CN117592778A - Risk early warning system - Google Patents
Risk early warning system Download PDFInfo
- Publication number
- CN117592778A CN117592778A CN202311550992.0A CN202311550992A CN117592778A CN 117592778 A CN117592778 A CN 117592778A CN 202311550992 A CN202311550992 A CN 202311550992A CN 117592778 A CN117592778 A CN 117592778A
- Authority
- CN
- China
- Prior art keywords
- risk
- contract
- module
- information
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012502 risk assessment Methods 0.000 claims abstract description 103
- 238000012545 processing Methods 0.000 claims abstract description 100
- 238000013058 risk prediction model Methods 0.000 claims abstract description 53
- 238000012544 monitoring process Methods 0.000 claims description 48
- 238000000034 method Methods 0.000 claims description 42
- 238000004458 analytical method Methods 0.000 claims description 32
- 230000008569 process Effects 0.000 claims description 27
- 230000008451 emotion Effects 0.000 claims description 24
- 230000008859 change Effects 0.000 claims description 22
- 238000004806 packaging method and process Methods 0.000 claims description 16
- 238000009933 burial Methods 0.000 claims description 11
- 238000012795 verification Methods 0.000 claims description 7
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000005538 encapsulation Methods 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 230000000007 visual effect Effects 0.000 claims description 5
- 238000003058 natural language processing Methods 0.000 claims description 4
- 230000000295 complement effect Effects 0.000 claims description 3
- 238000013480 data collection Methods 0.000 claims 1
- 238000013479 data entry Methods 0.000 claims 1
- 238000011156 evaluation Methods 0.000 abstract description 13
- 230000006978 adaptation Effects 0.000 abstract description 4
- 238000012550 audit Methods 0.000 description 8
- 238000012549 training Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 239000003795 chemical substances by application Substances 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 101100264022 Drosophila melanogaster wgn gene Proteins 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 238000010561 standard procedure Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Game Theory and Decision Science (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application provides a risk early warning system. The system comprises: the business processing module is used for processing various contract businesses; the data acquisition module is used for acquiring contract form filling information and acquiring risk analysis information according to the contract form filling information; the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to each dimension risk prediction model according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by each dimension risk prediction model and the corresponding weight factor. The system provided by the embodiment realizes the automatic adaptation of a proper risk evaluation system to different service processing scenes, performs automatic intelligent comprehensive risk evaluation from multiple dimensions, and prompts related service departments in real time, thereby reducing risk occurrence probability or loss caused by risk occurrence.
Description
Technical Field
The application relates to the field of computer technology, in particular to a risk early warning system.
Background
The internal audit is an independent and objective confirmation and consultation activity, and is used for examining and evaluating the appropriateness and effectiveness of business activities, internal control and risk management of organizations by using a systematic and standard method so as to promote the organizations to perfect management and help the enterprises to achieve the aim. The prior audit and the prior audit can timely find and feed back the problems, correct the deviation as early as possible, and carry out scientific decision and control management, thereby ensuring that the economic activities of enterprises are legal, reasonable and effective according to the expected targets.
At present, the traditional monitoring model is audited mainly to reveal problems caused after risks occur, and the prior risk identification capability is insufficient. In the prior risk assessment process, different risk assessment systems should be adopted for the same object because the risk is concerned differently in different business scenes. If the process is performed manually by means of a traditional model, the operation process is extremely complex, the operation amount is huge, the subjective influence of individuals cannot be avoided, and risk judgment deviation is easily caused. Meanwhile, as the current digital level is improved, each business department generally uses a closed system to process business, after an audit supervision department finishes risk assessment, the business department is informed of a wind control post in a traditional way such as mail, telephone and the like, and the risk control requirement is difficult to be implemented to actual operators of a business system in time, so that the risk cannot be restrained in advance.
Therefore, how to design a suitable real-time audit risk early warning method to adapt to the current risk assessment business scenario becomes a problem to be solved urgently.
Disclosure of Invention
The application provides a risk early warning system for reduce risk probability or loss that risk takes place to cause.
The application provides a risk early warning system, the system includes:
the business processing module is used for processing various contract businesses, wherein the various contract businesses comprise contract drafting and contract approval;
the data acquisition module is used for acquiring contract form filling information when the business processing module processes contract business, and acquiring risk analysis information according to the contract form filling information;
the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, inputting the risk analysis information into risk prediction models of different risk dimensions, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by the risk prediction model of each dimension and the corresponding weight factor; the risk prediction model in the same dimension has different corresponding weight factors in different service processing scenes.
Optionally, the data acquisition module comprises a service embedded point engine and a unified message engine; wherein the service burial point engine is buried in the service processing module;
the business embedded point engine is used for collecting contract form filling information of a front-end page when the business processing module processes contract business, and packaging and sending the contract form filling information to the unified message engine;
the unified message engine is used for receiving the contract form filling information sent by the service embedded point engine, acquiring contract side risk data, acquiring risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to the intelligent risk analysis module; the contract risk data comprises at least one of historical contract information, historical cooperation service, historical score and contractor enterprise information.
Optionally, the service burial point engine includes: the system comprises an event acquisition module, a buried point message encapsulation module, a buried point message dispatch module and a buried point message feedback module;
the event acquisition module is used for responding to a triggering event of a user on a front-end page, and packaging the event type, the operation value and the triggering event of the triggering operation to obtain single form filling information;
The embedded point message packaging module is used for packaging a plurality of single form filling information in the contract drafting and contract approving stages to obtain the contract form filling information;
the embedded point message dispatching module is used for sending the contract form filling message to the intelligent risk analysis module and receiving the risk rating sent by the intelligent risk analysis module;
and the buried point message feedback module is used for feeding back the risk rating to the service processing module.
Optionally, the embedded point message feedback module is specifically configured to, if the risk rating is higher than a preset rating, feed back the risk rating and risk prompt information to the service processing module, so that the service processing module displays the risk prompt information.
Optionally, the unified message engine is specifically configured to:
caching the received contract form filling information into a message queue;
sequentially acquiring contract form filling information from a message queue, and checking data integrity and validity;
and after the verification is passed, acquiring contractual risk data, obtaining risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to an intelligent risk analysis module.
Optionally, the service burial point engine further includes:
the flow event monitoring module is used for monitoring the state change of the contract business in the link according to a predefined monitoring interface, and the monitoring interface is configured to monitor the monitored object and the monitoring event and output the contract state change message obtained by monitoring.
Optionally, the intelligent risk analysis module comprises a data transmission module, an identification and scene classification module, a model prediction module, a risk weight adjustment module and a risk identification output module;
the data input module is used for receiving risk analysis information sent by the unified message engine;
the identification and scene classification module is used for intelligently processing the risk analysis information through a natural language processing technology to obtain a service processing scene corresponding to the current contract service;
the model prediction module is used for inputting the risk analysis information into risk prediction models of different risk dimensions and obtaining a risk score output by each risk prediction model;
the risk weight adjustment module is used for acquiring weight factors corresponding to the risk prediction models of each dimension aiming at different types of service processing scenes;
And the risk identification output module is used for obtaining the risk rating corresponding to the contract business according to the risk scores and the corresponding weight factors output by the risk prediction model of each dimension.
Optionally, the recognition and scene classification module comprises at least one of a semantic analysis unit, an entity recognition unit or an emotion analysis unit, and a scene classification module;
the semantic analysis unit is used for identifying a theme in the risk analysis information through an LDA probability model and/or an NLU algorithm;
the entity identification unit is used for identifying the named entity in the risk analysis information through an entity identification algorithm;
the emotion analysis unit is used for identifying emotion in the risk analysis information through a machine learning algorithm;
the scene classification module is used for determining a service processing scene according to one of the theme, the entity or the emotion.
Optionally, the risk prediction model is an XGBoost-based prediction model.
Optionally, the risk early warning system further includes: a risk convergence module, the risk convergence module comprising:
a flow event unit for storing the contract state change message to a message queue;
The risk event filtering unit is used for acquiring the risk rating corresponding to the contract state change message in the message queue, storing the risk event data set if the risk rating is higher than a preset risk water line, and discarding the contract state change message if the risk rating is not higher than the preset risk water line.
The risk event backtracking unit is used for backtracking the contract business in the risk event data set so as to complement the contract business missing link according to the log of the contract business;
and the risk event analysis unit is used for carrying out visual analysis according to the risk event data set.
The application provides a risk early warning system, include: the business processing module is used for processing various contract businesses, wherein the various contract businesses comprise contract drafting and contract approval; the data acquisition module is used for acquiring contract form filling information when the business processing module processes contract business, and acquiring risk analysis information according to the contract form filling information; the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, inputting the risk analysis information into risk prediction models of different risk dimensions, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by the risk prediction model of each dimension and the corresponding weight factor; the risk prediction model in the same dimension has different corresponding weight factors in different service processing scenes. The system provided by the embodiment realizes the automatic adaptation of a proper risk evaluation system to different service processing scenes, performs automatic intelligent comprehensive risk evaluation from multiple dimensions, and prompts related service departments in real time, thereby reducing risk occurrence probability or loss caused by risk occurrence.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.
Fig. 1 is a schematic structural diagram of a risk early warning system provided in an embodiment of the present application;
fig. 2 is a schematic diagram of a risk early warning system according to a second embodiment of the present disclosure.
Specific embodiments thereof have been shown by way of example in the drawings and will herein be described in more detail. These drawings and the written description are not intended to limit the scope of the inventive concepts in any way, but to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as detailed in the accompanying claims.
In this application, the terms "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.
Aiming at the problems that in the actual enterprise contract signing process, a risk evaluation system is difficult to automatically match according to a service scene, related risks are difficult to prompt in advance by combining the service scene, and the like, the risk early warning system is designed, the embedded point deployment is carried out on the contract system, the form filling information in the contract system is collected in real time, the historical contract information, the enterprise information and the like of a partner company are collected for intelligent risk analysis, the risk evaluation system which accords with the contract service scene is automatically adapted by combining the actual service scene, the multi-dimensional intelligent risk judgment is carried out on the partner, and the risk prompt is carried out in real time.
The following describes the technical solutions of the present application and how the technical solutions of the present application solve the above technical problems in detail with specific embodiments. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.
Fig. 1 is a schematic structural diagram of a risk early-warning system provided in an embodiment of the present application, as shown in fig. 1, where the risk early-warning system includes a service processing module, a data acquisition module, and an intelligent risk analysis module; wherein:
the business processing module is used for processing various contract businesses, wherein the various contract businesses comprise contract drafting and contract approval;
the data acquisition module is used for acquiring contract form filling information when the business processing module processes contract business, and acquiring risk analysis information according to the contract form filling information;
the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, inputting the risk analysis information into risk prediction models of different risk dimensions, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by the risk prediction model of each dimension and the corresponding weight factor; the risk prediction model in the same dimension has different corresponding weight factors in different service processing scenes.
In this embodiment, the service processing module is configured to process a plurality of contract services, which may be contract systems for making contracts for enterprises, and the contract services may include contract drafting and contract approval in the contract systems.
The data acquisition module is used for acquiring contract form filling information when the service processing module processes contract service. Specifically, the business monitoring point can be embedded in the business processing module to collect the form filling information of the contract in real time, as shown in fig. 2, the business monitoring point can be embedded in the drafting stage of the contract to collect form filling information such as the province, place, signing department, contract client, contract amount, contract type and the like of the contract, the business monitoring point is embedded in the examining stage of the contract to collect form filling information such as whether the contract is continuously signed or not. Further, the data acquisition module acquires risk analysis data based on contract form filling information, and sends the risk analysis data to the intelligent risk analysis module for analysis and processing.
The business processing scene is used for indicating the type of the contract which is currently signed, for example, the type of the contract which is signed is the contract of the operation and maintenance class; the risk dimension refers to a plurality of dimensions such as model monitoring, history problem discovery, business information, associated enterprise risks and the like; and each risk dimension corresponds to one risk prediction model, data of different business types is input into the corresponding risk prediction model, and a risk score of each risk dimension can be obtained.
Specifically, the intelligent risk analysis module analyzes and processes the risk analysis information transmitted by the data acquisition module to obtain a service processing scene corresponding to the current contract service, inputs the risk analysis information into risk prediction models of different risk dimensions to obtain a risk score output by each risk prediction model, further obtains a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and then obtains a risk rating corresponding to the current contract service according to the risk score and the corresponding weight factor output by the risk prediction model of each dimension, for example, the risk score and the corresponding weight factor output by the risk prediction model of each dimension can be weighted and summed to obtain the risk rating corresponding to the current contract service.
It should be noted that, because some risk dimensions may be more important in some business scenarios, the corresponding weight factors will be larger, so that the risk prediction model in the same dimension has different weight factors in different business processing scenarios.
The risk early warning system provided in this embodiment includes: the business processing module is used for processing various contract businesses, wherein the various contract businesses comprise contract drafting and contract approval; the data acquisition module is used for acquiring contract form filling information when the business processing module processes contract business, and acquiring risk analysis information according to the contract form filling information; the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, inputting the risk analysis information into risk prediction models of different risk dimensions, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by the risk prediction model of each dimension and the corresponding weight factor; the risk prediction model in the same dimension has different corresponding weight factors in different service processing scenes. The system provided by the embodiment realizes the automatic adaptation of a proper risk evaluation system to different service processing scenes, performs automatic intelligent comprehensive risk evaluation from multiple dimensions, and prompts related service departments in real time, thereby reducing risk occurrence probability or loss caused by risk occurrence.
Fig. 2 is a schematic diagram of a risk early warning system provided in the embodiment of the present application, where, based on the above embodiment, a data acquisition module and an intelligent risk analysis module are described in detail, as shown in fig. 2, the data acquisition module includes a service embedded point engine and a unified message engine; wherein the service burial point engine is buried in the service processing module;
the business embedded point engine is used for collecting contract form filling information of a front-end page when the business processing module processes contract business, and packaging and sending the contract form filling information to the unified message engine;
the unified message engine is used for receiving the contract form filling information sent by the service embedded point engine, acquiring contract side risk data, acquiring risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to the intelligent risk analysis module; the contract risk data comprises at least one of historical contract information, historical cooperation service, historical score and contractor enterprise information.
Specifically, the data acquisition module is mainly used for acquiring and analyzing the form filling information. By embedding the service monitoring point in the service processing module, the contract form filling information is collected in real time, for example, the drafting information and the approval information are returned in the contract drafting stage and the contract approval stage. After receiving the contract form filling information, the data acquisition module combines the pre-collected contract side risk data and pushes the contract side risk data to the intelligent risk analysis module. After intelligent risk analysis, pushing the result to a business processing module.
The service burial point engine is mainly responsible for collecting contract form filling information of a front page of the contract system when the service processing module processes contract service, and packaging and distributing the form filling information to the unified message engine.
In a specific embodiment, the service burial point engine includes: the system comprises an event acquisition module, a buried point message encapsulation module, a buried point message dispatch module and a buried point message feedback module;
the event acquisition module is used for responding to a triggering event of a user on a front-end page, and packaging the event type, the operation value and the triggering event of the triggering operation to obtain single form filling information;
the embedded point message packaging module is used for packaging a plurality of single form filling information in the contract drafting and contract approving stages to obtain the contract form filling information;
the embedded point message dispatching module is used for sending the contract form filling message to the intelligent risk analysis module and receiving the risk rating sent by the intelligent risk analysis module;
and the buried point message feedback module is used for feeding back the risk rating to the service processing module.
The event collection module captures the operation of the user on the front-end page through DOM event monitoring, and performs the operation of preliminary format encapsulation on the original data. A global message processor is first designed that can identify the particular event generated by the DOM element. Specifically, DOM elements and corresponding event types that need to be listened to are first defined, and then registered in the global message processor so that they can be captured when they occur on the front-end page. When an event occurs on an element, the global message processor cleans and reconstructs the value generated by the event and encapsulates it into a message object. The message object contains all necessary information so that the server can accurately understand the operation of the user. For example, the DOM type, DOM element identification, operation value, trigger time, and the like of the trigger event are included. In this way, the operation of the user on the front page is captured, and single form filling information is obtained.
Further, the embedded point message encapsulation module is triggered in two stages of contract drafting and contract approval. The module comprises two steps of service data cleaning and unified message packaging, and is used for packaging a plurality of single form filling information in the contract drafting and contract approval stages to obtain the contract form filling information.
In the specific implementation process, the single form filling information generated in the DOM event acquisition process is cleaned. And removing null values and repeated data, and merging similar data, such as two new and old contract body names of the contract body change operation records. And then checking the cleaned data, such as data types, data ranges and the like, so as to remove illegal data and perform fault-tolerant processing in time.
And the unified message packaging is to package the cleaned single form filling information into contract form filling information and send the contract form filling information to the intelligent risk analysis module. One piece of packaged contract form filling information can completely represent all single form filling information of one business operation. In addition, the contract form filling information contains all the single form filling information, and further adds contents such as an operating system identifier, a package time stamp, a browser of the current system login, an IP address, a digital signature of service operation information and the like so as to ensure the integrity and the validity of data.
Further, the embedded point message dispatching module sends the contract form filling message to the unified message engine, the unified message engine combines the contract side risk data, the incoming contract form filling message is packaged to obtain risk analysis information, then the unified message engine sends the risk analysis information to the intelligent risk analysis module, the intelligent risk analysis module carries out intelligent processing and analysis to obtain the current risk rating, the embedded point message dispatching module receives the risk rating sent by the intelligent risk analysis module, and finally the embedded point message feedback module feeds back the risk rating to the business processing module. Specifically, if the risk rating is higher than the preset rating, the buried point message feedback module feeds back the risk rating and the risk prompt information to the service processing module, so that the service processing module displays the risk prompt information.
Optionally, the service burial point engine further includes:
the flow event monitoring module is used for monitoring the state change of the contract business in the link according to a predefined monitoring interface, and the monitoring interface is configured to monitor the monitored object and the monitoring event and output the contract state change message obtained by monitoring.
In a specific implementation process, the process event monitoring module defines a monitoring interface firstly, and the monitoring interface is used for defining a data exchange format and a transmission mode of monitoring information between the data acquisition module and the service processing module. The monitoring interface comprises definitions of monitoring objects, monitoring events, monitoring data and the like, so that the data acquisition module can effectively monitor the flow change of the service processing module.
In order to interface the monitoring interface with the service processing module, a monitoring agent needs to be implemented in the service processing module, and the monitoring agent is used for receiving the monitoring information sent by the data acquisition module and converting the monitoring information into an event and a data format in the service processing module.
Furthermore, the object to be monitored is registered in the service processing module, so that the monitoring agent can effectively monitor the object. The monitoring object is typically a specific business process or operation in a business processing module, such as contract drafting, contract auditing, etc., that requires that a unique identifier of the monitoring object and a monitoring event be specified at registration. When a specified monitoring event occurs to a monitoring object, the monitoring agent captures the event, converts the event into an event format defined in a monitoring interface, and outputs the event through the monitoring interface.
In a specific embodiment, the unified message engine is specifically configured to:
caching the received contract form filling information into a message queue;
sequentially acquiring contract form filling information from a message queue, and checking data integrity and validity;
and after the verification is passed, acquiring contractual risk data, obtaining risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to an intelligent risk analysis module.
Specifically, the contract form filling information submitted by the service processing module is cached in a message queue mode, the unified message engine sequentially acquires the contract form filling information from the message queue, and data integrity and validity verification is carried out, wherein verification contents comprise key information such as that the contract flow is empty, the client code is empty, the contract type is empty and the like, or whether the contract key word information is valid or not. After verification, the information of the contract form is completed, for example, various wind control related data including the cooperators (such as management data, enterprise data and other dimension risk data of the company), employee information (such as dimension risk data of employee departments, ground markets, channels and the like), contract text and the like can be obtained from a third party service system. And then, corresponding information fields are associated through conditions such as names, unique identifiers and the like, risk analysis information sent by the business processing module is supplemented and perfected, and the intelligent risk analysis module is pushed to process the risk analysis information.
Optionally, the unified message engine further includes:
the log recording module is used for recording contract form filling information requested by all contract systems so as to be backtracked and reproduced in the later period; recording risk analysis information after information complementation, and assisting in optimizing the evaluation accuracy of the intelligent risk analysis module; recording an intelligent risk analysis result for system risk capability analysis capability; and recording risk feedback information to form a risk evaluation data set for the risk aggregation module to trace back risk items.
The module is mainly realized by the following steps:
creating a section class, realizing a section interface in the AOP, and adding an annotation to identify the section;
in the section class, a method is written for recording log information. The parameters of the method should contain a JoinPoint object, the object contains the information of the currently executed method, and the information of the name, parameters and the like of the method can be acquired through the object;
in the Spring configuration file, declaring the section class, and configuring an entry point expression to specify which methods need to record logs;
on the method requiring logging, a section annotation is added to specify the use of the section. When a method requiring logging is called, the tangent plane automatically intercepts the method and executes the logging method.
As shown in fig. 2, the intelligent risk analysis module comprises a data transmission module, an identification and scene classification module, a model prediction module, a risk weight adjustment module and a risk identification output module;
the data input module is used for receiving risk analysis information sent by the unified message engine;
the identification and scene classification module is used for intelligently processing the risk analysis information through a natural language processing technology to obtain a service processing scene corresponding to the current contract service;
the model prediction module is used for inputting the risk analysis information into risk prediction models of different risk dimensions and obtaining a risk score output by each risk prediction model;
the risk weight adjustment module is used for acquiring weight factors corresponding to the risk prediction models of each dimension aiming at different types of service processing scenes;
and the risk identification output module is used for obtaining the risk rating corresponding to the contract business according to the risk scores and the corresponding weight factors output by the risk prediction model of each dimension.
The data input module acquires risk analysis information sent by the unified message engine, including income contracts, expenditure contracts, enterprise transaction flow, historical contract information, historical cooperation business, historical scores, historical evaluation, cooperation party information and the like.
Further, the recognition and scene classification module performs intelligent processing on the risk analysis information through a natural language processing technology to acquire a service processing scene corresponding to the current contract service.
Optionally, the recognition and scene classification module comprises at least one of a semantic analysis unit, an entity recognition unit or an emotion analysis unit, and a scene classification module;
the semantic analysis unit is used for identifying a theme in the risk analysis information through an LDA probability model and/or an NLU algorithm;
the entity identification unit is used for identifying the named entity in the risk analysis information through an entity identification algorithm;
the emotion analysis unit is used for identifying emotion in the risk analysis information through a machine learning algorithm;
the scene classification module is used for determining a service processing scene according to one of the theme, the entity or the emotion.
In the specific implementation process, the semantic analysis unit performs semantic analysis on the risk analysis information through an LDA probability model and/or an NLU algorithm, analyzes sentence structures, word senses and the like, and extracts key information to identify topics in the risk analysis information. Specifically, the LDA probability model, that is, the latent dirichlet allocation algorithm, is a text topic model based on a probability model, and is used for discovering topics or topic distribution in a document set. NLU algorithms, natural language understanding, are used to extract meaning from complex sentences, understand context, and parse and interpret the grammar and semantics of the language.
And the entity identification unit is used for identifying the named entity in the risk analysis information through an entity identification algorithm.
By way of example, the entity recognition algorithm may be a bi-directional encoder representation (BERT) of Conditional Random Fields (CRFs) and morphers to identify named entities in text data, where:
the goal of the CRF algorithm is to find the most likely output tag sequence given the input feature sequence. This objective can be achieved by maximizing the conditional probability between the input feature sequence and the tag sequence. The conditional probability may be calculated using the local potential and the transition potential. The local potentials depend on interactions between the input features and the labels, while the transitional potentials depend on interactions between adjacent labels. By combining the local and transitional potentials, the CRF can capture complex patterns in the text, thereby more accurately identifying named entities. In implementing the CRF algorithm, the Viterbi algorithm or the Forward-Backward algorithm may be used to find the most likely output tag sequence. The Viterbi algorithm finds the most likely tag sequence by dynamic programming, while the Forward-Backward algorithm finds the most likely tag sequence by calculating the Forward and Backward probabilities.
BERT is a deep neural network model that can use a converter architecture to encode the context of each word in a text sequence. The BERT model is pre-trained on a large corpus of text data using a mask language modeling target and a next sentence prediction target. By adding a task-specific output layer and performing end-to-end training of the entire model on the label data, downstream tasks (e.g., NER) can be fine-tuned.
And the emotion analysis unit is used for identifying emotion in the risk analysis information through a machine learning algorithm. Specifically, the emotion analysis unit may perform emotion analysis on the text data using a supervised machine learning algorithm, such as a support vector machine (svm) and a naive bayes classifier. Dictionary-based methods, such as price-aware dictionaries and emotion inferors (VADERs), may also be used to identify specific emotions in text.
Illustratively, the dictionary-based approach uses a predefined emotion dictionary that contains a list of words with associated polarity scores (e.g., positive, negative, neutral) to calculate the overall emotion score of the text. The emotion score may be calculated using various methods, such as simple summation, weighted summation, or normalized scoring.
The deep learning method learns emotion representations of text from the marker data using neural network models, such as convolutional neural network (cnn) and recurrent neural network (rnn). The neural network model may be trained using various techniques, such as supervised learning, semi-supervised learning, or transfer learning.
Further, the scene classification module determines a business processing scene according to one of the subject, the entity or the emotion. It should be noted that, the theme acquired by the semantic analysis unit may be different from the service processing scenario corresponding to the entity identified by the entity identification unit, and the emotion analysis needs to determine the service processing scenario more in line with the actual service.
The risk analysis information comprises dimension data such as historical transaction records, bad account verification, arrearage records, blacklist records, multi-place account opening, project contracts, bidding, business data and the like, and after the multi-dimension data in the risk analysis information are input into risk prediction models in different dimensions, risk scores output by each risk prediction model are obtained, wherein the risk prediction models can be prediction models based on XGBoost.
Exemplary, XGBoost machine learning model building is performed on the historical audit discovery problem risk dimension, so as to predict risk scores of the audit discovery problem risk dimension, and the model building steps are as follows:
First, the data are subjected to a washing and preprocessing operation: preprocessing original sample data and index features, including removing repeated data, abnormal data and invalid data, correcting error data, filling missing values (the missing values are filled with 0) according to conditions, and converting category variables into numerical variables.
Splitting the processed sample set into a training sample set and a test sample set, establishing a following model by using the response variable and interpretation variable relation of the training sample set, and optimizing model parameters by using the test sample set.
Wherein F represents the space of the regression tree (CART tree), Φ (·) represents the XGBoost model,represents x i And (5) corresponding audit discovery problem risk scores.
Further, constructing a loss function based on the risk score prediction result and the true value, and evaluating the prediction capability and stability of the XGBoost model according to the loss function, so that multiple rounds of parameter adjustment are performed on the XGBoost model, wherein the adjustable parameters comprise the maximum iteration times, the number of sub-models, the learning rate, the loss function and the maximum depth. The evaluation index of XGBoost regression model is as follows:
wherein MAE represents the mean absolute error; RMSE represents root mean square error; r is R 2 Representing the decision coefficients; y is i A target value representing an ith partner; Representing a predicted value of an ith partner; />Representing the mean; n represents the number of cooperators;
the training mode of the XGBoost model by using the training set is mainly K-fold cross validation. After training the XGBoost model to expectations, it can be deployed in a specific environment, outputting risk scores of different dimensions of the partner on a monthly regular basis. And the prediction effect of the model can be monitored in the using process, and the XGBoost model is updated periodically according to the prediction effect.
Optionally, the risk early warning system further includes a risk convergence module, as shown in fig. 2, where the risk convergence module includes:
a flow event unit for storing the contract state change message to a message queue;
the risk event filtering unit is used for acquiring the risk rating corresponding to the contract state change message in the message queue, storing the risk event data set if the risk rating is higher than a preset risk water line, and discarding the contract state change message if the risk rating is not higher than the preset risk water line.
The risk event backtracking unit is used for backtracking the contract business in the risk event data set so as to complement the contract business missing link according to the log of the contract business;
and the risk event analysis unit is used for carrying out visual analysis according to the risk event data set.
In a specific embodiment, the risk convergence module is mainly configured to receive the contract state change information returned by the flow event monitoring unit. The complete risk contract responsibility chain is formed and recorded through the processes of cleaning, backtracking and the like of the checked contract, wherein the complete risk contract responsibility chain comprises contract information, drafting person information and approver information. And visually analyzing the contract signing data in a period of time in the form of a bar graph, a pie chart and the like. The module comprises a flow event unit, a risk event filtering unit, a risk event backtracking unit and a risk event analysis unit.
The flow event unit records the contract state change notification in a message queue mode so as to relieve the system pressure in a large concurrency scene and improve the expansibility and flexibility of the service system. When the module receives the state change notification pushed by the contract system, the module can complete the processing by only inserting the state change notification into a message queue.
The risk event filtering unit firstly acquires a contract approval state change type message in the message queue, and then backtracks the risk evaluation corresponding to the contract recorded by the data acquisition module. If the risk rating of the contract is higher than the preset risk water line, the message is stored in the risk item data set, otherwise, the message is discarded.
And for the contracts in the risk event data set, the module pulls full contract sketch and approval full flow data from the logs recorded by the data acquisition module according to contract running water, so that a complete risk contract responsibility chain is formed.
The risk event analysis unit performs visual analysis on contract signing data in a period of time through a histogram, a pie chart and other forms, wherein the visual analysis comprises counting the total quantity and detail of contract signing, the total quantity and detail of contract risk matters and the total quantity and detail of contract signing and active termination.
Finally, it should be noted that: other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It is to be understood that the present application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (10)
1. A risk early warning system, comprising:
the business processing module is used for processing various contract businesses, wherein the various contract businesses comprise contract drafting and contract approval;
the data acquisition module is used for acquiring contract form filling information when the business processing module processes contract business, and acquiring risk analysis information according to the contract form filling information;
the intelligent risk analysis module is used for analyzing and processing the risk analysis information, acquiring a service processing scene corresponding to the current contract service, inputting the risk analysis information into risk prediction models of different risk dimensions, acquiring a risk score output by each risk prediction model, acquiring a weight factor corresponding to the risk prediction model of each dimension according to the service processing scene, and acquiring a risk rating corresponding to the contract service according to the risk score output by the risk prediction model of each dimension and the corresponding weight factor; the risk prediction model in the same dimension has different corresponding weight factors in different service processing scenes.
2. The system of claim 1, wherein the data collection module comprises a service burial point engine and a unified message engine; wherein the service burial point engine is buried in the service processing module;
The business embedded point engine is used for collecting contract form filling information of a front-end page when the business processing module processes contract business, and packaging and sending the contract form filling information to the unified message engine;
the unified message engine is used for receiving the contract form filling information sent by the service embedded point engine, acquiring contract side risk data, acquiring risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to the intelligent risk analysis module; the contract risk data comprises at least one of historical contract information, historical cooperation service, historical score and contractor enterprise information.
3. The system of claim 2, wherein the service burial point engine comprises: the system comprises an event acquisition module, a buried point message encapsulation module, a buried point message dispatch module and a buried point message feedback module;
the event acquisition module is used for responding to a triggering event of a user on a front-end page, and packaging the event type, the operation value and the triggering event of the triggering operation to obtain single form filling information;
the embedded point message packaging module is used for packaging a plurality of single form filling information in the contract drafting and contract approving stages to obtain the contract form filling information;
The embedded point message dispatching module is used for sending the contract form filling message to the intelligent risk analysis module and receiving the risk rating sent by the intelligent risk analysis module;
and the buried point message feedback module is used for feeding back the risk rating to the service processing module.
4. The system of claim 3, wherein the embedded point message feedback module is specifically configured to, if the risk rating is higher than a preset rating, feed back the risk rating and risk prompt information to the service processing module, so that the service processing module displays the risk prompt information.
5. The system according to claim 2, wherein the unified message engine is specifically configured to:
caching the received contract form filling information into a message queue;
sequentially acquiring contract form filling information from a message queue, and checking data integrity and validity;
and after the verification is passed, acquiring contractual risk data, obtaining risk analysis information according to the contract form filling information and the contract risk data, and pushing the risk analysis information to an intelligent risk analysis module.
6. The system of claim 3, wherein the service burial point engine further comprises:
The flow event monitoring module is used for monitoring the state change of the contract business in the link according to a predefined monitoring interface, and the monitoring interface is configured to monitor the monitored object and the monitoring event and output the contract state change message obtained by monitoring.
7. The system of claim 1, wherein the intelligent risk analysis module comprises a data entry module, an identification and scene classification module, a model prediction module, a risk weight adjustment module, and a risk identification output module;
the data input module is used for receiving risk analysis information sent by the unified message engine;
the identification and scene classification module is used for intelligently processing the risk analysis information through a natural language processing technology to obtain a service processing scene corresponding to the current contract service;
the model prediction module is used for inputting the risk analysis information into risk prediction models of different risk dimensions and obtaining a risk score output by each risk prediction model;
the risk weight adjustment module is used for acquiring weight factors corresponding to the risk prediction models of each dimension aiming at different types of service processing scenes;
And the risk identification output module is used for obtaining the risk rating corresponding to the contract business according to the risk scores and the corresponding weight factors output by the risk prediction model of each dimension.
8. The system of claim 4, wherein the recognition and scene classification module comprises at least one of a semantic analysis unit, an entity recognition unit, or an emotion analysis unit, and a scene classification module;
the semantic analysis unit is used for identifying a theme in the risk analysis information through an LDA probability model and/or an NLU algorithm;
the entity identification unit is used for identifying the named entity in the risk analysis information through an entity identification algorithm;
the emotion analysis unit is used for identifying emotion in the risk analysis information through a machine learning algorithm;
the scene classification module is used for determining a service processing scene according to one of the theme, the entity or the emotion.
9. The system of claim 1, wherein the risk prediction model is an XGBoost-based prediction model.
10. The system of claim 5, wherein the risk early warning system further comprises: a risk convergence module, the risk convergence module comprising:
A flow event unit for storing the contract state change message to a message queue;
the risk event filtering unit is used for acquiring a risk rating corresponding to the contract state change message in the message queue, storing a risk event data set if the risk rating is higher than a preset risk water line, otherwise discarding the contract state change message;
the risk event backtracking unit is used for backtracking the contract business in the risk event data set so as to complement the contract business missing link according to the log of the contract business;
and the risk event analysis unit is used for carrying out visual analysis according to the risk event data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311550992.0A CN117592778A (en) | 2023-11-20 | 2023-11-20 | Risk early warning system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311550992.0A CN117592778A (en) | 2023-11-20 | 2023-11-20 | Risk early warning system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117592778A true CN117592778A (en) | 2024-02-23 |
Family
ID=89919414
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311550992.0A Pending CN117592778A (en) | 2023-11-20 | 2023-11-20 | Risk early warning system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117592778A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118171920A (en) * | 2024-05-15 | 2024-06-11 | 山东浪潮智慧建筑科技有限公司 | LLM model-based park safety emergency response method, device and medium |
-
2023
- 2023-11-20 CN CN202311550992.0A patent/CN117592778A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118171920A (en) * | 2024-05-15 | 2024-06-11 | 山东浪潮智慧建筑科技有限公司 | LLM model-based park safety emergency response method, device and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20150269495A1 (en) | Methods and system for providing simultaneous multi-task ensemble learning | |
| CN110598070B (en) | Application type identification method and device, server and storage medium | |
| CN117592778A (en) | Risk early warning system | |
| CN113298121B (en) | Message sending method and device based on multi-data source modeling and electronic equipment | |
| Djedovic et al. | Innovative approach in modeling business processes with a focus on improving the allocation of human resources | |
| CN115565038A (en) | Content audit, content audit model training method and related device | |
| CN113674846A (en) | Hospital intelligent service public opinion monitoring platform based on LSTM network | |
| CN115809796B (en) | Project intelligent dispatching method and system based on user portrait | |
| CN115438190B (en) | Power distribution network fault auxiliary decision knowledge extraction method and system | |
| CN111241288A (en) | Emergency sensing system of large centralized power customer service center and construction method | |
| CN112712270B (en) | Information processing method, device, equipment and storage medium | |
| CN114797114A (en) | Real-time intelligent identification method and system for game chat advertisement | |
| CN108491481B (en) | Intelligent identification active early warning system and method for party involved in intellectual property right infringement | |
| CN116993307B (en) | Collaborative office method and system with artificial intelligence learning capability | |
| Corpuz | ISO 9001: 2015 quality management system requirements and audit findings classification using support vector machine and long short-term memory neural network: An optimization method | |
| CN113379211B (en) | Logistics information platform default risk management and control system and method based on blockchain | |
| CN118313727A (en) | Enterprise comprehensive assessment method and system based on large language model technology | |
| CN117076664A (en) | Text-based public opinion identification method, system and electronic device | |
| CN116208361A (en) | Cloud network fusion network and content security method and system | |
| CN117875706A (en) | Digital management method for grading process based on AI | |
| CN117648214A (en) | Exception log processing method and device | |
| CN117951260A (en) | Decision simulation execution method and device | |
| CN116257803A (en) | User state detection method and device, electronic equipment and storage medium | |
| CN118134657A (en) | Method for automatically acquiring claim information based on waybill number without certification | |
| CN118211832A (en) | Financial tax data risk monitoring method, system, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |