CN117807406A - Enterprise account management method, system, equipment and storage medium of payment platform - Google Patents
Enterprise account management method, system, equipment and storage medium of payment platform Download PDFInfo
- Publication number
- CN117807406A CN117807406A CN202410236060.7A CN202410236060A CN117807406A CN 117807406 A CN117807406 A CN 117807406A CN 202410236060 A CN202410236060 A CN 202410236060A CN 117807406 A CN117807406 A CN 117807406A
- Authority
- CN
- China
- Prior art keywords
- account
- enterprise
- enterprise account
- account management
- abnormal
- 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.)
- Granted
Links
- 238000007726 management method Methods 0.000 title claims abstract description 166
- 230000002159 abnormal effect Effects 0.000 claims abstract description 169
- 238000012795 verification Methods 0.000 claims abstract description 107
- 238000009826 distribution Methods 0.000 claims abstract description 66
- 238000004458 analytical method Methods 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 50
- 238000012706 support-vector machine Methods 0.000 claims abstract description 23
- 239000011159 matrix material Substances 0.000 claims description 16
- 238000012544 monitoring process Methods 0.000 claims description 12
- 230000004927 fusion Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 230000005856 abnormality Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 9
- 108091026890 Coding region Proteins 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 25
- 230000006399 behavior Effects 0.000 description 18
- 230000009471 action Effects 0.000 description 7
- 238000012502 risk assessment Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 3
- 230000010485 coping Effects 0.000 description 3
- 238000007418 data mining Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002349 favourable effect Effects 0.000 description 3
- 230000010354 integration Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 206010000117 Abnormal behaviour Diseases 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012954 risk control Methods 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000010223 real-time analysis Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/08—Payment architectures
- G06Q20/10—Payment architectures specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
- G06Q20/102—Bill distribution or payments
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Accounting & Taxation (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Probability & Statistics with Applications (AREA)
- Strategic Management (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application relates to the technical field of account management, and discloses an enterprise account management method, system, equipment and storage medium of a payment platform. The method comprises the following steps: generating a first account management logic structure and a second account management logic structure; creating a two-layer Bayesian network of each enterprise account verification requirement information and setting a constant probability discrimination threshold set; acquiring a historical enterprise account association data set of a target enterprise account through a payment platform and generating a plurality of enterprise account target item data to be verified; calculating abnormal probability distribution information and carrying out root cause analysis to obtain abnormal state data of a plurality of accounts; carrying out account abnormal state classification through a support vector machine model to obtain an account abnormal state classification result; and generating enterprise account management decision information according to the account abnormal state classification result, so that the accuracy of enterprise account management decision is improved.
Description
Technical Field
The present disclosure relates to the field of account management technologies, and in particular, to a method, a system, an apparatus, and a storage medium for managing an enterprise account of a payment platform.
Background
In the digital economic age today, paymate plays a key role in facilitating business economic activities. Enterprise account management, one of the core functions of paymate, is increasingly highlighting its importance to business operations and financial security. However, as enterprise size and complexity increases, traditional account management methods face increasing challenges. Traditional methods often appear too static and lack intelligent characteristics when dealing with large account data and multi-level account relationships, thereby limiting accurate prediction and management of potential risks.
In the current business environment, paymate is faced with diversified account verification requirements and complex account associations. The traditional account management method is difficult to meet the requirements of quick response and accurate judgment on the abnormal state of the account, and the efficiency and the accuracy of account management are insufficient. In addition, with the continuous expansion of paymate services, the past account management methods have not been able to accommodate the ever-increasing account size and data complexity, which further increases the difficulty of account management.
Disclosure of Invention
The application provides an enterprise account management method, system, equipment and storage medium of a payment platform, so that the accuracy of enterprise account management decisions is improved.
The first aspect of the present application provides a method for managing an enterprise account of a payment platform, where the method for managing an enterprise account of a payment platform includes:
acquiring N first-level enterprise account management logics, generating a first account management logic structure according to the N first-level enterprise account management logics, acquiring S second-level enterprise account management logics, and generating a second account management logic structure according to the S second-level enterprise account management logics, wherein N and S are positive integers, and N is more than S;
acquiring a plurality of enterprise account verification requirement information, and creating a two-layer Bayesian network of each enterprise account verification requirement information according to the first account management logic structure and the second account management logic structure;
acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network, and setting an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the parent node account management variables and the child node conditional probability distribution;
acquiring a historical enterprise account association data set of a target enterprise account through a preset payment platform, and generating a plurality of enterprise account target item data to be verified of the historical enterprise account association data set according to the abnormal probability discrimination threshold set;
Calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network, and respectively carrying out root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information to obtain abnormal state data of a plurality of accounts;
inputting the account abnormal state data into a preset support vector machine model to classify the account abnormal state, and obtaining an account abnormal state classification result;
and generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
A second aspect of the present application provides a business account management system for a paymate, the business account management system for a paymate comprising:
the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring N first-level enterprise account management logics, generating a first account management logic structure according to the N first-level enterprise account management logics, acquiring S second-level enterprise account management logics, and generating a second account management logic structure according to the S second-level enterprise account management logics, wherein N and S are positive integers, and N is more than S;
The creation module is used for acquiring a plurality of enterprise account verification requirement information and creating a two-layer Bayesian network of each enterprise account verification requirement information according to the first account management logic structure and the second account management logic structure;
the setting module is used for acquiring father node account management variables and child node conditional probability distribution of the two-layer Bayesian network, and setting an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the father node account management variables and the child node conditional probability distribution;
the processing module is used for acquiring a historical enterprise account associated data set of a target enterprise account through a preset payment platform and generating a plurality of enterprise account target item data to be verified of the historical enterprise account associated data set according to the abnormal probability judging threshold set;
the analysis module is used for calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network, and respectively carrying out root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information to obtain abnormal state data of a plurality of accounts;
The classification module is used for inputting the abnormal state data of the plurality of accounts into a preset support vector machine model to classify the abnormal states of the accounts, and obtaining an abnormal state classification result of the accounts;
and the generation module is used for generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
A third aspect of the present application provides a computer device comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the computer device to perform the enterprise account management method of the paymate described above.
A fourth aspect of the present application provides a computer readable storage medium having instructions stored therein that, when executed on a computer, cause the computer to perform the enterprise account management method of the paymate described above.
According to the technical scheme, the two-layer Bayesian network and the support vector machine model are used, so that an intelligent account management decision can be realized. The system can dynamically adjust the decision strategy according to the specific enterprise account verification requirement information, and the flexibility and accuracy of decision are improved. The system can carry out self-adaptive adjustment according to real-time data and changed situations by acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network and setting an abnormal probability judgment threshold set according to the information. This allows the sensitivity of the system to anomaly probabilities to be optimized over time and traffic environment variations. By fusing the enterprise account management logic structures of the first level and the second level into the two-level Bayesian network, the effective integration of multi-level logic is realized. This helps to more fully understand the enterprise account verification needs and improves the global perceptibility of the system to account status. By performing root cause analysis and abnormal state data monitoring according to the abnormal probability distribution information, the system can more accurately identify and understand the cause of the abnormal state of the account. This helps to quickly respond to potential problems and take appropriate action to improve the efficiency and accuracy of account management. By inputting the abnormal state data of a plurality of accounts into the support vector machine model for classification, efficient abnormal state classification can be realized. Support vector machines perform well in processing complex data sets, which facilitates accurate classification and identification of account states. By constructing an initial account decision strategy and a dynamic adjustment decision strategy, the system can flexibly adjust the decision according to actual conditions. The method is favorable for timely coping with the changed business environment and account state, improves the adaptability of the decision, and further improves the accuracy of enterprise account management decision.
Drawings
FIG. 1 is a schematic diagram of one embodiment of a method for enterprise account management for a paymate in an embodiment of the present application;
FIG. 2 is a schematic diagram of one embodiment of an enterprise account management system for a paymate in an embodiment of the present application.
Detailed Description
The embodiment of the application provides an enterprise account management method, system, equipment and storage medium of a payment platform, so that the accuracy of enterprise account management decisions is improved.
The terms "first," "second," "third," "fourth" and the like in the description and in the claims of this application and in the above-described figures, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For ease of understanding, a specific flow of an embodiment of the present application is described below, referring to fig. 1, and one embodiment of an enterprise account management method for a paymate in an embodiment of the present application includes:
step 101, acquiring N first-level enterprise account management logics, generating a first account management logic structure according to the N first-level enterprise account management logics, acquiring S second-level enterprise account management logics, and generating a second account management logic structure according to the S second-level enterprise account management logics, wherein N and S are positive integers, and N is more than S;
it may be understood that the executing body of the present application may be an enterprise account management system of a payment platform, and may also be a terminal or a server, which is not limited herein. The embodiment of the present application will be described by taking a server as an execution body.
Specifically, N first-level enterprise account management logics are obtained through a built-in rule base or an external professional database, and the logics comprise account verification rules, transaction limits, security monitoring standards and the like. These logics are parsed and classified by algorithms to understand their respective functions and roles. And then, organizing the N logics according to a certain structural relation, and constructing a logic structure capable of comprehensively reflecting the first-level account management logic through a tree structure, a graph network or other structural forms suitable for expressing the logic relation, wherein the structure is used as a basis for subsequent processing and decision. S second-level enterprise account management logic is obtained, which is generally more complex and advanced, and relates to a risk assessment model, an anomaly detection algorithm, a market dynamic adaptation strategy and the like. After these second level logics are obtained, they are subjected to deep analysis to understand their inherent logic relationships and mechanisms of action, and then based on these understandings, a second account management logic structure is constructed, which is more focused on high-level decision support and risk control. In this process, N and S are positive integers and N is greater than S, ensuring that the first level of logical structure has sufficient base and breadth, while the second level of logical structure provides deeper, specialized support.
102, acquiring a plurality of enterprise account verification requirement information, and creating a two-layer Bayesian network of each enterprise account verification requirement information according to a first account management logic structure and a second account management logic structure;
in particular, verification requirement information of a plurality of enterprise accounts is obtained from a database of a payment platform or a real-time transaction monitoring system, wherein the information comprises key data points such as account activities, transaction modes, fund flows and the like. And carrying out demand characteristic analysis on the verification demand information of each enterprise account, and reading and understanding the specific content of each verification demand, including the type, frequency, amount, participant and the like of the transaction, so as to obtain the verification demand characteristics of the verification demand information of each enterprise account. Further, verification type analysis is performed, and the specific type of account verification is judged according to the demand characteristics, such as whether the account verification is high-volume transaction verification, new account verification or frequent transaction verification, etc., so that each verification demand is treated more accurately. Matching the first account management logic structure according to the enterprise account verification type of each enterprise account verification requirement information, searching logic substructures corresponding to the verification requirements in the first logic structure, wherein the substructures represent basic logic and rules required by processing specific verification types, and simultaneously performing similar matching work on the second account management logic structure to find higher-level and more complex logic substructures. After the two logic sub-structures are obtained, a first layer Bayesian network and a second layer Bayesian network of each enterprise account verification requirement information are respectively created according to the two logic sub-structures, and the two networks respectively map basic logic and high-level logic for processing the verification requirements, so that the system can better understand and infer account behaviors and potential risks. And (3) carrying out network fusion on the first-layer Bayesian network and the second-layer Bayesian network, and organically combining the basic logic network and the advanced logic network to form a unified and collaborative decision support system. Through network fusion, two layers of Bayesian networks of each enterprise account verification requirement information can mutually supplement information and mutually verify assumptions, so that more comprehensive and more accurate verification support is provided.
Step 103, acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network, and setting an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the parent node account management variables and the child node conditional probability distribution;
it should be noted that, node extraction is performed on the two-layer bayesian network, and all parent nodes and their corresponding child nodes in the network are identified and extracted through analysis of the network structure, where the nodes represent various factors and their interrelationships in the account management process, such as transaction amount, frequency, account age, and their probability relations. The extracted parent nodes are subjected to variable analysis, account management variables represented by each parent node are understood and defined, the variables are key factors in the decision making process, and the states and changes of the variables can significantly influence the risk assessment and management decision of the account. And carrying out probability distribution calculation on the extracted child nodes through a parameter estimation algorithm, and estimating the conditional probability distribution of the child node states under the given parent node states according to historical data and a statistical model, wherein the distribution is the basis for understanding and predicting account behaviors. Based on father node account management variables and child node conditional probability distribution, a plurality of node abnormality discrimination rules are matched for each enterprise account verification requirement information, and the rules are formulated based on various resources such as historical cases, industry standards, expert knowledge and the like, and aim at risk and abnormal behavior to be identified. Each rule corresponds to a certain probability threshold, and when the probability distribution of the account behavior meets the rule condition, the system marks the account behavior as a potential risk. And setting a plurality of abnormal probability discrimination thresholds for each enterprise account verification requirement information according to the node abnormal discrimination rules, and performing set conversion on the thresholds to form a final abnormal probability discrimination threshold set. The set integrates risk assessment of multiple angles and levels, and can provide a comprehensive and detailed risk judgment basis for the system.
104, acquiring a historical enterprise account association data set of a target enterprise account through a preset payment platform, and generating a plurality of enterprise account target item data to be verified of the historical enterprise account association data set according to the abnormal probability discrimination threshold set;
specifically, a historical enterprise account association data set of the target enterprise account is obtained through API call, database query or other data interface modes, and the data sets comprise transaction records, account activities, associated account information and the like. The historical enterprise account association data set is classified into a plurality of candidate account target item data according to the attribute, the behavior mode, the association relation and the like of the data through data mining and machine learning technologies, and each data represents an account behavior or event needing further verification. Thus, account activities with risks or anomalies are more effectively identified and focused, and pertinence and efficiency of verification work are improved. And according to the anomaly probability discrimination threshold set, performing anomaly discrimination on the target item data of the plurality of candidate accounts, wherein the discrimination process utilizes probability models such as Bayesian networks, decision trees, logistic regression and the like to calculate the anomaly probability of each candidate item, and compares the anomaly probability with the criteria in the threshold set to obtain the anomaly discrimination result of the target item data of each candidate account. This result identifies whether each candidate meets the probability criteria for abnormal behavior, helping the system identify those account behaviors that require further attention and verification. And further screening the candidate account target item data according to the abnormal judgment results, and screening out items judged to be abnormal from all the candidate items to form a plurality of enterprise account target item data to be verified. These data to be validated are the result of a preliminary risk assessment, which will be the basis for subsequent deep validation and analysis.
Step 105, calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to a two-layer Bayesian network, and respectively carrying out root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information to obtain abnormal state data of the accounts;
specifically, the anomaly probability distribution calculation is carried out on the target item data of each enterprise account to be verified through a two-layer Bayesian network. The calculation not only needs to consider the direct influence factor of each data point, but also comprehensively considers the indirect influence and interaction in the whole network structure, so that the comprehensiveness and accuracy of the calculation result are ensured. And carrying out rule matching on a preset root cause analysis rule base through the abnormal probability distribution information. The rule base is constructed by the historical case analysis and the data mining technology, and comprises a large number of root causes and corresponding probability models thereof. The system determines a corresponding target root cause rule tree from the rule base according to the specific condition of the abnormal probability distribution information, and the rule tree guides the subsequent root cause analysis process to provide a structured analysis path and logic. Traversing the determined target root cause rule tree to obtain root cause analysis rules corresponding to the target item data of the enterprise account to be verified. These rules are specific implementations of the target root cause rule tree that detail how potential root causes are inferred from specific anomaly probability distribution information. By means of these rules, complex, abstract probability distribution information is transformed into concrete, operational root cause hypotheses. And monitoring abnormal state data of the enterprise account target item data to be verified according to the root cause analysis rule. Root cause analysis rules are applied to identify the potential root cause of each of the to-be-verified items and further monitor the state changes of those items to verify the accuracy of the root cause assumptions and determine the final abnormal state.
Step 106, inputting the abnormal state data of the plurality of accounts into a preset support vector machine model to classify the abnormal states of the accounts, and obtaining an abnormal state classification result of the accounts;
specifically, the abnormal state data of a plurality of accounts are coded and mapped, each account contains various types of information, such as transaction abnormality, login abnormality, frequent account information change and the like, and the multidimensional and diversified abnormal state information is converted into a standardized state coding sequence through the coding and mapping. These sequences express the abnormal state information of each account in the form of values or symbols so that it can be recognized and processed by a computer program. And carrying out matrix fusion on the state coding sequence of the abnormal state data of each account. Matrix fusion is a method of combining multiple sequences or data points into a single structure, typically used to enhance the expressive power and integrity of the data. Through this process, a plurality of state code sequences are fused into a comprehensive account abnormal state matrix. Inputting the account abnormal state matrix into a preset support vector machine model to classify the account abnormal state, and obtaining an account abnormal state classification result. Support vector machines are a machine learning algorithm that distinguishes between different classes of data points by constructing one or more hyperplanes. The SVM model classifies the input account abnormal state matrix by utilizing the learned discrimination boundary, and classifies the abnormal state of each account into a predefined category, such as normal, suspicious, high risk and the like.
And 107, generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
Specifically, according to the account abnormal state classification result, an initial account decision strategy of target item data of each enterprise account to be verified is respectively constructed, and the initial account decision strategy defines preliminary management measures which should be adopted for enterprise accounts with different abnormal state classifications, such as adding monitoring, limiting certain transactions, conducting further examination and the like. And dynamically adjusting the primary account decision strategy of the target item data of each enterprise account to be verified. This adjustment process takes into account a variety of factors such as historical behavior of the account, associated networks, changes in the market environment, etc., and dynamically updates and optimizes the decision-making policies of each account via algorithms and models to ensure that these policies can adapt to changes in the external environment and dynamic characteristics of account behavior. And carrying out enterprise account management decision on the target enterprise account according to the target account decision strategy. And comprehensively considering the target strategy of each account, the specific condition of the account and the overall risk management framework, and preparing specific management measures and action plans. These measures and plans will provide personalized, accurate management decision support for the specific risk status and management needs of each account. And carrying out information combination on the established enterprise account management decisions to obtain corresponding enterprise account management decision information. This merging process not only simplifies the decision information, but also ensures consistency and operability of the decisions so that they can be efficiently implemented and tracked by the paymate.
In the embodiment of the application, the two-layer Bayesian network and the support vector machine model are used, so that an intelligent account management decision can be realized. The system can dynamically adjust the decision strategy according to the specific enterprise account verification requirement information, and the flexibility and accuracy of decision are improved. The system can carry out self-adaptive adjustment according to real-time data and changed situations by acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network and setting an abnormal probability judgment threshold set according to the information. This allows the sensitivity of the system to anomaly probabilities to be optimized over time and traffic environment variations. By fusing the enterprise account management logic structures of the first level and the second level into the two-level Bayesian network, the effective integration of multi-level logic is realized. This helps to more fully understand the enterprise account verification needs and improves the global perceptibility of the system to account status. By performing root cause analysis and abnormal state data monitoring according to the abnormal probability distribution information, the system can more accurately identify and understand the cause of the abnormal state of the account. This helps to quickly respond to potential problems and take appropriate action to improve the efficiency and accuracy of account management. By inputting the abnormal state data of a plurality of accounts into the support vector machine model for classification, efficient abnormal state classification can be realized. Support vector machines perform well in processing complex data sets, which facilitates accurate classification and identification of account states. By constructing an initial account decision strategy and a dynamic adjustment decision strategy, the system can flexibly adjust the decision according to actual conditions. The method is favorable for timely coping with the changed business environment and account state, improves the adaptability of the decision, and further improves the accuracy of enterprise account management decision.
In a specific embodiment, the process of executing step 102 may specifically include the following steps:
(1) Acquiring a plurality of enterprise account verification requirement information, and respectively carrying out requirement characteristic analysis on each enterprise account verification requirement information to obtain verification requirement characteristics of each enterprise account verification requirement information;
(2) Performing verification type analysis on the verification requirement characteristics to obtain enterprise account verification types of the verification requirement information of each enterprise account;
(3) According to the enterprise account verification type of each enterprise account verification requirement information, performing account management logic structure matching on the first account management logic structure to obtain a first logic substructure of each enterprise account verification requirement information, and performing account management logic structure matching on the second account management logic structure to obtain a second logic substructure of each enterprise account verification requirement information;
(4) Creating a first layer Bayesian network of each enterprise account verification requirement information according to the first logic sub-structure, and creating a second layer Bayesian network of each enterprise account verification requirement information according to the second logic sub-structure;
(5) And carrying out network fusion on the first-layer Bayesian network and the second-layer Bayesian network to obtain two-layer Bayesian networks of the verification requirement information of each enterprise account.
Specifically, various enterprise account verification requirement information is obtained from the payment platform, wherein the information comprises login data of an account, transaction records, fund flows and the like. And carrying out demand characteristic analysis on the verification demand information of each enterprise account. For example, for an account with frequent large transactions, the verification requirement characteristics include transaction frequency, size of the amount, payee account, and the like. And (3) performing verification type analysis on the verification demand characteristics, matching the verification demand characteristics with a predefined verification type, such as classifying frequent small-amount transfer behaviors as 'suspicious scattered transactions', classifying large-amount burst transactions as 'large-amount abnormal transactions', and the like. Thereby converting the disordered verification requirement information into a clear and standardized verification type, and facilitating subsequent processing and decision-making. And carrying out account management logic structure matching on the first account management logic structure according to the verification type of the verification requirement information of each enterprise account. Logical substructures corresponding to authentication types are found in a first account management logical structure, which generally represents the basic rules and procedures required to handle various conventional authentication requirements. The second account management logic structures are matched to find higher level, more complex logic sub-structures that typically involve more complex decision support and risk control logic. Next, a first layer bayesian network of each enterprise account verification requirement information is created from the first logical sub-structure, and a second layer bayesian network is created from the second logical sub-structure. The two layers of networks respectively map basic logic and high-level logic for processing verification requirements, wherein the first layer of network comprises basic account behavior rules, and the second layer of network comprises complex risk assessment and decision models. And performing network fusion on the first-layer Bayesian network and the second-layer Bayesian network. By establishing connection and interaction between two layers of networks, the two networks can mutually supplement information and mutually verify assumptions, so that a unified and collaborative decision support system is formed.
In a specific embodiment, the process of executing step 103 may specifically include the following steps:
(1) Node extraction is carried out on the two-layer Bayesian network to obtain father nodes and corresponding child nodes;
(2) Performing variable analysis on the parent node to obtain a parent node account management variable, and performing probability distribution calculation on the child node through a parameter estimation algorithm to obtain child node conditional probability distribution;
(3) Based on the parent node account management variables and child node conditional probability distribution, respectively matching a plurality of node abnormality discrimination rules of each enterprise account verification requirement information;
(4) And respectively setting a plurality of abnormal probability discrimination thresholds corresponding to the verification requirement information of each enterprise account according to a plurality of node abnormal discrimination rules, and carrying out set conversion on the plurality of abnormal probability discrimination thresholds to obtain an abnormal probability discrimination threshold set.
Specifically, node extraction is performed on the two-layer Bayesian network, the structure of the Bayesian network is traversed, and all father nodes and the child nodes corresponding to the father nodes are identified. The parent node typically represents the primary factors that affect account behavior, such as transaction amount, login frequency, account qualification, etc., while the child nodes represent the results or states that occur in a particular state of the parent node, such as transaction being denied, account being frozen, etc. And carrying out variable analysis on the father nodes, and determining the specific meaning and effect of the account management variable represented by each father node. For example, a parent node represents the transaction amount, and information such as the normal range, trend, etc. of this amount needs to be understood by variable analysis. And carrying out probability distribution calculation on each child node by using a parameter estimation algorithm, and calculating the occurrence probability of each state of the child node under the specific state of each parent node. This calculation relies on historical data and statistical models, which are intended to provide a quantitative basis for subsequent risk decisions and decisions. Based on the parent node account management variables and child node conditional probability distribution, a plurality of node abnormality discrimination rules of each enterprise account verification requirement information are respectively matched. Each rule describes a corresponding account anomaly behavior and risk level in the case of a particular parent node state and child node probability distribution. And determining a series of abnormal judgment rules for each piece of verification requirement information through rule matching. And respectively setting a plurality of abnormal probability discrimination thresholds for each enterprise account verification requirement information according to the node abnormal discrimination rules. These thresholds represent probability criteria for determining account behavior as normal or abnormal, and are only marked as abnormal if the child node probability distribution caused by the account behavior meets these thresholds. And performing set conversion on the plurality of abnormal probability discrimination thresholds, and combining the abnormal probability discrimination thresholds into a comprehensive abnormal probability discrimination threshold set. This set will be used for final account behavior determination and classification, providing a unified, coordinated risk assessment criteria.
In a specific embodiment, the process of executing step 104 may specifically include the following steps:
(1) Acquiring a historical enterprise account association data set of a target enterprise account through a preset payment platform;
(2) Performing association data classification on the historical enterprise account association data set to obtain target item data of a plurality of candidate accounts;
(3) Performing anomaly discrimination on the target item data of the plurality of candidate accounts according to the anomaly probability discrimination threshold set to obtain an anomaly discrimination result of the target item data of each candidate account;
(4) And carrying out data screening on the plurality of candidate account target item data according to the abnormal discrimination result to generate a plurality of enterprise account target item data to be verified.
Specifically, through interaction with an interface of a preset payment platform, historical data of a target enterprise account is obtained. Such data typically includes transaction records, account balance changes, account login information, and other account association information, etc., resulting in a historical enterprise account association data set. And carrying out association data classification on the historical enterprise account association data set. Through data mining and machine learning techniques, data is classified according to a plurality of dimensions such as transaction type, frequency, amount, associated account type and the like, and candidate account target item data with specific characteristics and modes are identified from a cluttered data set. These candidate data are the focus of subsequent analysis, and represent a potential risk or act requiring further verification. And carrying out anomaly discrimination on the candidate account target item data according to the set anomaly probability discrimination threshold set. In this process, various risk indexes and probabilities for each candidate item are calculated, and these calculation results are compared with an abnormality probability discrimination threshold. If the risk indicator of a candidate item exceeds a corresponding threshold, the item is determined to be abnormal, and otherwise, is determined to be normal. This discrimination process is dynamic and complex, requiring comprehensive consideration of multiple factors and indicators to ensure accuracy and reliability of the discrimination results. And further screening the target item data of the plurality of candidate accounts according to the abnormal discrimination result. And selecting all the matters judged to be abnormal to form final to-be-verified enterprise account target matter data. These data represent account behaviors that the system currently deems most problematic or risky, and they will be prioritized for subsequent in-depth investigation and analysis.
In a specific embodiment, the process of executing step 105 may specifically include the following steps:
(1) Calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network;
(2) Rule matching is carried out on a preset root cause analysis rule base through the abnormal probability distribution information, and a target root cause rule tree corresponding to the abnormal probability distribution information is determined;
(3) Traversing the target root cause rule tree to obtain root cause analysis rules corresponding to the target item data of the enterprise account to be verified;
(4) And monitoring abnormal state data of the enterprise account target item data to be verified according to the root cause analysis rule to obtain a plurality of account abnormal state data.
Specifically, according to the two-layer Bayesian network, the abnormal probability distribution of the target item data of each enterprise account to be verified is calculated. The probability of each target item under different conditions is calculated through the two layers of networks, so that a comprehensive abnormal probability distribution diagram is formed. And carrying out rule matching on a preset root cause analysis rule base through the abnormal probability distribution information. The rule base contains a plurality of root cause analysis rules, each of which describes a relationship between a particular abnormal probability distribution pattern and a root cause. And determining which rules are consistent with the current abnormal probability distribution information through rule matching, thereby selecting one or more most target root cause rule trees. These rule trees will provide structured paths and logic for subsequent root cause analysis. Traversing the target root cause rule tree, exploring different branches and nodes of each rule tree, and finding out all root cause analysis rules applicable to the current target item data. And monitoring abnormal state data of the enterprise account target item data to be verified according to the root cause analysis rule. The state changes of each target item are closely monitored while root cause analysis rules are applied to evaluate and interpret the changes. If the actual state change of a target item matches an anomaly pattern described by a root cause analysis rule, the system records the state of the item as anomaly and generates corresponding account anomaly state data. These data will detail the current status of each target item, the root cause, and other relevant information.
In a specific embodiment, the process of executing step 106 may specifically include the following steps:
(1) Performing coding mapping on the abnormal state data of the accounts to obtain a state coding sequence of the abnormal state data of each account;
(2) Carrying out matrix fusion on the state coding sequence of each account abnormal state data to generate an account abnormal state matrix;
(3) Inputting the account abnormal state matrix into a preset support vector machine model to classify the account abnormal state, and obtaining an account abnormal state classification result.
Specifically, the plurality of account abnormal state data are coded and mapped. These abnormal states include various signals such as transaction abnormality, login abnormality, frequent account data change, etc. The purpose of the code mapping is to convert these diversified and complex abnormal state data into a standardized, easily handled format. In this process, a unique code is defined for each abnormal state, and then a corresponding code is assigned to each account based on its actual abnormal state, forming a code sequence describing the abnormal state of the account. And (3) carrying out matrix fusion on the state coding sequence of the abnormal state data of each account, combining a plurality of data sources or data points into a unified structure, and enhancing the expressive force and the integrity of the data. And combining a plurality of state coding sequences from the same account into a unified account abnormal state matrix through matrix fusion. And inputting the account abnormal state matrix into a preset support vector machine model to classify the account abnormal state. Support vector machines are a supervised learning algorithm that distinguishes between different classes of data points by constructing one or more hyperplanes in the data space. At this stage, the SVM model will classify each of the account anomaly state matrices according to the learned classification boundaries, determining to which predefined category the account anomaly state represented by each matrix belongs, e.g., normal, suspicious, high risk, etc. This classification process depends on the discriminatory power of the SVM model and the quality of the previous data preprocessing.
In a specific embodiment, the process of performing step 107 may specifically include the following steps:
(1) Respectively constructing an initial account decision strategy of target item data of each enterprise account to be verified according to the account abnormal state classification result;
(2) Dynamically adjusting the primary account decision strategy of the target item data of each enterprise account to be verified to obtain the target account decision strategy of the target item data of each enterprise account to be verified;
(3) And carrying out enterprise account management decision and decision information combination on the target enterprise account according to the target account decision strategy to obtain corresponding enterprise account management decision information.
Specifically, an initial account decision strategy is respectively constructed according to the account abnormal state classification result. These classification results divide account status into multiple levels of normal, suspicious, high risk, etc., each of which requires a corresponding initial decision strategy. The primary account decision policy is formulated based on the current anomaly status of the account and the risk management policy of the paymate, and includes a series of predefined actions such as increasing the monitoring strength of the account, limiting the transaction rights of the account, or requiring the account holder to provide more verification information, etc. And dynamically adjusting the primary account decision strategy of the target item data of each enterprise account to be verified. The adjustment process is based on real-time analysis and evaluation of various information such as account behavior, history records, market environment and the like. For example, if an account marked as suspicious has been recorded with a false positive in the past, or if the market environment changes, the system will reduce its risk level accordingly, reducing the restrictions on that account. Also, if the transaction pattern of an account suddenly changes abnormally, the system can raise the risk level and increase monitoring and limiting measures. This process requires the system to constantly collect and analyze data to ensure that the decision strategy always matches the current situation. Each enterprise account target item data to be verified can obtain a dynamically adjusted target account decision strategy, and the strategy can reflect the actual risk condition and management requirement of the account more accurately. And carrying out enterprise account management decision on the target enterprise account according to the target account decision strategy, and carrying out decision information merging. The target decision strategy formulated for each account is translated into specific management actions and measures. For example, for an account that is marked as suspicious, the system may decide to increase the strength of the audit of its transaction and require the account holder to provide additional authentication information. These specific management decisions will be tailored to account specific situations and target decision policies. The information combination of the decisions ensures that all decision information is consistent and coordinated without conflicting or repeated measures. The merging process not only simplifies the decision information, but also improves the decision efficiency and execution force.
The method for managing an enterprise account of a payment platform in the embodiment of the present application is described above, and the enterprise account management system of a payment platform in the embodiment of the present application is described below, referring to fig. 2, an embodiment of the enterprise account management system of a payment platform in the embodiment of the present application includes:
an obtaining module 201, configured to obtain N first-level enterprise account management logics and generate a first account management logic structure according to the N first-level enterprise account management logics, obtain S second-level enterprise account management logics, and generate a second account management logic structure according to the S second-level enterprise account management logics, where N and S are positive integers, and N > S;
a creating module 202, configured to obtain a plurality of enterprise account verification requirement information, and create a two-layer bayesian network of each enterprise account verification requirement information according to the first account management logic structure and the second account management logic structure;
the setting module 203 is configured to obtain a parent node account management variable and a child node conditional probability distribution of the two-layer bayesian network, and set an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the parent node account management variable and the child node conditional probability distribution;
The processing module 204 is configured to obtain a historical enterprise account related data set of a target enterprise account through a preset payment platform, and generate a plurality of enterprise account target item data to be verified of the historical enterprise account related data set according to the anomaly probability discrimination threshold set;
the analysis module 205 is configured to calculate, according to the two-layer bayesian network, abnormal probability distribution information of target item data of each enterprise account to be verified, and perform root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information, so as to obtain abnormal state data of a plurality of accounts;
the classification module 206 is configured to input the plurality of account abnormal state data into a preset support vector machine model to classify the account abnormal state, so as to obtain an account abnormal state classification result;
and the generating module 207 is configured to generate enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
Through the cooperation of the components, an intelligent account management decision can be realized by using a two-layer Bayesian network and a support vector machine model. The system can dynamically adjust the decision strategy according to the specific enterprise account verification requirement information, and the flexibility and accuracy of decision are improved. The system can carry out self-adaptive adjustment according to real-time data and changed situations by acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network and setting an abnormal probability judgment threshold set according to the information. This allows the sensitivity of the system to anomaly probabilities to be optimized over time and traffic environment variations. By fusing the enterprise account management logic structures of the first level and the second level into the two-level Bayesian network, the effective integration of multi-level logic is realized. This helps to more fully understand the enterprise account verification needs and improves the global perceptibility of the system to account status. By performing root cause analysis and abnormal state data monitoring according to the abnormal probability distribution information, the system can more accurately identify and understand the cause of the abnormal state of the account. This helps to quickly respond to potential problems and take appropriate action to improve the efficiency and accuracy of account management. By inputting the abnormal state data of a plurality of accounts into the support vector machine model for classification, efficient abnormal state classification can be realized. Support vector machines perform well in processing complex data sets, which facilitates accurate classification and identification of account states. By constructing an initial account decision strategy and a dynamic adjustment decision strategy, the system can flexibly adjust the decision according to actual conditions. The method is favorable for timely coping with the changed business environment and account state, improves the adaptability of the decision, and further improves the accuracy of enterprise account management decision.
The present application also provides a computer device including a memory and a processor, where the memory stores computer readable instructions that, when executed by the processor, cause the processor to perform the steps of the method for enterprise account management for paymate in the above embodiments.
The present application also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, and may also be a volatile computer readable storage medium, where instructions are stored in the computer readable storage medium, when the instructions are executed on a computer, cause the computer to perform the steps of the enterprise account management method of the paymate.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (10)
1. The enterprise account management method of the payment platform is characterized by comprising the following steps of:
acquiring N first-level enterprise account management logics, generating a first account management logic structure according to the N first-level enterprise account management logics, acquiring S second-level enterprise account management logics, and generating a second account management logic structure according to the S second-level enterprise account management logics, wherein N and S are positive integers, and N is more than S;
acquiring a plurality of enterprise account verification requirement information, and creating a two-layer Bayesian network of each enterprise account verification requirement information according to the first account management logic structure and the second account management logic structure;
Acquiring parent node account management variables and child node conditional probability distribution of the two-layer Bayesian network, and setting an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the parent node account management variables and the child node conditional probability distribution;
acquiring a historical enterprise account association data set of a target enterprise account through a preset payment platform, and generating a plurality of enterprise account target item data to be verified of the historical enterprise account association data set according to the abnormal probability discrimination threshold set;
calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network, and respectively carrying out root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information to obtain abnormal state data of a plurality of accounts;
inputting the account abnormal state data into a preset support vector machine model to classify the account abnormal state, and obtaining an account abnormal state classification result;
and generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
2. The method of claim 1, wherein the obtaining a plurality of enterprise account verification requirement information and creating a two-tier bayesian network for each enterprise account verification requirement information based on the first account management logic structure and the second account management logic structure comprises:
acquiring a plurality of enterprise account verification requirement information, and respectively carrying out requirement characteristic analysis on each enterprise account verification requirement information to obtain verification requirement characteristics of each enterprise account verification requirement information;
performing verification type analysis on the verification requirement characteristics to obtain enterprise account verification types of the verification requirement information of each enterprise account;
according to the enterprise account verification type of each enterprise account verification requirement information, performing account management logic structure matching on the first account management logic structure to obtain a first logic substructure of each enterprise account verification requirement information, and performing account management logic structure matching on the second account management logic structure to obtain a second logic substructure of each enterprise account verification requirement information;
creating a first-layer Bayesian network of each enterprise account verification requirement information according to the first logic sub-structure, and creating a second-layer Bayesian network of each enterprise account verification requirement information according to the second logic sub-structure;
And carrying out network fusion on the first-layer Bayesian network and the second-layer Bayesian network to obtain two-layer Bayesian networks of the verification requirement information of each enterprise account.
3. The method for managing enterprise accounts of a paymate of claim 1, wherein the obtaining parent node account management variables and child node conditional probability distributions of the two-layer bayesian network and setting the abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the parent node account management variables and the child node conditional probability distributions comprises:
node extraction is carried out on the two-layer Bayesian network to obtain father nodes and corresponding child nodes;
performing variable analysis on the parent node to obtain a parent node account management variable, and performing probability distribution calculation on the child node through a parameter estimation algorithm to obtain child node conditional probability distribution;
based on the parent node account management variable and the child node conditional probability distribution, respectively matching a plurality of node abnormality discrimination rules of each enterprise account verification requirement information;
and respectively setting a plurality of abnormal probability discrimination thresholds corresponding to the verification requirement information of each enterprise account according to the plurality of node abnormal discrimination rules, and carrying out set conversion on the plurality of abnormal probability discrimination thresholds to obtain an abnormal probability discrimination threshold set.
4. The method for managing enterprise accounts of paymate as claimed in claim 1, wherein the obtaining, by the preset paymate, a historical enterprise account related data set of the target enterprise account, and generating a plurality of to-be-verified enterprise account target item data of the historical enterprise account related data set according to the abnormal probability discrimination threshold set, comprises:
acquiring a historical enterprise account association data set of a target enterprise account through a preset payment platform;
performing association data classification on the historical enterprise account association data set to obtain target item data of a plurality of candidate accounts;
performing anomaly discrimination on the plurality of candidate account target item data according to the anomaly probability discrimination threshold set to obtain an anomaly discrimination result of each candidate account target item data;
and carrying out data screening on the plurality of candidate account target item data according to the abnormal discrimination result to generate a plurality of enterprise account target item data to be verified.
5. The method for managing enterprise accounts of paymate of claim 1, wherein the calculating anomaly probability distribution information of each target item of enterprise account to be verified according to the two-layer bayesian network, and performing root cause analysis on the target item data of enterprise account to be verified based on the anomaly probability distribution information, respectively, to obtain a plurality of account anomaly state data, comprises:
Calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network;
rule matching is carried out on a preset root cause analysis rule base through the abnormal probability distribution information, and a target root cause rule tree corresponding to the abnormal probability distribution information is determined;
traversing the target root cause rule tree to obtain root cause analysis rules corresponding to the enterprise account target item data to be verified;
and monitoring abnormal state data of the enterprise account target item data to be verified according to the root cause analysis rule to obtain a plurality of account abnormal state data.
6. The method for managing enterprise accounts of the payment platform according to claim 1, wherein inputting the plurality of account abnormal state data into a preset support vector machine model for account abnormal state classification, obtaining an account abnormal state classification result, comprises:
performing coding mapping on the plurality of account abnormal state data to obtain a state coding sequence of each account abnormal state data;
carrying out matrix fusion on the state coding sequence of each account abnormal state data to generate an account abnormal state matrix;
And inputting the account abnormal state matrix into a preset support vector machine model to classify the account abnormal state, and obtaining an account abnormal state classification result.
7. The method for managing enterprise accounts of a payment platform according to claim 1, wherein the generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result comprises:
respectively constructing an initial account decision strategy of target item data of each enterprise account to be verified according to the account abnormal state classification result;
dynamically adjusting the primary account decision strategy of the target item data of each enterprise account to be verified to obtain the target account decision strategy of the target item data of each enterprise account to be verified;
and carrying out enterprise account management decision and decision information combination on the target enterprise account according to the target account decision strategy to obtain corresponding enterprise account management decision information.
8. A business account management system for a paymate, the business account management system for a paymate comprising:
the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring N first-level enterprise account management logics, generating a first account management logic structure according to the N first-level enterprise account management logics, acquiring S second-level enterprise account management logics, and generating a second account management logic structure according to the S second-level enterprise account management logics, wherein N and S are positive integers, and N is more than S;
The creation module is used for acquiring a plurality of enterprise account verification requirement information and creating a two-layer Bayesian network of each enterprise account verification requirement information according to the first account management logic structure and the second account management logic structure;
the setting module is used for acquiring father node account management variables and child node conditional probability distribution of the two-layer Bayesian network, and setting an abnormal probability discrimination threshold set of each enterprise account verification requirement information according to the father node account management variables and the child node conditional probability distribution;
the processing module is used for acquiring a historical enterprise account associated data set of a target enterprise account through a preset payment platform and generating a plurality of enterprise account target item data to be verified of the historical enterprise account associated data set according to the abnormal probability judging threshold set;
the analysis module is used for calculating abnormal probability distribution information of target item data of each enterprise account to be verified according to the two-layer Bayesian network, and respectively carrying out root cause analysis on the target item data of the enterprise accounts to be verified based on the abnormal probability distribution information to obtain abnormal state data of a plurality of accounts;
The classification module is used for inputting the abnormal state data of the plurality of accounts into a preset support vector machine model to classify the abnormal states of the accounts, and obtaining an abnormal state classification result of the accounts;
and the generation module is used for generating enterprise account management decision information corresponding to the target enterprise account according to the account abnormal state classification result.
9. A computer device, the computer device comprising: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invoking the instructions in the memory to cause the computer device to perform the enterprise account management method of the paymate of any of claims 1-7.
10. A computer readable storage medium having instructions stored thereon, which when executed by a processor, implement the method of enterprise account management for a paymate as claimed in any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410236060.7A CN117807406B (en) | 2024-03-01 | 2024-03-01 | Enterprise account management method, system, equipment and storage medium of payment platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410236060.7A CN117807406B (en) | 2024-03-01 | 2024-03-01 | Enterprise account management method, system, equipment and storage medium of payment platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117807406A true CN117807406A (en) | 2024-04-02 |
| CN117807406B CN117807406B (en) | 2024-04-30 |
Family
ID=90433885
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410236060.7A Active CN117807406B (en) | 2024-03-01 | 2024-03-01 | Enterprise account management method, system, equipment and storage medium of payment platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117807406B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118537018A (en) * | 2024-07-25 | 2024-08-23 | 深圳市拜特科技股份有限公司 | Enterprise payment risk management system based on deep learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112801800A (en) * | 2021-04-14 | 2021-05-14 | 深圳格隆汇信息科技有限公司 | Behavior fund analysis system, behavior fund analysis method, computer equipment and storage medium |
| CN113935819A (en) * | 2021-10-25 | 2022-01-14 | 国网江苏省电力有限公司营销服务中心 | Method for extracting checking abnormal features |
| CN115063143A (en) * | 2022-06-14 | 2022-09-16 | 中国工商银行股份有限公司 | Account data processing method and device, computer equipment and storage medium |
| CN116188020A (en) * | 2023-02-20 | 2023-05-30 | 中国工商银行股份有限公司 | Abnormal account identification method, device and storage medium |
-
2024
- 2024-03-01 CN CN202410236060.7A patent/CN117807406B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112801800A (en) * | 2021-04-14 | 2021-05-14 | 深圳格隆汇信息科技有限公司 | Behavior fund analysis system, behavior fund analysis method, computer equipment and storage medium |
| CN113935819A (en) * | 2021-10-25 | 2022-01-14 | 国网江苏省电力有限公司营销服务中心 | Method for extracting checking abnormal features |
| CN115063143A (en) * | 2022-06-14 | 2022-09-16 | 中国工商银行股份有限公司 | Account data processing method and device, computer equipment and storage medium |
| CN116188020A (en) * | 2023-02-20 | 2023-05-30 | 中国工商银行股份有限公司 | Abnormal account identification method, device and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118537018A (en) * | 2024-07-25 | 2024-08-23 | 深圳市拜特科技股份有限公司 | Enterprise payment risk management system based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117807406B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9661010B2 (en) | Security log mining devices, methods, and systems | |
| CN111461216B (en) | Case risk identification method based on machine learning | |
| Zanke et al. | Predictive Modelling for Insurance Pricing: A Comparative Analysis of Machine Learning Approaches | |
| US20200286095A1 (en) | Method, apparatus and computer programs for generating a machine-learning system and for classifying a transaction as either fraudulent or genuine | |
| CN117235811A (en) | Internet finance-based payment big data analysis method and big data analysis system | |
| KR102499181B1 (en) | Loan regular auditing system using artificia intellicence | |
| Baumann | Improving a rule-based fraud detection system with classification based on association rule mining | |
| CN117807406B (en) | Enterprise account management method, system, equipment and storage medium of payment platform | |
| CN118586706A (en) | Enterprise risk assessment system, method, device, storage medium and program product | |
| CN111949852A (en) | Macroscopic economy analysis method and system based on internet big data | |
| Zhang et al. | A hierarchical clustering strategy of processing class imbalance and its application in fraud detection | |
| KR102409019B1 (en) | System and method for risk assessment of financial transactions and computer program for the same | |
| Ball et al. | Anomaly detection using autoencoders with network analysis features | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| US20200285895A1 (en) | Method, apparatus and computer program for selecting a subset of training transactions from a plurality of training transactions | |
| Gangopadhyay et al. | LAD in finance: accounting analytics and fraud detection | |
| CN118569655B (en) | Staged data life cycle safety assessment method and system | |
| CN113742472B (en) | Data mining method and device based on customer service marketing scene | |
| Thirumagal et al. | AI-Enhanced IoT Data Analytics for Risk Management in Banking Operations | |
| Bäßler et al. | Predictive Process Monitoring for Internal Audit: Forecasting Payment Punctuality from the Perspective of the Three Lines Model | |
| Dileep et al. | Financial Fraud Detection Using Deep Learning Techniques | |
| Yao et al. | Intelligent Network Awareness | |
| Enzinger et al. | Use Case—Fraud Detection Using Machine Learning Techniques | |
| Niu et al. | Leveraging Rough Set Theory to Enhance the Performance of Financial Statement Fraud Detection Model | |
| Patel | AI-DRIVEN FRAUD DETECTION IN CRYPTOCURRENCY TRANSACTIONS |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |