A ship leasing customer stratification compliance verification method and system based on big data

Abstract

The application discloses a ship leasing customer hierarchical compliance verification method and system based on big data, and the method comprises the following steps: customer information initialization and correlation graph construction, incremental customer flow real-time preliminary screening, sanction compliance information association acquisition, project review, contract making, payment approval node embedded compliance verification, and ship international sanction risk whole-process monitoring. At the key nodes of business affairs such as customer creation / change, project review, contract making and payment approval, the preset compliance data interface is automatically called to capture customer basic information and sanction compliance information, and embedded compliance review is automatically performed. Through ship information docking, pre-warning, in-process reporting and post-history backtracking of ships related to sanction regions are realized. Through automatic data capture and multi-node embedded review, the application improves customer compliance verification efficiency, constructs a ship risk "early warning-reporting-backtracking" closed loop, and strengthens the real-time performance and whole-process traceability of ship leasing business risk control.

Description

Technical Field

[0001] This invention relates to the field of risk management technology in the ship leasing industry, and in particular to a big data-based method and system for tiered compliance verification of ship leasing customers. Background Technology

[0002] Due to its globalized nature, the ship leasing industry faces increasingly stringent international sanctions and regulatory compliance requirements. Customer qualification verification and vessel navigation risk monitoring are directly related to a company's compliance level and risk control capabilities. Traditional methods relying on manual screening of customer backgrounds and manual verification of vessel status are insufficient to handle massive amounts of dynamic data and real-time monitoring needs, necessitating the automation and intelligentization of compliance management through digital technologies.

[0003] The current ship leasing compliance management solutions commonly used in the industry have the following significant shortcomings:

[0004] (1) Customer compliance verification is inefficient, such as information collection relying on manual labor, disconnected review nodes, and delayed response;

[0005] (2) There is a lack of dynamic monitoring of ships, especially in terms of risk warning in areas involving international sanctions, there is a lack of effective pre-event, during-event and post-event automated control.

[0006] (3) Insufficient system collaboration. Data sources such as the company's internal compliance system and third-party shipping service platforms are isolated and have not built a unified big data interface scheduling system, forming "information silos" that further restrict review efficiency and risk response speed. Summary of the Invention

[0007] This invention addresses the technical problems of existing ship leasing compliance management solutions, such as low efficiency in customer compliance verification and lack of fully automated control over international sanctions risks. It proposes a big data-based tiered compliance verification method for ship leasing customers. Through a pre-defined compliance data interface, it automatically retrieves relevant customer and vessel compliance data, performing automated customer compliance reviews at key business nodes throughout the entire process, from customer registration and project review to contract drafting and payment approval. It also achieves integrated automated control of international sanctions risks, including pre-emptive warnings, in-process reporting, and post-event review, significantly improving customer compliance verification efficiency and enhancing the closed-loop risk management capabilities across the entire ship leasing chain. This invention also relates to a big data-based tiered compliance verification system for ship leasing customers.

[0008] The technical solution of the present invention is as follows:

[0009] A big data-based method for tiered compliance verification of ship leasing customers, characterized by the following steps:

[0010] Customer information initialization and association graph construction steps: In response to customer creation or information change business transactions, the system automatically obtains relevant compliant business data streams of customers through a preset compliant data interface, and performs field mapping, de-identification cleaning, and unified format normalization on the business data streams to obtain standardized customer basic data. The standardized customer basic data is collected and stored in the customer relationship management system. Entity feature vectors of customers and their related transaction entities are extracted from the standardized customer basic data. A customer association graph is constructed based on the entity feature vectors, and a snapshot of the graph is generated. The standardized customer basic data and the generated customer association graph snapshot are written into the customer relationship time series database.

[0011] The incremental customer streaming real-time initial screening process is as follows: At the customer creation or information change business transaction submission node, the preset compliance data interface is automatically called to obtain real-time compliance return data. Then, the streaming real-time verification engine is called to incrementally traverse the customers and their related transaction entities covered by the customer incremental business data corresponding to this submission based on the constructed customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock is automatically applied to the current business transaction to block the subsequent business process flow. If the risk score is less than the preset risk threshold, the business transaction is allowed and the streaming verification record is stored to complete the compliance initial screening of the customers and related entities corresponding to this incremental transaction.

[0012] Sanctions compliance information acquisition steps: At the trigger node of the ship leasing project review, the preset compliance data interface is automatically called to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph are updated. Relying on the customer relationship graph, the customer and all related transaction entities under the current project are recursively traversed, and the sanctions compliance related information of each entity is automatically extracted. For new customers cooperating in this ship leasing project, a new customer sanctions compliance due diligence assessment and audit form is generated in a structured manner, and the new customer sanctions compliance due diligence assessment and audit form is bound to the current project review as a mandatory attachment.

[0013] Embedded compliance verification steps at project review nodes: At the initial and final review submission nodes of the ship leasing project, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and related transaction entities under the current project based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. The entity compliance risk level is divided according to the preset risk threshold, and the complete compliance verification results of this project are output simultaneously to complete the embedded compliance verification of all entities of the project.

[0014] Embedded compliance verification steps at the contract creation node: At the transaction trigger node for creating and saving the ship leasing contract, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current contract based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. The entity compliance risk level is divided according to the preset risk threshold, and the complete compliance verification result of this contract is output simultaneously to complete the embedded compliance verification of the entire contract entity.

[0015] Embedded compliance verification steps at the payment approval node: At the payment approval transaction submission node, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current payment approval item based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. The entity compliance risk level is divided according to the preset risk threshold, and the complete compliance verification result of this payment is output synchronously to complete the embedded compliance verification of the entire payment entity.

[0016] The process of monitoring the risks of international sanctions against ships involves: obtaining real-time and historical information on ship routes, ship status, and destination ports through big data interfaces, and implementing full-process monitoring of ships involved in international sanctions areas, including pre-warning, in-process reporting, and post-event review.

[0017] Preferably, after the embedded compliance verification step at the payment approval node, a periodic compliance verification step for existing customers / projects is also included:

[0018] Configure a periodic compliance verification mechanism for existing clients or projects; the periodic compliance verification mechanism is a manual trigger mode, which is divided into manually performing single or all compliance verifications for clients and manually performing single or all compliance verifications for projects.

[0019] It also centrally stores and visualizes all records of compliance verification operations conducted by calling preset compliance data interfaces and the corresponding verification results. The records of compliance verification operations include manual compliance verification records and automatic compliance verification records generated by automatically calling preset compliance data interfaces at key nodes of each business transaction. The key nodes of the business transaction include the customer creation or information change business transaction submission node, the ship leasing project review transaction trigger node, the project preliminary review and final review transaction submission node, the ship leasing contract creation and saving transaction trigger node, and the payment approval transaction submission node.

[0020] Preferably, in the process of monitoring the risks of international sanctions against ships, the advance warning includes: configuring a daily scheduled task to call the API of a third-party shipping service platform to synchronously obtain the destination port information and international sanctioned area data of the planned voyage segment to be executed by the ship, and verifying in advance whether the destination port of the planned voyage segment to be executed by the ship belongs to the international sanctioned area; if there are ships planning to enter the international sanctioned area, an early warning reminder message is pushed.

[0021] The in-process reporting includes: configuring a daily scheduled task to call the API of a third-party shipping service platform to synchronously obtain the real-time dynamic location information of the vessel and data on international sanctioned areas, and determining whether the vessel's current location is within an international sanctioned area; if a vessel is already within an international sanctioned area, a reporting reminder message will be pushed.

[0022] The post-event retrospective includes: centrally storing and visually displaying all historical data on ships entering and leaving international sanctioned areas, supporting historical data retrospective queries, and regularly pushing relevant data reports on international sanctioned areas.

[0023] Preferably, in the customer information initialization and association graph construction step, the preset compliance data interface serves as a unified compliance standardized big data intermediate interface, with a built-in unified data transmission protocol and metadata conversion mapping table, and provides a standardized risk feature vector output port to the outside world; the compliance return data obtained by subsequent steps calling the preset compliance data interface includes identification data used to distinguish whether a customer or vessel is compliant;

[0024] The standardized customer base data includes customer name, project source, and customer characteristics;

[0025] The customer relationship management system is used to perform customer relationship management operations, which include maintaining customer needs, pricing plans, and visit record information.

[0026] Preferably, in the incremental customer streaming real-time initial screening step, the incremental traversal adopts a local directional traversal method, loading only the local graph subgraphs in the customer relationship graph corresponding to the current incremental customer business for calculation; the entity feature vectors corresponding to each entity are weighted and fused with the real-time compliance return data and then input into the blacklist fuzzy matching model to complete the risk scoring; when the risk score is greater than or equal to the preset risk threshold and a database transaction lock is applied, only the data row corresponding to the current business transaction is locked, and the business transaction is marked as pending manual review and a risk warning is pushed; the stored streaming verification record includes entity identifier, matching similarity, risk score value, preset risk threshold, process handling result, and verification timestamp.

[0027] Preferably, in the step of obtaining the associated sanctions compliance information, only the entity feature vector corresponding to the entity that has obtained the latest sanctions compliance data is updated in the customer relationship graph, while the entity feature vectors of other entities remain unchanged. The recursive traversal is carried out along the multi-level relationship links of customer equity, actual controller, guarantor, and partner, traversing all levels of related transaction entities of the customer under the current project. The compliance fields are automatically extracted from the updated entity feature vectors and the latest sanctions compliance data and structuredly filled into the new customer sanctions compliance due diligence assessment and review form. If the new customer sanctions compliance due diligence assessment and review form is not generated and bound as a mandatory attachment to the current project review transaction, the project review transaction is blocked from flowing to the next approval node. The updated entity feature vectors are synchronously written back to the customer relationship time-series database to retain the corresponding version snapshot.

[0028] Preferably, in the embedded compliance verification steps of the project review node, contract preparation node, and payment approval node, when performing a full traversal, all related entity links of the customer relationship graph are fully loaded, all related entities under the current business item are extracted at once, and the entity feature vectors of all entities are input into the blacklist fuzzy matching model in parallel batches to simultaneously complete batch risk scoring calculations; multi-level entity compliance risk levels are divided according to the risk scoring intervals and corresponding handling rules are matched; the output complete compliance verification result includes verification type, business document number, entity identifier, entity risk score, entity compliance risk level, blacklist matching basis, and verification timestamp; the complete compliance verification result is bound to the corresponding ship leasing project review, ship leasing contract, or payment approval application and stored in the customer relationship time-series database for compliance evidence preservation; if any entity is determined to be of high risk level during the traversal, the current ship leasing project review, contract preparation, or payment approval transaction is marked as suspended, and a high-risk warning is pushed.

[0029] Preferably, in the process of monitoring the risk of international sanctions against ships, the big data interface is connected to the enterprise's internal compliance risk system and the API of a third-party shipping service platform. By connecting to the enterprise's internal compliance risk system, the historical compliance risk control files of customers and their related transaction entities can be retrieved. By configuring daily scheduled tasks to call the API of the third-party shipping service platform, the destination port information or real-time dynamic location information of the ship's planned voyage segment and the corresponding international sanctions area data can be obtained synchronously.

[0030] The aforementioned early warning and notification messages, reporting reminder messages, and data reports related to international sanctioned zones are all pushed via email.

[0031] In the aforementioned advance warning, the warning reminder information pushed by the email includes ship information and the international sanctions region matching determination result. The ship information includes ship name, country of origin of voyage, port of origin of voyage, country of destination of voyage, start time of voyage, and end time of voyage.

[0032] In the aforementioned in-process reporting, the reporting reminder information pushed via email includes vessel information and the international sanctions region matching determination results. The vessel information includes the vessel name, the region where the vessel is located, location information, and current latitude and longitude.

[0033] In the post-event retrospective analysis, the data reports related to international sanctioned zones, which are regularly sent via email, adopt a fixed format and summarize all historical records of ships entering and leaving international sanctioned zones.

[0034] A tiered compliance verification system for ship leasing customers based on big data is characterized by comprising a customer information initialization and association graph construction module, an incremental customer streaming real-time screening module, a sanctions compliance information association acquisition module, an embedded compliance verification module for project review nodes, an embedded compliance verification module for contract preparation nodes, an embedded compliance verification module for payment approval nodes, and a full-process monitoring module for international sanctions risks of ships.

[0035] The customer information initialization and association graph construction module responds to customer creation or information change business transactions, automatically obtains relevant compliant business data streams of customers through a preset compliant data interface, and performs field mapping, de-identification cleaning, and unified format normalization processing on the business data streams to obtain standardized customer basic data. The standardized customer basic data is collected and stored in the customer relationship management system. Entity feature vectors of customers and their related transaction entities are extracted from the standardized customer basic data. A customer association graph is constructed based on the entity feature vectors, and a snapshot of the graph is generated. The standardized customer basic data and the generated customer association graph snapshot are written into the customer relationship time series database.

[0036] The incremental customer streaming real-time initial screening module, at the customer creation or information change business transaction submission node, first automatically calls the preset compliance data interface to obtain real-time compliance return data, and then calls the streaming real-time verification engine. Based on the constructed customer relationship graph, it incrementally traverses the customers and their related transaction entities covered by the incremental customer business data corresponding to this submission. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock is automatically applied to the current business transaction to block the subsequent business process flow. If the risk score is less than the preset risk threshold, the business transaction is allowed and the streaming verification record is stored, completing the compliance initial screening of the customer and related entities corresponding to this increment.

[0037] The sanctions compliance information acquisition module, at the trigger node of the ship leasing project review, first automatically calls a preset compliance data interface to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, it updates the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph. Relying on the customer relationship graph, it recursively traverses the customers and all related transaction entities under the current project, automatically extracts the sanctions compliance related information corresponding to each entity, and for new customers cooperating in this ship leasing project, it structurally fills in and generates a new customer sanctions compliance due diligence assessment and audit form, and binds the new customer sanctions compliance due diligence assessment and audit form to the current project review as a mandatory attachment.

[0038] The embedded compliance verification module at the project review node automatically calls a preset compliance data interface to obtain real-time compliance return data at the initial review and final review submission nodes of the ship leasing project. It reuses the streaming real-time verification engine to fully traverse the customers and related transaction entities under the current project based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on the preset risk threshold, it classifies the entity compliance risk level and outputs the complete compliance verification results of this project simultaneously, thus completing the embedded compliance verification of all entities in the project.

[0039] The embedded compliance verification module at the contract creation node automatically calls a preset compliance data interface to obtain real-time compliance return data at the ship leasing contract creation and saving transaction trigger node. It reuses the streaming real-time verification engine to fully traverse the customers and their related transaction entities under the current contract based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on the preset risk threshold, it classifies the entity compliance risk level and outputs the complete compliance verification result of this contract simultaneously, thus completing the full entity embedded compliance verification of the contract.

[0040] The embedded compliance verification module at the payment approval node automatically calls a preset compliance data interface to obtain real-time compliance return data at the payment approval transaction submission node. It reuses the streaming real-time verification engine to fully traverse the customers and their related transaction entities under the current payment approval item based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on the preset risk threshold, it classifies the entity compliance risk level and outputs the complete compliance verification result of this payment simultaneously, thus completing the embedded compliance verification of the entire payment entity.

[0041] The aforementioned full-process monitoring module for international sanctions risks of ships obtains real-time and historical information on ship routes, ship status, and destination ports through big data interfaces, and performs full-process monitoring of ships involved in international sanctions areas, including pre-warning, in-process reporting, and post-event retrospective.

[0042] Preferably, between the embedded compliance verification module at the payment approval node and the full-process monitoring module for international ship sanctions risks, a periodic compliance verification module for existing clients / projects is also included:

[0043] Configure a periodic compliance verification mechanism for existing clients or projects; the periodic compliance verification mechanism is a manual trigger mode, which is divided into manually performing single or all compliance verifications for clients and manually performing single or all compliance verifications for projects.

[0044] The system centrally stores and visualizes all records of compliance verification operations performed by calling preset compliance data interfaces, along with the corresponding verification results. These records include manual compliance verification records and automatic compliance verification records generated by automatically calling preset compliance data interfaces at key business transaction nodes. The key business transaction nodes include the customer creation or information change business transaction submission node, the ship leasing project review transaction trigger node, the project preliminary and final review transaction submission node, the ship leasing contract creation and saving transaction trigger node, and the payment approval transaction submission node.

[0045] The technical effects of this invention are as follows:

[0046] This invention relates to a big data-based method for tiered compliance verification of ship leasing customers. Its customer information initialization and association graph construction steps automatically capture customer compliance business data streams through a preset compliance data interface, replacing traditional manual offline collection and manual entry of customer information. This avoids data omissions, entry errors, and information update delays caused by manual operations. Through field mapping, desensitization and cleaning, and unified format normalization, it completes the standardization processing of multi-source customer data, solving the pain points of inconsistent data formats and inability to perform unified calculations and analyses across different channels. Based on the standardized data, it extracts entity feature vectors to build a customer relationship graph and retains a snapshot of the graph. It persistently stores the relationships of all customers and upstream and downstream related transaction entities in a customer relationship time-series database, providing a unified, traceable, and complete underlying relationship data foundation for compliance verification of all subsequent business nodes, enabling visualized risk control analysis of the customer relationship network. Its incremental customer streaming real-time initial screening step embeds real-time verification logic at the business submission nodes for customer additions and information changes. Relying on a preset compliance data interface, it pulls compliance return data in real time and uses an incremental traversal mechanism to process only the customers and related transaction entities involved in the current change. Compared to full traversal, this significantly reduces computational power consumption and risk. The system offers faster response times, reusing a blacklist fuzzy matching model combined with real-time compliance data to automatically quantify risk scores. This eliminates the shortcomings of inconsistent standards in manual subjective screening and judgment. It automatically applies database transaction locks to high-risk transactions, blocking business flow and intercepting compliance risks at the very beginning of customer onboarding. Simultaneously, it fully retains every streaming verification record, completing the initial compliance risk screening. This achieves both proactive risk interception and meets the regulatory requirement for full traceability of compliance operations. Furthermore, the step of obtaining compliance information related to sanctions is triggered during project review, updating the entity relationships within the customer association graph in real time through a pre-set compliance data interface. Feature vectors ensure the real-time effectiveness of graph-based risk control data, avoiding compliance oversights caused by reviews based on expired customer information. Relying on the customer relationship graph, it recursively traverses all related transaction entities under the project, capturing the sanctions compliance information of indirect cooperating entities without omission. This addresses the industry shortcoming of manual review, which only verifies direct customers and omits related parties. It automatically generates standardized new customer sanctions compliance due diligence assessment and review forms in a structured manner and forcibly binds them to project review tasks, saving the workload of manually organizing and archiving due diligence materials. At the same time, it retains due diligence documents in a strong association with the project, meeting the international sanctions regulatory due diligence data retention standards.Its project review node embeds compliance verification steps into the initial and final review submission nodes, enabling business review and compliance risk control to proceed in parallel without the need for a separate offline risk control process. This simplifies the business approval chain, calls a preset compliance data interface to obtain real-time compliance data, and reuses the same streaming real-time verification engine to complete a full traversal and batch risk scoring of all customers and related entities under the project. This unifies the risk control model and judgment standards, avoids differences in review standards among different reviewers, and automatically classifies the compliance risk level of entities based on preset thresholds and outputs complete project verification results, thereby displaying the overall compliance status of the project and achieving comprehensive risk control for all cooperating entities during the ship leasing project initiation stage. Its contract preparation node embeds compliance verification steps within the ship... The ship leasing contract storage process automatically triggers a full-entity compliance review, forming a dual risk control defense line between project review and contract signing. This compensates for the inability to identify changes in customer qualifications or new sanctions risks after project review. It reuses pre-set compliance data interfaces, a streaming real-time verification engine, and a blacklist fuzzy matching model throughout the process, maintaining consistent risk control rules across all business nodes. Complete contract compliance verification results are bound to the lease contract for storage, and risk control records at the signing stage can be directly traced and queried, mitigating operational losses caused by compliance risks arising after signing. Furthermore, the embedded compliance verification step at the payment approval stage forcibly triggers compliance verification before the payment application is submitted, constructing a three-tiered risk control closed loop of "project initiation-signing-payment," preventing payments to entities with potential sanctions risks. For settlement of payments with partners involved in insurance, the latest compliance data is synchronously obtained through a pre-set compliance data interface. This allows for targeted traversal of all clients and related entities involved in the payment application and batch completion of risk scoring, eliminating the need for approval personnel to conduct additional offline verification of client qualifications. Payment compliance verification records and payment approval documents are simultaneously archived and retained, meeting the regulatory requirements for risk control traceability of transaction entities under anti-sanctions and cross-border fund supervision. The entire process of monitoring ship international sanctions risks independently utilizes a big data interface to collect ship-related data. This interface differs from the pre-set compliance data interfaces that only connect to internal systems in the aforementioned steps. This big data interface can simultaneously connect to the company's internal compliance risk system and third-party shipping service platform APIs, enabling daily... Scheduled tasks automatically synchronize data on vessel planned voyage segments, real-time location, destination ports, and international sanctioned areas, replacing manual periodic queries of vessel navigation status. Furthermore, it can access internally stored historical compliance and risk control files for clients, enabling cross-verification of vessel navigation risks and the qualifications of leasing clients. This breaks down the traditionally fragmented internal risk control systems and external shipping platform data sources, eliminating information silos. Relying on a fully automated management mechanism encompassing pre-emptive warnings, in-process reporting, and post-event retrospective analysis, it comprehensively covers international sanction risk management throughout the entire vessel voyage cycle. This addresses the shortcomings of traditional solutions that only offer post-event verification and lack fully automated risk control measures throughout the entire vessel process, significantly improving client compliance verification efficiency and enhancing the closed-loop risk management capabilities across the entire vessel leasing chain.

[0047] Overall, this invention presents a big data-based tiered compliance verification method for ship leasing customers. By deeply integrating two types of differentiated compliance data interaction interfaces, it embeds automated compliance review nodes into all critical business processes, including customer relationship creation and modification, project review, contract drafting, and payment approval. It also includes a periodic compliance scanning mechanism for existing customers and projects. Furthermore, it innovatively establishes a fully automated monitoring chain for the entire process of international sanctions risks, encompassing pre-emptive warnings, in-process reporting, and post-event retrospective analysis. Based on this integrated risk control architecture, this invention significantly improves the efficiency, accuracy, and coverage of compliance review work in the ship leasing industry. It reduces risk omissions and regulatory response delays caused by manual screening, ensuring stable compliance throughout the entire ship leasing business process and significantly strengthening enterprises' proactive prevention capabilities against navigation risks related to transit through or docking in internationally sanctioned areas. The system also features a centralized review record storage mechanism, with all risk control operations, verification data, and ship monitoring logs uniformly archived, greatly simplifying compliance file tracing and regulatory audit verification.

[0048] (1) Layered embedded automated risk control across the entire business chain, completely solving the pain point of inefficiency of manual verification: This invention covers the entire business process of customer filing, incremental changes, project review, contract signing, payment approval, and ship navigation. It embeds automated compliance verification logic in each key business node, replaces manual screening of the entire process with machine streaming calculation, optimizes computing power consumption by distinguishing between incremental traversal and full traversal operation modes, reuses a unified risk control model across all nodes to quantify risks, eliminates industry pain points such as lag, omissions, and inconsistent evaluation standards in manual review, and significantly improves the efficiency of compliance review for the entire ship leasing business.

[0049] (2) Based on the customer association graph, realize risk identification of all related entities without blind spots: The unified customer association graph is built with standardized time-series customer data as the underlying support. It can recursively traverse all upstream and downstream related transaction entities of customers, break through the limitation of traditional manual review that only checks direct cooperative customers, and comprehensively identify the indirect sanctions risks hidden in the complete customer association network, and improve the risk prevention and control system from the subject dimension.

[0050] (3) Two types of compliance interfaces work together to connect multiple channels of compliance data links: This invention configures two types of differentiated data interaction interfaces: one is a pre-set compliance data interface for customer business, which serves as a standardized intermediate interface to uniformly process internal customer compliance data flow; the other is a dedicated big data interface for ship monitoring, which links internal and external dual data sources to synchronize ship dynamics and customer risk control files; the two types of interfaces perform their respective functions and work together to break down the data barriers between the internal risk control system and the third-party shipping platform, solving the information silo problem that is common in the industry.

[0051] (4) Constructing a two-dimensional integrated compliance closed-loop management system for both customer entities and vessel navigation: On the one hand, a three-tiered progressive risk control interception mechanism is set up for the entire life cycle of customers to intercept customer compliance risks in advance at all stages of access, project establishment, contract signing and payment; on the other hand, an automated monitoring process for sanctions risks throughout the entire life cycle of vessel navigation is provided to deeply link customer entity qualification risk control with vessel route dynamic risk control, forming a complete compliance risk closed loop covering leasing partners and vessel navigation behavior, while taking into account pre-event risk interception, in-event dynamic monitoring and post-event full-link retrospective of archives, fully adapting to the stringent international sanctions regulatory requirements of the globalized ship leasing industry.

[0052] (5) Full-process data standardization and operation traceability meet regulatory archiving requirements: The whole solution completes customer data cleaning and normalization through preset compliance data interface. All customer basic data, correlation map snapshots, node verification records and ship navigation monitoring logs are uniformly and persistently stored in time series database. All compliance verification and risk monitoring actions are automatically stored in standardized archives, fully supporting regulatory audit and compliance traceability verification work.

[0053] Furthermore, this invention adds a periodic compliance verification mechanism for existing customers and projects after the full-process verification of payment approval. It provides two trigger modes: single manual verification and full batch verification, distinguishing between customers and projects. This achieves two-way coverage of incremental risk control for new business and periodic review of existing business, making up for the loopholes that only automatic verification at the business creation node cannot be timely identified when changes in the qualifications of existing customers occur. At the same time, it uniformly collects both automatic node verification and manual verification operation records, completing the centralized storage and visualization of all types of compliance records. It fully covers the risk control behavior of all business nodes, including customer creation and changes, project review, contract preparation, and payment approval, making it easy for managers to intuitively view the full-dimensional compliance operation logs and providing a unified and visualized data carrier for regulatory audits and historical risk reviews.

[0054] Furthermore, this invention refines the monitoring of international sanctions risks for ships into three segments: each segment is configured with an independent daily scheduled task to retrieve data from a third-party shipping platform, enabling automatic data synchronization before and during the event without manual querying; before the event, sanctions risks are predicted based on the destination port of the planned voyage, and early warnings are pushed out in advance to achieve proactive risk prevention; during the event, the ship's current geographical location is verified in real time, and reporting information is pushed out immediately to ships that have entered the sanctioned area, achieving dynamic control of the navigation process; after the event, complete historical navigation data of ships across sanctioned areas is centrally retained, supporting retrospective query at any time and regular automatic push of standardized statistical reports, constructing a full-cycle automated ship risk control link covering navigation plans, real-time navigation, and historical navigation records, completely solving the shortcomings of traditional ship risk control that can only perform manual verification after the event and lacks the ability to provide early warnings and dynamic tracking.

[0055] Furthermore, the pre-defined compliance data interface is limited to a unified compliance standardized big data middleware interface. Through a built-in fixed transmission protocol and metadata conversion mapping table, it completes unified conversion of multi-source customer data, outputting standardized risk feature vectors externally. This unifies the data input and output standards across all business nodes, eliminating issues of inconsistent data formats and incompatible model calculations across different business stages. The interface output carries customer and vessel compliance identification data, which can be directly used for risk assessment, eliminating the need for additional data parsing steps. Simultaneously, it defines standardized customer basic data fields and defines the business operations carried by the customer relationship management system, achieving structured and unified storage of basic customer information and full lifecycle customer business management. This provides standardized and complete underlying customer data support for end-to-end compliance verification.

[0056] Furthermore, the traversal calculation logic for incremental customer verification is optimized. Local targeted traversal is adopted, loading only the local graph subgraph corresponding to the change for calculation. Compared with loading the entire graph, this significantly reduces memory usage and computing power consumption, improving the verification response speed in incremental customer change scenarios. Entity feature vectors are weighted and fused with real-time compliance data before being input into the matching model, improving the accuracy of comprehensive risk scoring. In high-risk scenarios, only the data row corresponding to the current business is locked instead of a global lock, without interfering with the normal flow of other parallel businesses. At the same time, it is marked for manual review and a notification is pushed, balancing the stability of business system operation and the timeliness of risk handling. Multi-dimensional streaming verification detail fields are fully stored, and each incremental verification process can be accurately traced.

[0057] Furthermore, the steps for obtaining compliance information related to sanctions are limited to updating the feature vectors of entities only in this update, while retaining the original values ​​of unchanged entity data. This avoids meaningless recalculation of the entire graph and reduces the computational cost of data updates during the project review process. The system recursively traverses multiple layers of the chain, including equity, actual controllers, guarantors, and partners, fully covering indirect related entities at multiple levels and preventing missed reviews of related clients. It automatically extracts compliance fields to generate standardized due diligence forms and sets up a mandatory attachment blocking mechanism; approval cannot proceed without completed due diligence archiving, meeting the regulatory requirement for mandatory retention of due diligence data. Updated entity feature vectors are synchronously written back to the time-series database to retain version snapshots, fully recording changes in risk control data of client-related entities at each stage, achieving full traceability of data versions throughout the project review process.

[0058] Furthermore, the system standardizes the full-scale calculation logic for the three major business nodes of projects, contracts, and payments, fully loads all related entity links, completes model scoring in parallel batches, and outputs all entity risk results at once, significantly improving the efficiency of batch risk control calculations. It divides risk levels into multiple levels based on scoring intervals and binds corresponding handling rules, standardizing risk control judgment criteria. It fully records multi-dimensional verification details and binds them to business documents, storing them in a time-series database for compliant evidence preservation. If any entity is deemed high-risk, the current business is suspended and an alert is pushed. High-risk partners are promptly intercepted at key funding nodes such as project initiation, signing, and payment, forming a multi-layered fund security risk control defense line to avoid compliance losses in cross-border leasing settlements.

[0059] Furthermore, the full-process monitoring steps for international sanctions risks on ships limit the use of a dedicated big data interface for ship monitoring, which can simultaneously link with the enterprise's internal compliance and risk control system and third-party shipping platform APIs. This allows for the retrieval of historical risk control files from clients to achieve cross-verification between ships and leasing clients, and can also synchronize planned voyages, real-time locations, and geographic data of sanctioned areas on a regular basis, breaking down information silos by connecting heterogeneous data sources internally and externally. All warnings, reports, and statements are automatically pushed via email, eliminating the need for manual offline processing and notification. Differentiated ship detail push fields are defined for pre-event and in-event scenarios, ensuring complete and clear risk information for quick identification by auditors. Post-event reports automatically summarize historical voyage records using a fixed standardized format, unifying the archiving standards for ship risk control files and meeting the regulatory requirements of international sanctions for ship navigation risk notifications and historical record retention.

[0060] This invention also relates to a big data-based tiered compliance verification system for ship leasing customers. Corresponding to the aforementioned big data-based tiered compliance verification method for ship leasing customers, this system can be understood as a system for implementing the big data-based tiered compliance verification method for ship leasing customers. It includes a customer information initialization and association graph construction module, an incremental customer streaming real-time initial screening module, a sanctions compliance information association acquisition module, a project review node embedded compliance verification module, a contract preparation node embedded compliance verification module, a payment approval node embedded compliance verification module, and a full-process monitoring module for international ship sanctions risks. Based on a modular and integrated architecture design, it highly integrates core functions such as customer data processing, incremental risk initial screening, sanctions information association verification, multi-business node embedded risk control, and full-process monitoring of international ship sanctions. Each module operates in an orderly and coordinated manner, achieving fully automated, standardized, and systematic operation of ship leasing customer compliance verification and ship navigation risk monitoring. This system abandons the traditional model of decentralized manual verification, fragmented risk control at business nodes, and lack of dynamic vessel monitoring. It utilizes dual-type differentiated data interfaces to achieve unified scheduling and standardized processing of internal and external compliance data. Based on customer relationship graphs, it achieves comprehensive risk screening of multi-level related entities. It can perform intelligent compliance verification throughout the customer's entire lifecycle and at each key business node of the leasing project, while simultaneously achieving closed-loop management of vessel sanctions risks through pre-emptive warnings, in-process monitoring, and post-event retrospective analysis. The system's overall architecture features low coupling and strong functionality, effectively breaking down industry data silos, significantly improving the overall efficiency of compliance review and the accuracy of risk identification. It comprehensively avoids compliance risks such as omissions, delayed responses, and inconsistent standards inherent in manual verification. It uniformly collects and retains risk control logs and verification data throughout the entire process, providing stable, reliable, and traceable intelligent system support for compliant operation, dynamic risk prevention and control, and regulatory audit traceability in vessel leasing business. Attached Figure Description

[0061] Figure 1 This is a flowchart of the big data-based tiered compliance verification method for ship leasing customers according to the present invention.

[0062] Figure 2 This is a schematic diagram of the business flow in different scenarios according to the present invention; wherein, Figure 2 a) Provides a diagram for customers to create / modify the main business flow. Figure 2 b is a flowchart for the timer-triggered periodic verification process for existing customers / projects. Figure 2 c represents the flowchart for ship data source-driven ship sanctions monitoring.

[0063] Figure 3 This is a screenshot of the interface of the new customer sanctions compliance due diligence assessment and audit form in the big data-based ship leasing customer tiered compliance verification system of this invention.

[0064] Figure 4This is a screenshot of the manual compliance verification interface for customers in the big data-based tiered compliance verification system for ship leasing customers of this invention.

[0065] Figure 5 This is a screenshot of the customer verification record interface in the big data-based ship leasing customer tiered compliance verification system of this invention. Detailed Implementation

[0066] The present invention will now be described in detail with reference to the accompanying drawings.

[0067] This invention provides a big data-based method for tiered compliance verification of ship leasing customers, such as... Figure 1 The overall flowchart shown lists all the execution steps (S1~S8) of the method of the present invention from top to bottom. Figure 1 This is only a list of steps; the arrows in the diagram are for illustrative purposes only and do not represent the actual execution sequence. The existing customer / project periodic compliance verification step (S7) is a timer-triggered parallel process, while the ship international sanctions risk full-process monitoring step (S8) is an independently driven branch based on the ship data source. Both can run synchronously with the main customer business line on the left, without needing to wait sequentially. See also... Figure 2 The schematic diagram of the business flow of this invention in different scenarios shown includes three sub-diagrams, among which, Figure 2 a creates / modifies the main business flow diagram for customers, fully demonstrating the layered embedded risk control logic of all business nodes from initial collection of customer information and real-time initial screening of incremental customers to project review, contract preparation and payment approval. If the verification result of any business node is non-compliant, the current business process will be terminated directly and a risk warning will be pushed. Figure 2 b is a flowchart for the timer-triggered periodic verification of existing customers / projects; Figure 2 c is the flowchart for the independent branch of ship sanctions monitoring driven by ship data sources.

[0068] This invention provides a big data-based method for tiered compliance verification of ship leasing customers, which specifically includes the following steps:

[0069] S1. Customer Information Initialization and Association Graph Construction Steps: In response to customer creation or information change business transactions, automatically acquire relevant compliant business data streams from the customer through a preset compliance data interface. Perform field mapping, de-identification cleaning, and unified format normalization on the business data streams to obtain standardized customer basic data. Collect and store the standardized customer basic data in the customer relationship management system. Extract entity feature vectors of customers and their related transaction entities from the standardized customer basic data. Construct a customer association graph based on the entity feature vectors and generate a snapshot corresponding to the graph. Write the standardized customer basic data and the generated customer association graph snapshot into the customer relationship time-series database. This step corresponds to... Figure 2Customer information initialization and collection are shown in figure a.

[0070] Furthermore, the preset compliance data interface serves as a unified compliance standardization big data intermediate interface, with a built-in unified data transmission protocol and metadata conversion mapping table, and provides a standardized risk feature vector output port to the outside world; the compliance return data obtained by calling the preset compliance data interface in subsequent steps includes identification data used to distinguish whether a customer or vessel is compliant.

[0071] The standardized customer base data includes customer name, project source, customer characteristics, etc.; the customer relationship management system is used to perform customer relationship management operations, including maintaining customer needs, pricing plans, visit record information, etc.

[0072] S2. Incremental Customer Streaming Real-Time Initial Screening Steps: At the customer creation or information change business transaction submission node, the preset compliance data interface is automatically called to obtain real-time compliance return data. Then, the streaming real-time verification engine is called to incrementally traverse the customers and their related transaction entities covered by the customer incremental business data corresponding to this submission based on the constructed customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock is automatically applied to the current business transaction to block the subsequent business process flow. If the risk score is less than the preset risk threshold, the business transaction is allowed and the streaming verification record for this transaction is stored, completing the initial compliance screening of the customer and related entities corresponding to this increment.

[0073] This step corresponds to Figure 2 As shown in Figure a, the initial compliance screening for customer relationship establishment / modification means that when the customer relationship creation or modification process is submitted, the compliance data interface is automatically invoked to conduct an initial compliance risk screening on the customer (customer, transaction entity, or entity). The compliance status is determined based on the return result of the compliance data interface. If compliant (not on the blacklist), the process proceeds to the next business process; if not compliant (on the blacklist), the process is terminated. This step is a pre-entry checkpoint for customer access. By locking business data at the bottom layer through database transaction locks, the process is blocked (i.e., new data is added at the source, and the bottom layer locks the database to block it), thus preventing high-risk customers from entering the business chain from the root.

[0074] Specifically, the incremental traversal in this step adopts a local directional traversal method, loading only the local subgraph of the customer relationship graph corresponding to the current customer incremental business for calculation; the entity feature vectors corresponding to each entity are weighted and fused with the real-time compliance return data and then input into the blacklist fuzzy matching model to complete the risk scoring; when the risk score is greater than or equal to the preset risk threshold and a database transaction lock is applied, only the data row corresponding to the current business transaction is locked, and the business transaction is marked as pending manual review and a risk warning is pushed; the stored streaming verification record includes entity identifier, matching similarity, risk score value, preset risk threshold, process handling result, and verification timestamp.

[0075] S3. Sanctions Compliance Information Acquisition Steps: At the trigger node of the ship leasing project review transaction, the preset compliance data interface is automatically called to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph are updated. Relying on the customer relationship graph, the customer and all related transaction entities under the current project are recursively traversed, and the sanctions compliance related information corresponding to each entity is automatically extracted. For new customers cooperating in this ship leasing project, a new customer sanctions compliance due diligence assessment and audit form is generated in a structured manner, and the new customer sanctions compliance due diligence assessment and audit form is bound to the current project review transaction as a mandatory attachment (i.e., a necessary attachment for project review).

[0076] Specifically, this step involves updating the entity feature vector corresponding to the entity that has received the latest sanctions compliance data only within the customer relationship graph, while the entity feature vectors of other entities remain unchanged. The recursive traversal unfolds along the multi-layered relationship chain of customer equity, actual controller, guarantor, and partner, traversing all levels of related transaction entities of the customer under the current project. Compliance fields are automatically extracted from the updated entity feature vectors and the latest sanctions compliance data, and structuredly populated into the new customer sanctions compliance due diligence assessment and review form. If the new customer sanctions compliance due diligence assessment and review form is not generated and bound as a mandatory attachment to the current project review transaction, the project review transaction is blocked from flowing to the next approval node. The updated entity feature vectors are synchronously written back to the customer relationship time-series database to retain the corresponding version snapshot.

[0077] Specifically, the new client sanctions compliance due diligence assessment audit form includes the following information: basic information (project name, department, project manager, remarks), client information (company name in Chinese and English, date of establishment, company website, registered address, contact information, registration certificate, whether it is listed on a sanctions list, etc.), shareholders and related personnel (actual controller, shareholder list, whether shareholders and related personnel are listed on any sanctions list, detailed description, etc.), and transaction participants (agents, registered addresses, etc.).

[0078] S4. Embedded Compliance Verification Steps in Project Review Nodes: At the initial review and final review submission nodes of the ship leasing project, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current project based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. The entity compliance risk level is divided according to the preset risk threshold, and the complete compliance verification results of this project are output simultaneously to complete the embedded compliance verification of all entities in the project.

[0079] This step reuses the same blacklist fuzzy matching model as step S2 to complete the risk scoring of customers and related entities; when the verification risk score reaches the preset risk threshold, since the customer's main information has been entered into the database, the current node flow is blocked at the business approval level (i.e., the mid-term node of the project, the business layer intercepts), directly terminating the entire business process of this project and pushing a risk warning; if the risk score is lower than the threshold, the verification is passed and the business process flows downward.

[0080] This step corresponds to Figure 2 As shown in Figure a, the embedded compliance review for project review automatically calls the compliance data interface to conduct compliance review when the project is submitted for the initial review and final review process. If the project complies with the rules, the subsequent S5 steps are executed; otherwise, the process is terminated.

[0081] Specifically, the compliance review scope of this step automatically covers all transaction objects under the current project, as well as all customers associated with each transaction object (i.e., customers and their related transaction entities, collectively referred to as related entities). When performing a full traversal, all related entity links of the customer relationship graph are fully loaded, and all related entities under the current project are extracted at once. The entity feature vectors of all entities are input into the blacklist fuzzy matching model in parallel batches, and batch risk scoring is calculated simultaneously. Based on the risk score range, multi-level entity compliance risk levels are divided and corresponding handling rules are matched. The output complete compliance verification result includes verification type, business document number, entity identifier, entity risk score, entity compliance risk level, blacklist matching basis, and verification timestamp. The complete compliance verification result is bound to the corresponding ship leasing project review and stored in the customer relationship time-series database for compliance evidence. If any entity is judged to be of high risk level during the traversal, the current ship leasing project review is first marked as suspended and a high-risk warning is pushed. After manual review confirms that the risk cannot be eliminated, the entire business process of this project is directly terminated. If the risk score is lower than the threshold, the verification is passed, and the business process flows down to the contract preparation node.

[0082] S5. Embedded Compliance Verification Steps at the Contract Creation Node: At the transaction trigger node for creating and saving the ship leasing contract, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and related transaction entities under the current contract based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into a blacklist fuzzy matching model and combined with the real-time compliance return data to batch complete risk scoring. Based on preset risk thresholds, entity compliance risk levels are divided, and the complete compliance verification results for this contract are output simultaneously, completing the embedded compliance verification of all contract entities. This step corresponds to... Figure 2 As shown in Figure a, the contract creation embedded compliance review automatically calls the compliance data interface to perform compliance review when the general contract is created and saved. If the contract complies with the rules, the subsequent S6 steps are executed; otherwise, the process is terminated.

[0083] Specifically, the compliance review scope of this step automatically covers all transaction objects under the current contract, as well as all customers associated with each transaction object (i.e., customers and their related transaction entities, collectively referred to as related entities). When performing a full traversal, all related entity links in the customer relationship graph are fully loaded, and all related entities under the current contract are extracted at once. The entity feature vectors of all entities are input into the blacklist fuzzy matching model in parallel batches, and batch risk scoring is calculated simultaneously. Based on the risk score range, multi-level entity compliance risk levels are divided and corresponding handling rules are matched. The output complete compliance verification result includes verification type, business document number, entity identifier, entity risk score, entity compliance risk level, blacklist matching basis, and verification timestamp. The complete compliance verification result is bound to the corresponding ship leasing contract and stored in the customer relationship time-series database for compliance evidence storage. If any entity is judged to be of high risk level during the traversal, the current contract creation transaction is first marked as suspended and a high-risk warning is pushed. After manual review confirms that the risk cannot be eliminated, the entire business process of this project is directly terminated. If the risk score is lower than the threshold, the verification is passed, and the business process flows down to the payment approval node.

[0084] S6. Embedded Compliance Verification Steps in Payment Approval Nodes: At the payment approval transaction submission node, the system first automatically calls a preset compliance data interface to obtain real-time compliance return data. It then reuses a streaming real-time verification engine to fully traverse the customers and their associated transaction entities under the current payment approval item based on the customer relationship graph. The system inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model, combining this with the real-time compliance return data to batch complete risk scoring. Based on preset risk thresholds, it classifies the entity compliance risk level and simultaneously outputs the complete compliance verification results for this payment, completing the embedded compliance verification of the entire payment entity. This step corresponds to... Figure 2 As shown in figure a, the payment approval embedded compliance review automatically calls the compliance data interface to conduct a compliance review when the payment application is submitted. If the payment is in compliance with the rules, the payment is completed; otherwise, the process is terminated.

[0085] Specifically, the compliance review scope of this step automatically covers all transaction objects under the current payment approval item, as well as all customers associated with each transaction object (i.e., customers and their related transaction entities, collectively referred to as related entities). When performing a full traversal, all related entity links in the customer relationship graph are fully loaded, and all related entities under the current payment approval item are extracted at once. The entity feature vectors of all entities are input into the blacklist fuzzy matching model in parallel batches, and batch risk scoring is calculated simultaneously. Based on the risk score range, multi-level entity compliance risk levels are divided and corresponding handling rules are matched. The output complete compliance verification result includes verification type, business document number, entity identifier, entity risk score, entity compliance risk level, blacklist matching basis, and verification timestamp. The complete compliance verification result is bound to the corresponding payment approval application and stored in the customer relationship time-series database for compliance evidence storage. If any entity is judged to be of high risk level during the traversal, the current payment approval transaction is first marked as suspended and a high-risk warning is pushed. After manual review confirms that the risk cannot be eliminated, the entire business process of this project is directly terminated. If the risk score is lower than the threshold, the verification is passed and the payment is completed.

[0086] S7. Periodic Compliance Verification Steps for Existing Customers / Projects: Configure a periodic compliance verification mechanism for existing customers or projects. This mechanism is a manually triggered mode, divided into manual single or full compliance verification for customers and manual single or full compliance verification for projects. This step also centrally stores and visualizes all records of compliance verification operations performed by calling preset compliance data interfaces and the corresponding verification results. The records of compliance verification operations include manual compliance verification records and automatic compliance verification records generated by automatically calling preset compliance data interfaces at key business transaction nodes. The key business transaction nodes include customer creation or information change business transaction submission nodes, ship leasing project review transaction trigger nodes, project preliminary and final review transaction submission nodes, ship leasing contract creation and saving transaction trigger nodes, and payment approval transaction submission nodes.

[0087] Specifically, in this regular compliance verification mechanism, the three dimensions of verification type, operation source, and verification scope can all be customized and selected. The verification type includes two categories: customer verification and project verification. The operation source is divided into manual triggering and automatic triggering of each business process. The verification scope supports two modes: single data verification and batch data verification.

[0088] This step corresponds to Figure 2 As shown in b, regular compliance checks on existing customers / projects are conducted, such as Figure 2 As shown in b, after the scheduled task starts, compliance verification is performed in batches. The verification results are divided into two categories: normal and abnormal. The compliance status of the customer / project is updated synchronously when the status is normal, and the abnormal status enters the special risk handling process and is marked for manual review and handling.

[0089] S8. Full-process monitoring steps for international sanctions risks of ships: Obtain real-time and historical information such as ship routes, ship status, and destination ports through big data interfaces, and realize automated pre-warning, in-process reporting, and post-event retrospective full-process monitoring for ships involved in international sanctions areas.

[0090] This step corresponds to Figure 2 As shown in c, the monitoring of international sanctions risks to ships, such as Figure 2 As shown in c, by connecting to a third-party shipping platform through a big data interface to synchronize ship navigation dynamic data (i.e., ship data source), three types of automated control actions are executed: pre-warning, real-time reporting during the event, and post-event retrospective analysis, which fully covers risk control throughout the entire ship navigation lifecycle.

[0091] Specifically, the big data interface connects to the enterprise's internal compliance risk system and (internal and external) third-party shipping service platform APIs. By connecting to the enterprise's internal compliance risk system, it retrieves the historical compliance risk control files of customers and their related transaction entities. By configuring daily scheduled tasks to call the third-party shipping service platform API, it synchronously obtains the destination port information or real-time dynamic location information of the vessel's planned voyage segment, as well as the corresponding international sanctioned area data.

[0092] The pre-warning includes: configuring a daily scheduled task to call the API of a third-party shipping service platform to synchronously obtain the destination port information and international sanctioned area data of the planned voyage segment to be executed by the ship, and verifying in advance whether the destination port of the planned voyage segment to be executed by the ship belongs to an international sanctioned area; if there is a ship planning to enter an international sanctioned area, an early warning reminder will be pushed by email. The early warning reminder includes ship information and the international sanctioned area matching judgment result. The ship information includes ship name, voyage start country, voyage start port, voyage end country, destination port, voyage start time, voyage end time, etc.

[0093] The in-process reporting includes: configuring a daily scheduled task to call a third-party shipping service platform API to synchronously obtain the real-time dynamic location information of the vessel and data on international sanctioned areas, and determining whether the vessel's current location is within an international sanctioned area; if a vessel is already within an international sanctioned area, a reporting reminder message will be pushed to the user via email. This reporting reminder message includes the vessel information and the international sanctioned area matching determination result. The vessel information includes the vessel name, the region where the vessel is located, the location information, and the current latitude and longitude.

[0094] The post-event retrospective includes: centrally storing and visually displaying all historical data on ships entering and leaving international sanctioned areas, supporting historical data retrospective queries, and regularly sending relevant data reports on international sanctioned areas via email. These data reports adopt a fixed format and summarize all historical records of ships entering and leaving international sanctioned areas.

[0095] This invention also relates to a big data-based tiered compliance verification system for ship leasing customers, corresponding to the aforementioned big data-based tiered compliance verification method for ship leasing customers. It can be understood as a system implementing the big data-based tiered compliance verification method for ship leasing customers. This system includes a customer information initialization and association graph construction module, an incremental customer streaming real-time initial screening module, a sanctions compliance information association acquisition module, an embedded compliance verification module for project review nodes, an embedded compliance verification module for contract preparation nodes, an embedded compliance verification module for payment approval nodes, a periodic compliance verification module for existing customers / projects, and a full-process monitoring module for international sanctions risks in the ship industry.

[0096] The customer information initialization and association graph construction module responds to customer creation or information change business transactions, automatically obtains relevant compliant business data streams of customers through a preset compliant data interface, and performs field mapping, de-identification cleaning, and unified format normalization processing on the business data streams to obtain standardized customer basic data. The standardized customer basic data is collected and stored in the customer relationship management system. Entity feature vectors of customers and their related transaction entities are extracted from the standardized customer basic data. A customer association graph is constructed based on the entity feature vectors, and a snapshot of the graph is generated. The standardized customer basic data and the generated customer association graph snapshot are written into the customer relationship time series database.

[0097] The front-end interface for entering original customer compliance information is the New Customer Sanctions Compliance Due Diligence Assessment and Review Form, as shown in the image. Figure 3 As shown, Figure 3 This is a screenshot of the interface of the new customer sanctions compliance due diligence assessment and review form of the system of this invention. The form collects all the compliance information required for the business in steps through multiple tabs, including basic information, customer information, shareholders and related personnel, transaction participants, transaction target, transaction currency and bank, assessment conclusion, etc. It fully enters key compliance information such as customer business registration information, equity related entities, transaction-related materials, and sanctions list verification results. After the form is saved, the original compliance business data stream is generated. The customer information is input into the initialization and association graph construction module to complete the automated cleaning, integration and association graph construction.

[0098] The incremental customer streaming real-time initial screening module, at the customer creation or information change business transaction submission node, first automatically calls the preset compliance data interface to obtain real-time compliance return data, and then calls the streaming real-time verification engine. Based on the constructed customer relationship graph, it incrementally traverses the customers and their related transaction entities covered by the incremental customer business data corresponding to this submission. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock is automatically applied to the current business transaction to block the subsequent business process flow. If the risk score is less than the preset risk threshold, the business transaction is allowed and the streaming verification record is stored, completing the compliance initial screening of the customer and related entities corresponding to this incremental transaction.

[0099] The sanctions compliance information acquisition module, at the trigger node of the ship leasing project review, first automatically calls a preset compliance data interface to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, it updates the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph. Relying on the customer relationship graph, it recursively traverses the customers and all related transaction entities under the current project, automatically extracts the sanctions compliance related information of each entity, and for new customers cooperating in this ship leasing project, it structurally fills in and generates a new customer sanctions compliance due diligence assessment and review form, and binds the new customer sanctions compliance due diligence assessment and review form to the current project review as a mandatory attachment. The new customer sanctions compliance due diligence assessment and review form automatically generated by this module is a system-filled archive attachment, which does not require manual filling. It is generated automatically by automatically retrieving external sanctions interface data and reading the completed customer relationship graph data, and is only used as a mandatory electronic archive attachment for project review.

[0100] The embedded compliance verification module at the project review node automatically calls a preset compliance data interface to obtain real-time compliance return data at the initial review and final review submission nodes of the ship leasing project. It reuses the streaming real-time verification engine to fully traverse the customers and related transaction entities under the current project based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on the preset risk threshold, it classifies the entity compliance risk level and outputs the complete compliance verification results of this project simultaneously, thus completing the embedded compliance verification of all entities in the project.

[0101] The embedded compliance verification module at the contract creation node automatically calls a preset compliance data interface to obtain real-time compliance return data at the transaction trigger node for creating and saving the ship leasing contract. It then reuses the streaming real-time verification engine to fully traverse the customers and their related transaction entities under the current contract based on the customer relationship graph. The module inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on preset risk thresholds, it classifies the entity compliance risk level and simultaneously outputs the complete compliance verification result of this contract, thus completing the full entity embedded compliance verification of the contract.

[0102] The embedded compliance verification module at the payment approval node automatically calls a preset compliance data interface to obtain real-time compliance return data at the payment approval transaction submission node. It reuses the streaming real-time verification engine to fully traverse the customers and their related transaction entities under the current payment approval item based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into a blacklist fuzzy matching model and combines them with the real-time compliance return data to complete risk scoring in batches. Based on the preset risk threshold, it classifies the entity compliance risk level and outputs the complete compliance verification result of this payment simultaneously, thus completing the full entity embedded compliance verification of the payment.

[0103] The existing customer / project periodic compliance verification module configures a periodic compliance verification mechanism for existing customers or projects. This mechanism is a manually triggered mode, divided into manual single or full compliance verification for customers and manual single or full compliance verification for projects. It centrally stores and visualizes all records of compliance verification operations performed by calling preset compliance data interfaces, along with the corresponding verification results. These records include manual compliance verification records and automatically generated compliance verification records at key business transaction nodes. Key business transaction nodes include customer creation or information change business transaction submission nodes, ship leasing project review transaction trigger nodes, project preliminary and final review transaction submission nodes, ship leasing contract creation and saving transaction trigger nodes, and payment approval transaction submission nodes.

[0104] The manual compliance verification interface for customers is as follows: Figure 4 As shown, Figure 4This is a screenshot of the manual compliance verification interface for customers in the system of this invention. The page title is "Customer Blacklist Verification," and it allows for multi-dimensional query and filtering conditions, including customer code, customer name, customer abbreviation, business department, project manager, customer category, and customer type. An export button is located in the upper right corner of the page. The interface supports three manual compliance verification operation modes: single customer query verification, batch blacklist-specific verification, and batch verification of all customers in the system. The list below displays the basic profile information of existing customers. After the operator selects the corresponding customer item, the system calls the blacklist fuzzy matching model to perform matching operations on the customer and all its associated entities and complete the risk scoring. Based on the scoring results, the entity's compliance risk level is divided, and finally, the corresponding manual compliance verification record is generated.

[0105] Specifically, this module centrally stores and visualizes all compliance verification operation records and corresponding verification results throughout the entire process. All records generated by automatic verification, manual periodic verification, and scheduled batch verification at all business nodes are archived uniformly. A complete compliance verification record can include the following information: verification type (customer / project), customer / project number, customer name, verification result (compliant / non-compliant), operator, operation time, operation source (manual / automatically triggered by business process node), and raw data returned by the interface. This set of fields completely distinguishes between manual verification records and automatic verification records of business nodes, facilitating unified storage, traceability, and visualization.

[0106] All compliance verification records are uniformly summarized and displayed in the record list interface, such as Figure 5 As shown, Figure 5 This is a screenshot of the customer verification record interface in the system of this invention. The list fully presents the verification type, project / customer number, customer name, verification result, operator, operation time, operation source and other fields for each verification record. It can clearly distinguish the verification records from three sources: scheduled tasks, automatic triggering of business process nodes and manual operations. The top of the page is configured with multi-dimensional query and filtering conditions, and the upper right corner supports the export of verification record data.

[0107] The aforementioned full-process monitoring module for international sanctions risks of ships obtains real-time and historical information on ship routes, ship status, and destination ports through big data interfaces, and performs full-process monitoring of ships involved in international sanctions areas, including pre-warning, in-process reporting, and post-event retrospective.

[0108] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail with reference to the accompanying drawings and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention. In short, all technical solutions and improvements that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention patent.

Claims

1. A big data-based method for tiered compliance verification of ship leasing customers, characterized in that, Includes the following steps: Customer information initialization and association graph construction steps: In response to customer creation or information change business transactions, the system automatically obtains relevant compliant business data streams of customers through a preset compliant data interface, and performs field mapping, de-identification cleaning, and unified format normalization on the business data streams to obtain standardized customer basic data. The standardized customer basic data is collected and stored in the customer relationship management system. Entity feature vectors of customers and their related transaction entities are extracted from the standardized customer basic data. A customer association graph is constructed based on the entity feature vectors, and a snapshot of the graph is generated. The standardized customer basic data and the generated customer association graph snapshot are written into the customer relationship time series database. Incremental customer streaming real-time initial screening steps: At the customer creation or information change business transaction submission node, the preset compliance data interface is automatically called to obtain real-time compliance return data. Then, the streaming real-time verification engine is called to incrementally traverse the customers and their related transaction entities covered by the customer incremental business data corresponding to this submission based on the constructed customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock will be automatically applied to the current business transaction to block the flow of subsequent business processes. If the risk score is less than the preset risk threshold, the business transaction is allowed and the current streaming verification record is stored to complete the initial compliance screening of the customer and related entities corresponding to this increment. Sanctions compliance information acquisition steps: At the trigger node of the ship leasing project review, the preset compliance data interface is automatically called to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph are updated. Relying on the customer relationship graph, the customer and all related transaction entities under the current project are recursively traversed, and the sanctions compliance related information of each entity is automatically extracted. For new customers cooperating in this ship leasing project, a new customer sanctions compliance due diligence assessment and audit form is generated in a structured manner, and the new customer sanctions compliance due diligence assessment and audit form is bound to the current project review as a mandatory attachment. Embedded compliance verification steps in project review nodes: At the initial review and final review task submission nodes of ship leasing projects, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current project based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity compliance risk level is classified, and the complete compliance verification results of this project are output simultaneously, thus completing the embedded compliance verification of the entire entity of the project. Embedded compliance verification steps in contract creation node: At the transaction trigger node for creating and saving ship leasing contracts, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current contract based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity compliance risk level is classified, and the complete compliance verification results of this contract are output simultaneously, thus completing the full entity embedded compliance verification of the contract. Embedded compliance verification steps in payment approval node: At the payment approval transaction submission node, the preset compliance data interface is automatically called to obtain real-time compliance return data. The streaming real-time verification engine is reused to fully traverse the customers and their related transaction entities under the current payment approval item based on the customer relationship graph. The entity feature vectors corresponding to each entity are input into the blacklist fuzzy matching model and combined with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity's compliance risk level is classified, and the complete compliance verification results for this payment are output simultaneously, thus completing the embedded compliance verification of the entire entity for the payment. The process of monitoring the risks of international sanctions against ships involves: obtaining real-time and historical information on ship routes, ship status, and destination ports through big data interfaces; and implementing full-process monitoring of ships involved in international sanctions areas, including pre-warning, in-process reporting, and post-event review.

2. The method for tiered compliance verification of ship leasing customers based on big data as described in claim 1, characterized in that, Following the embedded compliance verification step at the payment approval node, there is also a periodic compliance verification step for existing customers / projects: Configure a periodic compliance verification mechanism for existing clients or projects; the periodic compliance verification mechanism is a manual trigger mode, which is divided into manually performing single or all compliance verifications for clients and manually performing single or all compliance verifications for projects. It also centrally stores and visualizes all records of compliance verification operations conducted by calling preset compliance data interfaces and the corresponding verification results. The records of compliance verification operations include manual compliance verification records and automatic compliance verification records generated by automatically calling preset compliance data interfaces at key nodes of each business transaction. The key nodes of the business transaction include the customer creation or information change business transaction submission node, the ship leasing project review transaction trigger node, the project preliminary review and final review transaction submission node, the ship leasing contract creation and saving transaction trigger node, and the payment approval transaction submission node.

3. The method for tiered compliance verification of ship leasing customers based on big data as described in claim 2, characterized in that, In the process of monitoring the risks of international sanctions against ships, the advance warning includes: configuring a daily scheduled task to call the API of a third-party shipping service platform to synchronously obtain the destination port information and international sanctioned area data of the planned voyage segment to be executed by the ship, and verifying in advance whether the destination port of the planned voyage segment to be executed by the ship belongs to the international sanctioned area; if there are ships planning to enter the international sanctioned area, an early warning reminder message will be pushed. The in-process reporting includes: configuring a daily scheduled task to call the API of a third-party shipping service platform to synchronously obtain the real-time dynamic location information of the vessel and data on international sanctioned areas, and determining whether the vessel's current location is within an international sanctioned area; if a vessel is already within an international sanctioned area, a reporting reminder message will be pushed. The post-event retrospective includes: centrally storing and visually displaying all historical data on ships entering and leaving international sanctioned areas, supporting historical data retrospective queries, and regularly pushing relevant data reports on international sanctioned areas.

4. The method for tiered compliance verification of ship leasing customers based on big data as described in any one of claims 1 to 3, characterized in that, In the customer information initialization and association graph construction steps, the preset compliance data interface serves as a unified compliance standardization big data intermediate interface, with a built-in unified data transmission protocol and metadata conversion mapping table, and provides a standardized risk feature vector output port to the outside world; the compliance return data obtained by subsequent steps by calling the preset compliance data interface includes identification data used to distinguish whether a customer or vessel is compliant; The standardized customer base data includes customer name, project source, and customer characteristics; The customer relationship management system is used to perform customer relationship management operations, which include maintaining customer needs, pricing plans, and visit record information.

5. The method for tiered compliance verification of ship leasing customers based on big data as described in any one of claims 1 to 3, characterized in that, In the incremental customer streaming real-time initial screening step, the incremental traversal adopts a local directional traversal method, only loading the local graph subgraph in the customer relationship graph corresponding to the current incremental customer business to participate in the calculation; the entity feature vector corresponding to each entity is weighted and fused with the real-time compliance return data and then input into the blacklist fuzzy matching model to complete the risk scoring; When the risk score is greater than or equal to the preset risk threshold and a database transaction lock is applied, only the data row corresponding to the current business transaction is locked, and the business transaction is marked as pending manual review and a risk warning is pushed. The stored streaming verification records include entity identifier, matching similarity, risk score, preset risk threshold, process handling result, and verification timestamp.

6. The method for tiered compliance verification of ship leasing customers based on big data according to any one of claims 1 to 3, characterized in that, In the steps of obtaining the associated sanctions compliance information, only the entity feature vector corresponding to the latest sanction compliance data is updated in the customer relationship graph, while the entity feature vectors of other entities remain unchanged. The recursive traversal unfolds along the multi-layered relationship links of customer equity, actual controller, guarantor, and partner, traversing all levels of related transaction entities of the customer under the current project. Compliance fields are automatically extracted from the updated entity feature vectors and the latest sanction compliance data and structuredly populated into the new customer sanctions compliance due diligence assessment and review form. If the new customer sanctions compliance due diligence assessment and review form is not generated and bound as a mandatory attachment to the current project review transaction, the project review transaction is blocked from flowing to the next approval node. The updated entity feature vectors are synchronously written back to the customer relationship time-series database to retain the corresponding version snapshot.

7. The method for tiered compliance verification of ship leasing customers based on big data as described in any one of claims 1 to 3, characterized in that, In the embedded compliance verification steps at the project review node, contract creation node, and payment approval node, a full traversal is performed, fully loading all related entity links in the customer relationship graph. All related entities under the current business item are extracted at once, and the entity feature vectors of all entities are input into the blacklist fuzzy matching model in parallel batches, simultaneously completing batch risk scoring calculations. Based on the risk scoring range, multiple levels of entity compliance risk are defined and corresponding handling rules are matched. The complete output compliance verification result includes verification type, business document number, entity identifier, entity risk score, entity compliance risk level, blacklist matching basis, and verification timestamp. After the complete compliance verification results are linked with the corresponding ship leasing project review, ship leasing contract or payment approval application, they are stored in the customer relationship time sequence database for compliance record keeping. If any entity is determined to be of high risk level during the traversal, the current ship leasing project review, contract preparation, or payment approval process will be marked as suspended, and a high-risk warning will be sent.

8. The method for tiered compliance verification of ship leasing customers based on big data as described in claim 3, characterized in that, In the process of monitoring the risks of international sanctions against ships, the big data interface connects to the company's internal compliance risk system and the API of a third-party shipping service platform. By connecting to the company's internal compliance risk system, the historical compliance risk control files of customers and their related transaction entities can be retrieved. By configuring daily scheduled tasks to call the API of the third-party shipping service platform, the destination port information or real-time dynamic location information of the ship's planned voyage segment and the corresponding international sanctions area data can be obtained synchronously. The aforementioned early warning and notification messages, reporting reminder messages, and data reports related to international sanctioned zones are all pushed via email. In the aforementioned advance warning, the warning reminder information pushed by the email includes ship information and the international sanctions region matching determination result. The ship information includes ship name, country of origin of voyage, port of origin of voyage, country of destination of voyage, start time of voyage, and end time of voyage. In the aforementioned in-process reporting, the reporting reminder information pushed via email includes vessel information and the international sanctions region matching determination results. The vessel information includes the vessel name, the region where the vessel is located, location information, and current latitude and longitude. In the post-event retrospective analysis, the data reports related to international sanctioned zones, which are regularly sent via email, adopt a fixed format and summarize all historical records of ships entering and leaving international sanctioned zones.

9. A big data-based tiered compliance verification system for ship leasing customers, characterized in that, This includes a module for initializing and constructing a relational graph of customer information through sequential communication connections, a module for real-time streaming initial screening of incremental customers, a module for obtaining related sanctions compliance information, a module for embedded compliance verification at project review nodes, a module for embedded compliance verification at contract drafting nodes, a module for embedded compliance verification at payment approval nodes, and a module for full-process monitoring of international sanctions risks for ships. The customer information initialization and association graph construction module responds to customer creation or information change business transactions, automatically obtains relevant compliant business data streams of customers through a preset compliant data interface, and performs field mapping, de-identification cleaning, and unified format normalization processing on the business data streams to obtain standardized customer basic data. The standardized customer basic data is collected and stored in the customer relationship management system. Entity feature vectors of customers and their related transaction entities are extracted from the standardized customer basic data. A customer association graph is constructed based on the entity feature vectors, and a snapshot of the graph is generated. The standardized customer basic data and the generated customer association graph snapshot are written into the customer relationship time series database. The incremental customer streaming real-time initial screening module, at the customer creation or information change business transaction submission node, first automatically calls the preset compliance data interface to obtain real-time compliance return data, and then calls the streaming real-time verification engine. Based on the constructed customer relationship graph, it incrementally traverses the customers and their related transaction entities covered by the incremental business data corresponding to this submission, and inputs the entity feature vectors corresponding to each entity into the blacklist fuzzy matching model and combines them with the real-time compliance return data to complete the risk scoring. If the risk score is greater than or equal to the preset risk threshold, a database transaction lock will be automatically applied to the current business transaction to block the flow of subsequent business processes. If the risk score is less than the preset risk threshold, the business transaction is allowed and the current streaming verification record is stored to complete the initial compliance screening of the customer and related entities corresponding to this increment. The sanctions compliance information acquisition module, at the trigger node of the ship leasing project review, first automatically calls a preset compliance data interface to obtain the latest sanctions compliance return data of the customer and its related transaction entities. Based on the latest sanctions compliance return data, it updates the entity feature vectors corresponding to the customer and its related transaction entities in the customer relationship graph. Relying on the customer relationship graph, it recursively traverses the customers and all related transaction entities under the current project, automatically extracts the sanctions compliance related information corresponding to each entity, and for new customers cooperating in this ship leasing project, it structurally fills in and generates a new customer sanctions compliance due diligence assessment and audit form, and binds the new customer sanctions compliance due diligence assessment and audit form to the current project review as a mandatory attachment. The embedded compliance verification module in the project review node automatically calls the preset compliance data interface to obtain real-time compliance return data at the initial review and final review task submission nodes of the ship leasing project. It reuses the streaming real-time verification engine to fully traverse the customers and their related transaction entities under the current project based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into the blacklist fuzzy matching model and combines them with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity compliance risk level is classified, and the complete compliance verification results of this project are output simultaneously, thus completing the embedded compliance verification of the entire entity of the project. The embedded compliance verification module at the contract creation node automatically calls the preset compliance data interface to obtain real-time compliance return data at the transaction trigger node of ship leasing contract creation and saving. It reuses the streaming real-time verification engine and fully traverses the customers and their related transaction entities under the current contract based on the customer relationship graph. It inputs the entity feature vectors corresponding to each entity into the blacklist fuzzy matching model and combines them with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity compliance risk level is classified, and the complete compliance verification results of this contract are output simultaneously, thus completing the full entity embedded compliance verification of the contract. The payment approval node has an embedded compliance verification module. At the payment approval transaction submission node, it first automatically calls the preset compliance data interface to obtain real-time compliance return data, reuses the streaming real-time verification engine, and fully traverses the customers and their related transaction entities under the current payment approval item based on the customer relationship graph. It then inputs the entity feature vectors corresponding to each entity into the blacklist fuzzy matching model and combines them with the real-time compliance return data to complete the risk scoring in batches. Based on preset risk thresholds, the entity's compliance risk level is classified, and the complete compliance verification results for this payment are output simultaneously, thus completing the embedded compliance verification of the entire entity for the payment. The aforementioned full-process monitoring module for international sanctions risks of ships obtains real-time and historical information on ship routes, ship status, and destination ports through big data interfaces, and performs full-process monitoring of ships involved in international sanctions areas, including pre-warning, in-process reporting, and post-event retrospective.

10. The big data-based tiered compliance verification system for ship leasing customers as described in claim 9, characterized in that, Between the embedded compliance verification module at the payment approval node and the full-process monitoring module for international ship sanctions risks, there is also a periodic compliance verification module for existing clients / projects: Configure a periodic compliance verification mechanism for existing clients or projects; the periodic compliance verification mechanism is a manual trigger mode, which is divided into manually performing single or all compliance verifications for clients and manually performing single or all compliance verifications for projects. The system centrally stores and visualizes all records of compliance verification operations performed by calling preset compliance data interfaces, along with the corresponding verification results. These records include manual compliance verification records and automatic compliance verification records generated by automatically calling preset compliance data interfaces at key business transaction nodes. The key business transaction nodes include the customer creation or information change business transaction submission node, the ship leasing project review transaction trigger node, the project preliminary and final review transaction submission node, the ship leasing contract creation and saving transaction trigger node, and the payment approval transaction submission node.

Images