206-210.

Step 206: Based on the optical character recognition OCR technology, perform text recognition on each document to be reviewed to obtain the text information contained in each document to be reviewed.

Step 207: Extract text information contained in each document to be reviewed to obtain fields to be reviewed and corresponding field values contained in each text information.

Step 208: Based on the target business matters, query the knowledge graph to obtain the required fields in each target document.

In some embodiments, for any business matter, when creating a knowledge graph corresponding to the business matter, the knowledge graph can also be created based on the fields that need to be included in each document required to handle the business matter. Correspondingly, in addition to the first node corresponding to the business matter and the second node corresponding to the documents required to handle the business matter, the knowledge graph also includes the third node corresponding to each field that needs to be included in each document. The node is connected to the corresponding second node through an edge. Therefore, the document review device can query the knowledge graph corresponding to the pre-created target business matter based on the target business matter, determine the third node in the knowledge graph connected to each second node through the edge, and determine the fields corresponding to each third node. are the required fields in the corresponding target document (that is, the target document corresponding to the second node connected to the third node).

Step 209: Determine whether there is a target field consistent with the required field in the corresponding target document among the fields to be reviewed included in each text information, and determine whether there is a corresponding field value in the target field, so as to determine whether there is a corresponding field value in each document to be reviewed. whether the information is complete.

Step 210: When it is determined that the information in each document to be reviewed is incomplete, it is determined that each document to be reviewed has an information completion type problem.

In some embodiments, for each required field in each target document, the document review device determines that among the fields to be reviewed in the text information contained in the corresponding document to be reviewed, there is a target field that is consistent with the field, and When each target field has a corresponding field value, it can be determined that the information in each document to be reviewed is complete.

In some embodiments, for at least one required field in each target document, the document review device determines that among the fields to be reviewed in the text information contained in the corresponding document to be reviewed, there is no target field that is consistent with the field. In this case, it can be determined that the information in each document to be reviewed is incomplete, and further it can be determined that there is an information completion type problem in each document to be reviewed.

In some embodiments, it may happen that a certain field in a document to be reviewed does not have a corresponding field value. Then, if the text information contained in each document to be reviewed is extracted, a certain field to be reviewed may not be obtained. The corresponding field value of the field means that there is no corresponding field value for the field to be reviewed. For at least one required field in each target document, in the field to be reviewed in the text information contained in the corresponding document to be reviewed determined by the document review device, there is a target field that is consistent with the field, but there is no corresponding target field. In the case of field values, it can be determined that the information in each document to be reviewed is incomplete, and then it can be determined that there is an information completion type problem in each document to be reviewed.

In some embodiments, problems belonging to the information completion type may also include: information in the document to be reviewed is inconsistent. Correspondingly, after step 210, the document review device can also determine whether there are information completion type problems in each document to be reviewed in the manner shown in the following steps:

Get the same fields to be reviewed in all text messages;

When there are corresponding field values for the same field to be reviewed, compare the field values corresponding to the same field to be reviewed to determine whether the information in each document to be reviewed is consistent;

When it is determined that the information in each document to be reviewed is inconsistent, it is determined that each document to be reviewed has an information completion type problem.

In some embodiments, when the document review device determines that the field values corresponding to the same fields to be reviewed in all text information are the same, it can be determined that the information in the documents to be reviewed is consistent. When the document review device determines that in all text information, at least one field value corresponding to the same field to be reviewed is different, it can be determined that the information in each document to be reviewed is inconsistent, and further it can be determined that each document to be reviewed has an information completion type. question.

Step 211: After it is determined that each document to be reviewed has an information completion type problem, modification suggestion information corresponding to the document to be reviewed is generated.

In some embodiments, when the document review device determines that each document to be reviewed does not have any of the above problems, it can determine that each document to be reviewed does not have an information completion type problem. When the document review device determines that each document to be reviewed has at least one of the above problems, it can be determined that each document to be reviewed has an information completion type problem, and then modification suggestion information corresponding to the document to be reviewed can be generated.

In summary, the document review method provided by the embodiments of the present disclosure for implementing IA by combining RPA and AI realizes the automatic review based on AI technology to see whether the documents to be reviewed corresponding to the target business matters have information completion type problems, thereby reducing the need for document review. The required labor costs improve the efficiency of document review. In addition, by generating modification suggestion information corresponding to the document to be audited when it is determined that each document to be audited has an information completion type problem, modification suggestions can be provided to the provider of the document to be audited, and it is convenient for the provider to make information on the document to be audited. Complete or modify.

In some embodiments, the preset type may include a process specification type. In conjunction with FIG. 3 , in this embodiment of the present disclosure, each document to be reviewed is reviewed based on AI technology to determine whether there are issues with the process specification type in each document to be reviewed. The process is explained.

Figure 3 is a flow chart of a document review method for implementing IA by combining RPA and AI according to the third embodiment of the present disclosure. As shown in Figure 3, the document review method for implementing IA by combining RPA and AI may include steps 301 to 305.

Step 301: Obtain at least one document to be reviewed corresponding to the target business matter.

Step 302: Based on OCR technology, text recognition is performed on each document to be reviewed to obtain text information contained in each document to be reviewed.

For the specific implementation process and principles of steps 301-302, reference can be made to the description of the above embodiments and will not be described again here.

Step 303: Based on the target business item, query the pre-created knowledge graph corresponding to the target business item to obtain the process specification corresponding to the target business item.

In some embodiments, for any business matter, when creating a knowledge graph corresponding to the business matter, the knowledge graph can be created according to the specifications that should be followed in the handling process of the business matter. Correspondingly, in addition to the first node corresponding to the business matter, the knowledge graph may also include a fourth node corresponding to the specifications that should be followed in the handling process of the business matter. The fourth node is connected to the first node through an edge. Therefore, the document review device can query the pre-created knowledge graph corresponding to the target business matter based on the target business matter, determine the fourth node in the knowledge graph that is connected to the first node corresponding to the target business matter through an edge, and associate the fourth node with the first node corresponding to the target business matter. The specifications are determined as process specifications corresponding to the target business matters.

Step 304: Based on each document to be reviewed and the text information contained therein, determine whether each document to be reviewed satisfies the process specification.

For example, taking the target business matter as a medical device registration matter, it is assumed that the process specifications corresponding to the medical device registration matter include: If a new mandatory standard or national standard is released and implemented within the validity period of the medical device registration certificate, the registered product is If changes made to comply with new mandatory standards or national standards require change registration, the registrant should first go through the change registration procedures and obtain the change registration (filing) document approved by the original approval department before submitting an application for registration renewal. Documents to be reviewed include medical device registration certificates.

Then the document review device can determine the validity period of the medical device registration certificate based on the text information contained in the medical device registration certificate, and query whether there are new mandatory standards and national standards released and implemented within the validity period, and determine whether registered products comply with the new Changes made to mandatory standards and national standards require change registration. If so, the document review device can review whether the documents to be reviewed submitted by the registrant include the change registration (filing) document approved by the original approval department. If it is not included, the document review device may determine that each document to be reviewed does not meet the process specifications corresponding to medical device registration matters. If included, the document review device can determine that each document to be reviewed meets the process specifications corresponding to the medical device registration matters.

Step 305: When each document to be reviewed does not meet the process specification, it is determined that each document to be reviewed has a process specification type problem, and modification suggestion information corresponding to the document to be reviewed is generated.

In some embodiments, the document review device may generate modification suggestion information corresponding to the document to be reviewed based on the process specification type problems existing in the document to be reviewed. For example, continuing the above example, the document review device can generate modification suggestion information "Please go to the original approval department to register the change, and after obtaining the change registration (filing) document approved by the original approval department, submit an application for registration renewal."

In summary, the document review method provided by the embodiments of the present disclosure for implementing IA by combining RPA and AI realizes the automatic review based on AI technology to see whether the documents to be reviewed corresponding to the target business matters have process specification type issues, thereby reducing the time required for document review. Reduce required labor costs and improve the efficiency of document review. In addition, by generating modification suggestion information corresponding to the document to be audited when it is determined that each document to be audited has a problem of the process specification type, modification suggestions can be provided to the provider of the document to be audited, and it is convenient for the provider to modify the document to be audited.

In some embodiments, the preset type may include a writing standard type. In the following, with reference to Figure 4, in the embodiment of the present disclosure, each document to be reviewed is reviewed based on AI technology to determine whether there is a problem with the writing standard type in each document to be reviewed. The process is explained.

Figure 4 is a flow chart of a document review method for implementing IA by combining RPA and AI according to the fourth embodiment of the present disclosure. As shown in Figure 4, the document review method for implementing IA by combining RPA and AI may include steps 401 to 404.

Step 401: Obtain at least one document to be reviewed corresponding to the target business matter.

Step 402: Based on OCR technology, text recognition is performed on each document to be reviewed to obtain text information contained in each document to be reviewed.

For the specific implementation process and principles of steps 401-402, please refer to the description of the above embodiments and will not be described again here.

Step 403: Input the text information contained in each document to be reviewed into a pre-trained language model, so as to use the language model to determine whether each document to be reviewed has problems with writing standards.

In some embodiments, the language model can be pre-trained to generate a language model. The input of the language model is text information, and the output is the standard-type questions and corresponding confidence levels existing in the text information. In this way, each document to be reviewed can be The text information contained in the document is input into a pre-trained language model, so that the language model can predict the writing standard type problems existing in each text information, and determine the corresponding confidence level, so that the document review device can predict each text based on the language model. The standard type of problems existing in the information and the corresponding confidence level determine whether each document to be reviewed has a standard type of problem. For example, you can set a confidence threshold, and if the text information contained in a document to be reviewed contains a writing standard type question 1 and the confidence level corresponding to question 1 is greater than the confidence threshold, it is determined that the document to be reviewed has a writing standard type. Question 1. Among them, the confidence threshold can be set arbitrarily as needed, for example, it can be set to 0.7, 0.8, etc., and this disclosure does not limit this.

Step 404: When it is determined that each document to be reviewed has a standard-type problem, modification suggestion information corresponding to the document to be reviewed is generated.

In some embodiments, the document review device may generate modification suggestion information corresponding to the document to be reviewed based on the standard-type issues existing in the document to be reviewed. For example, assuming that the document to be reviewed has a writing standard type problem: the English translation of a certain Chinese word x in document A is incorrect, the document review device can generate a modification suggestion message "Please modify the English translation of word x."

In summary, the document review method provided by embodiments of the present disclosure that combines RPA and AI to implement IA can automatically review whether the document to be reviewed corresponding to the target business matter has a standard type of writing based on AI technology, thus reducing the complexity of document review. Reduce required labor costs and improve the efficiency of document review. In addition, by generating modification suggestion information corresponding to the document to be reviewed when it is determined that each document to be reviewed has a standard-type problem, modification suggestions can be provided to the provider of the document to be reviewed, and it is convenient for the provider to modify the document to be reviewed.

The document review method for implementing IA by combining RPA and AI provided by the embodiment of the present disclosure will be further described below with reference to FIG. 5 . Figure 5 is a flow chart of a document review method that combines RPA and AI to implement IA according to the fifth embodiment of the present disclosure. As shown in Figure 5, the method may include steps 501 to 506.

Step 501: Obtain at least one document to be reviewed corresponding to the target business matter.

Step 502: Review each document to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed. The multiple preset types include information completion type, process specification type and writing specification type. .

Step 503: When it is determined that each document to be reviewed has at least one preset type of problem, generate modification suggestion information corresponding to the document to be reviewed.

For the specific implementation process and principles of steps 501-503, reference can be made to the description of the above embodiments and will not be described again here.

Step 504: When it is determined that each document to be reviewed does not have multiple preset types of problems, the document to be reviewed is sent to the manual review platform.

In some embodiments, the document review method that combines RPA and AI to implement IA provided by the embodiments of the present disclosure can be applied to the pre-review process before manually reviewing documents. Correspondingly, in the document review device, determine each document to be reviewed. If the document does not have multiple preset types of problems, it can be determined that each document to be reviewed has passed the pre-review, so that each document to be reviewed can be sent to the manual review platform for further review of each document to be reviewed manually.

Each document to be reviewed is automatically reviewed through the document review device. After it is determined that each document to be reviewed does not have multiple preset types of problems, each document to be reviewed is then sent to the manual review platform for further review, which reduces the need for each document to be reviewed. The labor cost required to review the documents to be reviewed is reduced, and the number of interactions between the provider of each document to be reviewed and the approval department is reduced, and the efficiency of handling target business matters is improved.

For example, for pharmaceutical registration matters, the document review method that combines RPA and AI to implement IA provided by the embodiments of the present disclosure can be used to pre-review registration application documents submitted by pharmaceutical companies. When the document review device determines that the registration application document has at least one type of problem among the information completion type, the process specification type, and the writing specification type, corresponding modification suggestion information can be generated so that the pharmaceutical enterprise can make the review based on the modification suggestion information. The document is modified. When the document review device determines that the registration application document does not have any problems in the information completion type, process specification type, or writing specification type, each document to be reviewed can be sent to the manual review platform so that the approval department can review each document to be reviewed. Review documents for further review. As a result, after the pharmaceutical company submits the registration application documents, it can provide timely feedback to the pharmaceutical company, reduce the pharmaceutical registration application cycle, and improve the registration application efficiency.

Step 505: Call the robotic process automation RPA robot to access the business system to obtain the contact information of the provider of each document to be reviewed.

Among them, the contact information of the provider can be a phone number, email address, etc., and this disclosure does not limit this.

In some embodiments, the business system stores the contact information of the providers of each document to be reviewed, and the document review device can obtain the contact information of the providers of each document to be reviewed from the business system through background data access.

In some embodiments, the document review device can also call the RPA robot to access the business system through web page access to obtain the contact information of the provider of each document to be reviewed. Among them, a web page refers to a file on the World Wide Web organized in HTML (Hyper Text Markup Language) format.

Step 506: Use the RPA robot to feed back the review results of each document to be reviewed to the corresponding provider through contact information.

Among them, the review results may include whether each document to be reviewed has passed the review, or each document to be reviewed has failed to pass the review, as well as the reasons for the failure and modification suggestion information, etc.

In some embodiments, after the document review device completes the review of each document to be reviewed provided by a certain provider, the RPA robot can be used to feed back the review results to the corresponding provider through the provider's contact information, so that the provider can In order to obtain the review results of documents to be reviewed in a timely manner.

Therefore, by calling the RPA robot to obtain the contact information of the provider in the business system, the linkage between the document review device and the business system can be realized. The RPA robot is used to obtain the contact information of the provider, and the review results of each document to be reviewed are passed. Contact information is fed back to the corresponding provider, and RPA and AI can be combined to achieve IA acquisition and provision. The contact information of each party is provided, and the review results of each document to be reviewed are automatically fed back to the corresponding provider, thereby further reducing the labor cost required to feedback the review results.

In order to implement the above embodiments, embodiments of the present disclosure also propose a document review device that combines RPA and AI to implement IA. Figure 6 is a schematic structural diagram of a document review device that combines RPA and AI to implement IA according to the sixth embodiment of the present disclosure.

As shown in Figure 6, the document review device 600 that combines RPA and AI to implement IA includes: an acquisition module 610, an review module 620, and a generation module 630.

The acquisition module 610 is used to acquire at least one document to be reviewed corresponding to the target business matter.

The review module 620 is used to review each document to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each document to be reviewed.

The generation module 630 is configured to generate modification suggestion information corresponding to the document to be reviewed when it is determined that each document to be reviewed has at least one preset type of problem.

It should be noted that the document review device 600 that combines RPA and AI to implement IA in the embodiment of the present disclosure can execute the document review method that combines RPA and AI to implement IA provided in the above embodiments. Among them, the document review device 600 that combines RPA and AI to implement IA can be implemented by software and/or hardware. The document review device 600 that combines RPA and AI to implement IA can be an electronic device, or can also be configured in an electronic device to implement Automatic review of documents, thereby reducing the labor costs required for document review and improving the efficiency of document review. The electronic device may include but is not limited to a terminal device, a server, etc., and this embodiment does not specifically limit the electronic device.

In one embodiment of the present disclosure, the preset types include information completion types; the audit module 620 is used to:

Obtain the identification of each document to be reviewed;

Based on the target business matter, query the knowledge graph corresponding to the pre-created target business matter to obtain the identification of at least one target document required by the target business matter;

Compare the identification of each document to be reviewed with the identification of each target document to determine whether each document to be reviewed is complete.

In one embodiment of the present disclosure, the audit module 620 is also used to:

Based on optical character recognition (OCR) technology, text recognition is performed on each document to be reviewed to obtain the text information contained in each document to be reviewed;

Extract the text information contained in each document to be reviewed to obtain the fields to be reviewed and the corresponding field values contained in each text information;

Based on the target business matters, query the knowledge graph to obtain the required fields in each target document;

Determine whether among the fields to be reviewed included in each text information, there is a target field that is consistent with the required field in the corresponding target document, and whether there is a corresponding field value in the target field, so as to determine whether the information in each document to be reviewed is complete.

Get the same fields to be reviewed in all text messages;

When there are corresponding field values for the same field to be reviewed, the field values corresponding to the same field to be reviewed are compared to determine whether the information in each document to be reviewed is consistent.

In one embodiment of the present disclosure, the preset type includes a process specification type; the audit module 620 is also used to:

Based on OCR technology, text recognition is performed on each document to be reviewed to obtain the text information contained in each document to be reviewed;

Based on the target business matter, query the knowledge graph corresponding to the pre-created target business matter to obtain the process specification corresponding to the target business matter;

Based on each document to be reviewed and the text information it contains, determine whether each document to be reviewed meets the process specifications.

In one embodiment of the present disclosure, the preset types include writing specification types; the review module 620 is also used to:

The text information contained in each document to be reviewed is input into the pre-trained language model, so that through the language model, it is determined whether there are issues of the writing standard type in each document to be reviewed.

In one embodiment of the present disclosure, the document review device 600 that combines RPA and AI to implement IA also includes:

The first sending module is used to send each document to be reviewed to the manual review platform when it is determined that each document to be reviewed does not have multiple preset types of problems.

The calling module is used to call the robotic process automation RPA robot to access the business system to obtain the contact information of the provider of each document to be reviewed;

The second sending module is used to use RPA robots to feedback the review results of each document to be reviewed to the corresponding provider through contact information.

In one embodiment of the present disclosure, the target business matter is a pharmaceutical registration matter.

It should be noted that the aforementioned explanation of the embodiment of the document review method that combines RPA and AI to implement IA is also applicable to the document review device that combines RPA and AI to implement IA in this embodiment. The embodiment of the present disclosure combines RPA and AI to implement IA. The unpublished details of the document review device embodiment will not be described again here.

In summary, the document review device that combines RPA and AI to implement IA in the embodiment of the present disclosure obtains at least one document to be reviewed corresponding to the target business matter; each document to be reviewed is reviewed based on AI technology to determine whether each document to be reviewed exists. Problems of multiple preset types; when it is determined that each document to be reviewed has at least one problem of the preset type, modification suggestion information corresponding to the document to be reviewed is generated. As a result, it is possible to automatically review documents to be reviewed corresponding to target business matters based on AI technology, reducing the labor costs required for document review and improving the efficiency of document review. In addition, by generating modification suggestion information corresponding to the document to be audited when it is determined that each document to be audited has at least one preset type of problem, modification suggestions can be provided to the provider of the document to be audited, and it is convenient for the provider to modify the document to be audited. Revise.

In order to implement the above embodiments, embodiments of the present disclosure also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, As described in any of the foregoing method embodiments, the document review method of IA is implemented by combining RPA and AI.

In order to implement the above embodiments, embodiments of the present disclosure also provide a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the combination of RPA and AI as described in any of the foregoing method embodiments is implemented. Implement IA's document review method.

In order to implement the above embodiments, embodiments of the present disclosure also provide a computer program product. When the instruction processor in the computer program product is executed, the implementation of IA by combining RPA and AI as described in any of the foregoing method embodiments is realized. Document review methods.

In order to implement the above embodiments, embodiments of the present disclosure also provide a computer program. The computer program includes computer program code. When the computer program code is run on a computer, it causes the computer to perform the combination described in any embodiment of the present disclosure. RPA and AI implement IA's document review method.

7 illustrates a block diagram of an exemplary electronic device suitable for implementing embodiments of the present disclosure. The electronic device 10 shown in FIG. 7 is only an example and should not bring any limitations to the functions and scope of use of the embodiments of the present disclosure.

As shown in Figure 7, electronic device 10 is embodied in the form of a general computing device. The components of electronic device 10 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and a bus 18 connecting various system components, including memory 28 and processing unit 16.

Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics accelerated port, a processor, or a local bus using any of a variety of bus structures. For example, these architectures include but are not limited to Industry Standard Architecture (hereinafter referred to as: ISA) bus, Micro Channel Architecture (Micro Channel Architecture; hereafter referred to as: MAC) bus, enhanced ISA bus, video electronics Standards Association (Video Electronics Standards Association; hereinafter referred to as: VESA) local bus and Peripheral Component Interconnection (hereinafter referred to as: PCI) bus.

In some embodiments, electronic device 10 includes a variety of computer system readable media. These media may be any available media that can be accessed by electronic device 10, including volatile and nonvolatile media, removable and non-removable media.

The memory 28 may include computer system-readable media in the form of volatile memory, such as random access memory (Random Access Memory; hereinafter referred to as: RAM) 30 and/or cache memory 32 . Electronic equipment10 Other removable/non-removable, volatile/non-volatile computer system storage media may further be included. By way of example only, storage system 34 may be used to read and write to non-removable, non-volatile magnetic media (not shown in Figure 7, commonly referred to as a "hard drive"). Although not shown in FIG. 7, a disk drive may be provided for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and a disk drive for reading and writing a removable non-volatile optical disk (e.g., a compact disk read-only memory). A disc drive that reads and writes from Disc Read Only Memory (hereinafter referred to as: CD-ROM), Digital Video Disc Read Only Memory (hereinafter referred to as: DVD-ROM) or other optical media). In these cases, each drive may be connected to bus 18 through one or more data media interfaces. Memory 28 may include at least one program product having a set (eg, at least one) of program modules configured to perform the functions of embodiments of the present disclosure.

A program/utility 40 having a set of (at least one) program modules 42, including but not limited to an operating system, one or more application programs, other program modules, and program data, may be stored, for example, in memory 28 , each of these examples or some combination may include the implementation of a network environment. Program modules 42 generally perform functions and/or methods in the embodiments described in this disclosure.

Electronic device 10 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), may also communicate with one or more devices that enable a user to interact with electronic device 10, and/or with Any device (eg, network card, modem, etc.) that enables the electronic device 10 to communicate with one or more other computing devices. This communication may occur through input/output (I/O) interface 22. Moreover, the electronic device 10 can also communicate with one or more networks (such as a local area network (Local Area Network; hereinafter referred to as: LAN), a wide area network (Wide Area Network; hereinafter referred to as: WAN)) and/or a public network, such as the Internet, through the network adapter 20 ) communication. As shown in FIG. 7 , network adapter 20 communicates with other modules of electronic device 10 via bus 18 . It should be understood that, although not shown in Figure 7, other hardware and/or software modules may be used in conjunction with electronic device 10, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tapes drives and data backup storage systems, etc.

The processing unit 16 executes programs stored in the memory 28 to perform various functional applications and data processing, such as implementing the methods mentioned in the previous embodiments.

It should be noted that the foregoing explanations of the method and device embodiments also apply to the electronic equipment, computer-readable storage media, computer program products and computer programs of the above-mentioned embodiments, and will not be described again here.

In the description of this specification, reference to the terms "one embodiment,""someembodiments,""anexample,""specificexamples," or "some examples" or the like means that specific features are described in connection with the embodiment or example. , structures, materials, or features are included in at least one embodiment or example of the present disclosure. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the specific features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. In addition, to the extent not inconsistent with each other, the technical The skilled person may combine and combine the different embodiments or examples and features of different embodiments or examples described in this specification.

In addition, the terms “first” and “second” are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Therefore, features defined as "first" and "second" may explicitly or implicitly include at least one of these features. In the description of the embodiments of the present disclosure, "plurality" means at least two, such as two, three, etc., unless otherwise explicitly and specifically limited.

Any process or method descriptions in flowcharts or otherwise described herein may be understood to represent modules, segments, or portions of code that include one or more executable instructions for implementing customized logical functions or steps of the process. , and the scope of the preferred embodiments of the present disclosure includes additional implementations in which functions may be performed out of the order shown or discussed, including in a substantially simultaneous manner or in the reverse order, depending on the functionality involved, which shall It should be understood by those skilled in the art to which embodiments of the present disclosure belong.

The logic and/or steps represented in the flowcharts or otherwise described herein, for example, may be considered a sequenced list of executable instructions for implementing the logical functions, and may be embodied in any computer-readable medium, For use by, or in combination with, instruction execution systems, devices or devices (such as computer-based systems, systems including processors or other systems that can fetch instructions from and execute instructions from the instruction execution system, device or device) or equipment. For the purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections with one or more wires (electronic device), portable computer disk cartridges (magnetic device), random access memory (RAM), Read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, and subsequently edited, interpreted, or otherwise suitable as necessary. process to obtain the program electronically and then store it in computer memory.

It should be understood that various parts of the present disclosure may be implemented in hardware, software, firmware, or combinations thereof. In the above embodiments, various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if it is implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: discrete logic gate circuits with logic functions for implementing data signals; Logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

Those of ordinary skill in the art can understand that all or part of the steps involved in implementing the methods of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. The program can be stored in a computer-readable storage medium. When executed, one of the steps of the method embodiment or a combination thereof is included.

In addition, each functional unit in various embodiments of the present disclosure may be integrated into one processing module, each unit may exist physically alone, or two or more units may be integrated into one module. The above integrated modules can be implemented in the form of hardware or software function modules. If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

The storage media mentioned above can be read-only memory, magnetic disks or optical disks, etc. Although the embodiments of the present disclosure have been shown and described above, it can be understood that the above-mentioned embodiments are illustrative and should not be construed as limitations of the present disclosure. Those of ordinary skill in the art can make modifications to the above-mentioned embodiments within the scope of the present disclosure. The embodiments are subject to changes, modifications, substitutions and variations.

All embodiments of the present disclosure can be executed alone or in combination with other embodiments, which are considered to be within the scope of protection claimed by the present disclosure.

Claims

A document review method that combines robotic process automation RPA and artificial intelligence AI to realize intelligent automated IA, wherein the method includes:

Obtain at least one document to be reviewed corresponding to the target business matter;

Review each of the documents to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each of the documents to be reviewed; and

Based on determining that each of the documents to be reviewed has at least one problem of the preset type, modification suggestion information corresponding to the document to be reviewed is generated.
The method according to claim 1, wherein the preset type includes an information completion type;

The AI technology is used to review each of the documents to be reviewed to determine whether there are multiple preset types of problems in each of the documents to be reviewed, including:

Obtain the identification of each document to be reviewed;

Based on the target business matter, query the pre-created knowledge graph corresponding to the target business matter to obtain the identification of at least one target document required by the target business matter; and

Compare the identification of each document to be reviewed with the identification of each target document to determine whether each document to be reviewed is complete.
The method according to claim 2, wherein the auditing of each of the documents to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each of the documents to be reviewed also includes:

Based on optical character recognition (OCR) technology, perform text recognition on each of the documents to be reviewed to obtain the text information contained in each of the documents to be reviewed;

Perform information extraction on the text information contained in each of the documents to be reviewed to obtain the fields to be reviewed and the corresponding field values contained in each of the text information;

Based on the target business matters, query the knowledge graph to obtain the required fields in each of the target documents; and

Determining whether there is a target field consistent with the required field in the corresponding target document among the fields to be reviewed included in each of the text information, and determining whether the corresponding field value exists in the target field, To determine whether the information in each of the documents to be reviewed is complete.
The method according to claim 3, wherein the auditing of each of the documents to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each of the documents to be reviewed also includes:

Obtain the same fields to be reviewed in all the text messages; and

Based on the existence of the corresponding field value in the same field to be reviewed, the corresponding field values of the same field to be reviewed are compared to determine whether the information in each of the documents to be reviewed is consistent.
The method according to any one of claims 1 to 4, wherein the preset type comprises a process specification type;

The AI technology is used to review each of the documents to be reviewed to determine whether there are multiple preset types of problems in each of the documents to be reviewed, including:

Based on OCR technology, perform text recognition on each of the documents to be reviewed to obtain the text information contained in each of the documents to be reviewed;

Based on the target business matter, query the pre-created knowledge graph corresponding to the target business matter to obtain the process specification corresponding to the target business matter; and

Based on each of the documents to be reviewed and the text information contained therein, it is determined whether each of the documents to be reviewed satisfies the process specification.
The method according to any one of claims 1 to 5, wherein the preset type includes a writing specification type;

The AI technology is used to review each of the documents to be reviewed to determine whether there are multiple preset types of problems in each of the documents to be reviewed, including:

Based on OCR technology, perform text recognition on each of the documents to be reviewed to obtain the text information contained in each of the documents to be reviewed; and

The text information contained in each of the documents to be reviewed is input into a pre-trained language model, so that through the language model, it is determined whether each of the documents to be reviewed has problems of the writing standard type.
The method according to any one of claims 1 to 6, wherein the method further comprises:

Based on determining that each of the documents to be reviewed does not have problems of the plurality of preset types, each of the documents to be reviewed is sent to a manual review platform.
The method according to any one of claims 1 to 7, wherein the method further comprises:

Call the Robotic Process Automation RPA robot to access the business system to obtain the contact information of the provider of each of the documents to be reviewed; and

The RPA robot is used to feed back the review results of each document to be reviewed to the corresponding provider through the contact information.
The method according to any one of claims 1 to 8, wherein the target business matter is a pharmaceutical registration matter.
A document review device that combines RPA and AI to implement IA, wherein the device includes:

The acquisition module is used to acquire at least one document to be reviewed corresponding to the target business matter;

The review module is used to review each of the documents to be reviewed based on AI technology to determine whether there are multiple preset types of problems in each of the documents to be reviewed; and

A generating module, configured to generate modification suggestion information corresponding to the document to be reviewed based on determining that each of the documents to be reviewed has at least one problem of the preset type.
The device according to claim 10, wherein the preset type includes an information completion type;

The audit module is used for:

Obtain the identification of each document to be reviewed;

Based on the target business matter, query the pre-created knowledge graph corresponding to the target business matter to obtain the identification of at least one target document required by the target business matter; and

Compare the identification of each document to be reviewed with the identification of each target document to determine whether each document to be reviewed is complete.
The device according to claim 11, wherein the audit module is also used to:

Based on optical character recognition (OCR) technology, perform text recognition on each of the documents to be reviewed to obtain the text information contained in each of the documents to be reviewed;

Perform information extraction on the text information contained in each of the documents to be reviewed to obtain the fields to be reviewed and the corresponding field values contained in each of the text information;

Based on the target business matters, query the knowledge graph to obtain the required fields in each of the target documents; and

Determining whether there is a target field consistent with the required field in the corresponding target document among the fields to be reviewed included in each of the text information, and determining whether the corresponding field value exists in the target field, To determine whether the information in each of the documents to be reviewed is complete.
The device according to claim 12, wherein the audit module is further configured to: obtain the same fields to be audited in all the text information; and the corresponding field values exist based on the same fields to be audited, The corresponding field values of the same fields to be reviewed are compared to determine whether the information in each of the documents to be reviewed is consistent.
The device according to any one of claims 10 to 13, wherein the preset type includes a process specification type; the review module is also used to: perform text recognition on each of the documents to be reviewed based on OCR technology , to obtain the text information contained in each of the documents to be reviewed; based on the target business matters, query the knowledge graph corresponding to the pre-created target business matters to obtain the process specifications corresponding to the target business matters; and based on each of the target business matters to be reviewed Review the documents and the text information contained therein to determine whether each of the documents to be reviewed meets the process specifications.
The device according to any one of claims 10 to 14, wherein the preset type includes a writing specification type; the review module is also used to: perform text recognition on each of the documents to be reviewed based on OCR technology , to obtain the text information contained in each document to be reviewed; input the text information contained in each document to be reviewed into a pre-trained language model, so as to determine each document to be reviewed through the language model Does the document contain any issues of the type specified in the document specification?
The device according to any one of claims 10 to 15, wherein the document review device that combines RPA and AI to implement IA further includes: a first sending module for determining that each of the documents to be reviewed does not exist. Describe multiple preset types of questions and send each of the documents to be reviewed to the manual review platform.
The device according to any one of claims 10 to 16, wherein the document review device that combines RPA and AI to implement IA further includes: a calling module for calling a robotic process automation RPA robot to access the business system to obtain each The contact information of the provider of the documents to be reviewed; and a second sending module used to use the RPA robot to feed back the review results of each of the documents to be reviewed to the corresponding provider through the contact information.
The device according to any one of claims 10 to 17, wherein the target business matter is a pharmaceutical registration matter.
An electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, any one of claims 1 to 9 is implemented. the method described.
A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.
A computer program product includes a computer program, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.
A computer program, wherein the computer program includes computer program code, and when the computer program code is run on a computer, the computer performs the method according to any one of claims 1 to 9.