Method and apparatus for confirming export and import requirements through rule-based electronic document processing algorithm

Abstract

A method and apparatus for verifying import / export requirements through a rule-based electronic document processing algorithm are disclosed. A method for verifying import / export requirements according to one embodiment of the present disclosure is characterized by a method for verifying import / export requirements through a rule-based electronic document processing algorithm in which at least a portion of each step is performed by a processor to process any one of import / export customs clearance, verification of requirements, and exemption from requirements by digitizing information exchange with a Customs Service server through a rule-based electronic document processing algorithm.

Description

Technology Field

[0001] The present invention relates to a method and apparatus for verifying import / export requirements through a rule-based electronic document processing algorithm, which processes one of the procedures of import / export customs clearance, verification of requirements, and exemption from requirements by digitizing information exchange with an external server through a rule-based electronic document processing algorithm. Background Technology

[0003] International trade and import / export customs clearance procedures, involving the submission of various documents, data verification, and complex administrative processes, have a significant impact on the efficiency and reliability of trade transactions. Traditional customs systems have relied primarily on manual methods or limited automation techniques, exchanging information between agencies using fixed communication protocols and rigid rules. While this approach contributed to maintaining consistency in document formats, content, and processing results, it failed to effectively reflect the diverse variables and unpredictable data patterns arising in the rapidly changing international trade environment. Consequently, errors, data omissions, and inconsistencies have become frequent during the customs process, leading to delays in overall trade transactions and increased costs.

[0004] Conventional technologies utilized electronic document processing systems that relied primarily on static rules. While these systems facilitated information exchange between the Korea Customs Service, requirements verification agencies, and other relevant departments using standardized data formats, they suffered from data inconsistencies between systems, security vulnerabilities, and a lack of flexibility to adapt to the changing customs environment. For instance, as the data formats and verification standards used by each agency were fixed, manual supplementation or modification became necessary whenever international trade regulations or practices changed. Furthermore, limitations existed in preventing forgery and ensuring data integrity of electronic documents, making it impossible to fully guarantee the reliability of related records.

[0005] With the recent increase in international trade, the automation of customs clearance procedures and the advancement of data processing have emerged as essential tasks. As the complexity of trade transactions grows, there is a growing need for research on dynamic rule-based processing algorithms capable of strengthening the structural and semantic verification of electronic documents and reflecting real-time data patterns. In particular, research on methods to enhance the efficiency and accuracy of customs clearance procedures by utilizing artificial intelligence and machine learning techniques to analyze historical customs records and real-time data, and by dynamically updating rules based on the results, is emerging as a critical research topic.

[0006] Furthermore, alongside advancements in security technology, there is a growing demand for systems that integrate blockchain-based distributed ledger technology into electronic document processing systems to prevent data falsification and enable transparent audit trails. As such, research and development of advanced electronic document processing technologies are essential to overcome the limitations of existing customs clearance systems and proactively respond to changes in the international trade environment. This can significantly contribute to the streamlining of import and export customs clearance, requirement verification, and exemption procedures, as well as to the enhancement of trade competitiveness.

[0007] The aforementioned background technology is technical information that the inventor possessed for the derivation of the present invention or acquired during the process of deriving the present invention, and it cannot be considered as prior art disclosed to the general public prior to the filing of the present invention. Prior art literature

[0009] Prior Art 1: Korean Registered Patent Publication No. 10-0397868 (September 1, 2003) The problem to be solved

[0010] One objective of the embodiment of the present disclosure is to digitize information exchange between servers based on an electronic document processing algorithm to conveniently process any one of the procedures for import / export customs clearance, verification of requirements, and exemption from requirements.

[0011] One objective of the embodiment of the present disclosure is to minimize input errors in electronic documents and significantly improve the efficiency and accuracy of customs clearance operations by implementing a RULE-based electronic document distribution and import / export requirement verification procedure between servers that automatically completes the input of items according to the settings of total frequency, individual frequency, and quarterly frequency.

[0012] One objective of the embodiments of the present disclosure is to integrate the calculation and real-time updating of AI-based dynamic RULE mapping criteria, the enhancement of structural and semantic verification, the blockchain-based integrity record, and the auto-completion processing function into the electronic document processing process between servers.

[0013] One objective of the embodiment of the present disclosure is to ensure that the requirements verification application data and customs clearance history within the electronic document are precisely preprocessed through processes such as quantitative feature extraction, standardization, and dimensionality reduction, and that dynamic RULE mapping criteria are calculated through a deep learning or reinforcement learning model.

[0014] One objective of the embodiment of the present disclosure is to prevent forgery of electronic documents and enable transparent audit tracking through a distributed ledger by applying a secure consensus algorithm to all relevant data, including dynamic RULE update history and processing logs, after verification results, assigning a digital signature and a cryptographic hash value, and then recording it on a blockchain network.

[0015] One objective of the embodiments of the present disclosure is to enable rapid verification of requirements and customs clearance examination without errors or delays caused by manual work, since information exchange between servers, electronic documentation, and automatic completion processing functions are performed automatically by a processor.

[0016] One objective of the embodiments of the present disclosure is to detect and respond to errors that may occur in customs clearance procedures at an early stage by promptly generating an automatic warning and initiating an additional verification procedure when abnormal signs or discrepancies occur through real-time updates of dynamic RULE mapping criteria and an additional verification procedure.

[0017] The purpose of the embodiments of the present disclosure is not limited to the problems mentioned above, and other unmentioned purposes and advantages of the present invention may be understood from the following description and will be more clearly understood by the embodiments of the present invention. Furthermore, it will be understood that the purposes and advantages of the present invention can be realized by the means and combinations thereof set forth in the claims. means of solving the problem

[0019] An import / export requirement verification system for verifying import / export requirements between a RULE-based electronic document distribution system and the Korea Customs Service according to one embodiment of the present disclosure comprises, wherein each step is performed by a processor, a requirement application information electronic document generation step in which, when an application for verification of requirements for import / export products is received through a Korea Customs Service server, requirement application information data is received from the Korea Customs Service server using a SOAP standard communication protocol and an electronic document of requirement application information including processing information and application information is generated; and a final customs clearance requirement verification step in which a final customs clearance requirement verification procedure is performed to reflect the result of verification of requirements necessary for import / export customs clearance procedures in the electronic document of requirement application information.

[0020] A method for verifying import / export requirements through a rule-based electronic document processing algorithm according to one embodiment of the present disclosure is a method for verifying import / export requirements through a rule-based electronic document processing algorithm in which at least a portion of each step is performed by a processor to process any one of import / export customs clearance, requirement verification, and requirement exemption by digitizing information exchange with a Korea Customs Service server through a rule-based electronic document processing algorithm, wherein when an application for requirement verification for import / export products is received through the Korea Customs Service server, requirement application information data is received from the Korea Customs Service server using a SOAP standard communication protocol, and the received raw data is converted into a predefined XML electronic document format to generate a requirement application information electronic document in which Collaboration Protocol Profile (CPP) information, Collaboration Protocol Agreement (CPA) information, and essential customs clearance information are inserted into the electronic document; A preprocessing step for verifying an electronic document of application for requirements information, wherein the format, required items, data consistency, and accuracy of the generated electronic document of application for requirements information are verified according to a predefined schema (XSD) and related standards, and the application for requirements confirmation data and related customs clearance history included in the electronic document of application for requirements information are aggregated, then features are quantitatively extracted from each data item, noise in the input data is minimized by applying at least one technique among statistical outlier removal, Z-score normalization, and Min-Max scaling, and the input data is preprocessed into an optimal input data form that a machine learning model can learn through Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor Embedding (t-SNE) techniques;A dynamic RULE mapping standard calculation and update step, wherein a deep learning or reinforcement learning model is trained using the aforementioned preprocessed data to calculate an optimal RULE mapping standard that precisely reflects changes in the customs clearance environment and data patterns by comparing it with a pre-set statically configured RULE mapping standard, and a pre-set RULE mapping table is updated in real time based on the calculated learning result; an electronic document verification step, wherein the updated RULE mapping standard is applied to the aforementioned electronic document of requirement application information to automatically extract, compare, and analyze the format, consistency, and alignment of CPP / CPA information and other customs clearance-related data included in the aforementioned electronic document of requirement application information, and verify whether the aforementioned electronic document of requirement application information conforms to a pre-defined dynamic standard based on the comparison and analysis results; and a blockchain recording step, wherein, based on the verification result, all relevant processing data including dynamic RULE update history and electronic document processing logs are organized into a single data block, data integrity is ensured by assigning a digital signature and a cryptographic hash value, and the block is permanently stored in an immutable distributed ledger by applying a secure consensus algorithm to record the block on a blockchain network. and may include a final customs clearance requirement verification step that automatically receives the electronic document of the requirement application information and processing history recorded in the blockchain network and performs a final customs clearance requirement verification procedure that reflects the results of verifying the requirements necessary for the import / export customs clearance procedure.;

[0021] In some examples, after the final customs clearance requirement verification step, the method may further include: a step of automatically initiating and performing an import / export declaration examination procedure based on relevant electronic documents and data including requirement application information and examination history for import / export products for which the requirement examination has been completed through the Korea Customs Service server; a step of receiving the import / export declaration examination result from the Korea Customs Service server, analyzing the received import / export declaration examination result data to generate an electronic examination result document including the examination result, related processing history, and examination details; and a step of executing an algorithm that automatically maps the requirement application information electronic document and the import / export declaration examination result electronic document based on the electronic examination result document to verify consistency between the requirement application information and the final customs clearance result for the import / export products.

[0022] In some examples, the preprocessing step for verifying the electronic document of the requirement application information may include: parsing the electronic document of the requirement application information using an XML parser and automatically comparing the existence of elements specified in the schema, the order of elements, the data format of each data item, and the maximum and minimum limits of defined values ​​through an XSD verification engine, while precisely verifying whether each element is designated as a required element, whether the elements are arranged in the defined order, and whether the data value exists within the range of values ​​specified in the schema, and automatically detecting errors such as omission, duplication, order mismatch, format error, or exceeding the value range; calculating quantitative evaluation indicators for each data item, such as the frequency of error occurrence, the degree of data inconsistency, and the consistency of cross-reference relationships between elements, based on the XSD verification results, and comparing these with a predefined threshold to determine whether the consistency and accuracy of the data within the electronic document of the requirement application information have been secured; and, after the determining step, if a verification error or inconsistency is detected, automatically generating a verification result log containing detailed information such as the error code of the error, the time of occurrence, a specific description of the error, and the location of the data item where the error occurred.

[0023] In some examples, the preprocessing step for verifying the electronic document of the above-mentioned requirement application information comprises: a step of quantifying features extracted from the aggregated data items by applying a predefined transformation function, performing statistical analysis on each quantified feature value to identify extreme values ​​of the data distribution, and removing abnormal values ​​by applying a predefined outlier removal algorithm; a step of converting each feature value into a normal distribution with a mean of 0 and a standard deviation of 1 by applying a Z-score normalization formula, or standardizing each feature value by linearly transforming it using a Min-Max scaling technique and adjusting it to a fixed range of 0 and 1. The method may include the step of applying a Principal Component Analysis (PCA) algorithm to a standardized multidimensional data matrix to calculate the variance contribution of each feature, selecting principal components that exceed a pre-set variance contribution criterion to reduce the data dimensionality, or applying a t-distributed Stochastic Neighbor Embedding (t-SNE) algorithm to generate low-dimensional embeddings, and converting the input data into an optimal input data form that removes unnecessary noise from the input data and preserves only the core features.

[0024] In some examples, the step of calculating and updating dynamic RULE mapping criteria may include: a step of constructing a multilayer perceptron-based deep learning network using a mean squared error loss function with the preprocessed data as input, calculating the gradient of the loss function for the weights and biases of the current model for each learning iteration using an error backpropagation algorithm, and repeating a learning process of updating model parameters (weights and biases) while keeping the learning rate constant using stochastic gradient descent, until a predetermined number of epochs is completed or the change in the loss value is negligible and the convergence condition is satisfied; and, after the learning is completed, calculating a prediction result for the input data using the learned model, and comparing and analyzing the prediction result with the result according to a pre-set static RULE mapping criteria, wherein the comparison is performed by quantitatively evaluating the degree of agreement between the actual customs clearance history and the prediction result using accuracy and F1 score, and calculating an optimal RULE mapping criterion that precisely reflects changes in the customs clearance environment and data patterns through an optimization algorithm that determines the optimal RULE mapping criterion based on a pre-set threshold value calculated as an evaluation metric.

[0025] In some examples, the step of calculating and updating dynamic RULE mapping criteria may include: a step of numerically evaluating the need to update each item by directly comparing the optimal RULE mapping criteria calculated by the learned model with each item of the existing static RULE mapping table; a step of replacing the corresponding item of the existing RULE mapping table by applying the calculated optimal RULE mapping criteria to the item requiring update according to the evaluation result, and generating an update log including the change history of the replaced item, the update time, and the applied RULE mapping criteria; and a step of the processor automatically verifying the consistency between the generated update log and the updated RULE mapping table, and then monitoring the update results in real time to immediately reflect them in the customs data processing system.

[0026] In some examples, the blockchain recording step may include: serializing the entire contents of the processing data in bytes and applying a predetermined cryptographic hash function to produce a fixed-length hash value; combining the produced hash value with the serialized processing data, generating a digital signature according to the RSA algorithm using a pre-registered private key, and attaching the digital signature and hash value to the corresponding data block to ensure the integrity of the data; executing a secure consensus algorithm (applying a Proof-of-Work algorithm) on the data block to produce a Proof-of-Work value that satisfies a difficulty goal based on the cryptographic hash value of the data block, transmitting the data block containing the produced Proof-of-Work value to a network so that all participating nodes on the blockchain network reach a consensus, and confirming that the data block is permanently recorded in the distributed ledger without tampering through the consensus process.

[0027] In some examples, when inputting items within an electronic document for receiving an application for verification of requirements for import / export products through the aforementioned Customs Service server, the method further includes a step of automatically recommending an optimal input value based on a predefined frequency setting for words or numbers entered by the user, wherein the recommendation step comprises: a step of quantitatively calculating the usage pattern of each item by aggregating the total usage frequency, the input frequency by a specific user, and the quarterly input frequency of the word and number data recorded in the electronic document of the application for verification of requirements received from the aforementioned Customs Service server; a step of calculating the weight of each item by applying a predefined algorithm based on the aggregated frequency data, and determining the priority of each item by considering the ratio of the total frequency, the frequency by user, and the input frequency by quarter; and a step of, when the user inputs some characters into an item of the electronic document, sorting word and number items similar to the characters in real time based on the entered characters and the calculated weight value, and displaying a recommendation list with high priority in an autocomplete window, wherein the recommendation list may be provided in a differentiated manner according to the input frequency by user and the quarterly input pattern of the data recorded in the electronic document of the application for verification of requirements received from the aforementioned Customs Service server.

[0028] An export / import requirement verification device using a rule-based electronic document processing algorithm according to one embodiment of the present disclosure is an export / import requirement verification device using a rule-based electronic document processing algorithm for processing any one of export / import customs clearance, requirement verification, and requirement exemption by digitizing information exchange with a Customs Service server through a rule-based electronic document processing algorithm, comprising: a memory; and includes at least one processor connected to the memory and configured to execute computer-readable instructions contained in the memory, wherein when an application for verification of requirements for import / export products is received through the Korea Customs Service server, the at least one processor receives requirement application information data from the Korea Customs Service server using the SOAP standard communication protocol, converts the received raw data into a predefined XML electronic document format to generate an electronic document of requirement application information that inserts Collaboration Protocol Profile (CPP) information, Collaboration Protocol Agreement (CPA) information, and essential customs clearance information into the electronic document, verifies the format, required items, data consistency, and accuracy of the generated electronic document of requirement application information according to a predefined schema (XSD) and related standards, aggregates the requirement verification application data and related customs clearance history included in the electronic document of requirement application information, quantitatively extracts features from each data item, minimizes noise in the input data by applying at least one technique among statistical outlier removal, Z-score normalization, and Min-Max scaling, and performs Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor The input data is preprocessed into an optimal form that machine learning models can learn using the Embedding (t-SNE) technique, and by training a deep learning or reinforcement learning model using the preprocessed data, an optimal RULE mapping standard is calculated that precisely reflects changes in the customs clearance environment and data patterns by comparing it with a pre-set statically configured RULE mapping standard.Based on the learning results calculated above, a pre-configured RULE mapping table is updated in real time, and the updated RULE mapping criteria are applied to the electronic document of the requirement application information. This allows for the automatic extraction, comparison, and analysis of the format, consistency, and alignment of CPP / CPA information and other customs-related data included in the electronic document of the requirement application information. Based on the comparison and analysis results, it verifies whether the electronic document of the requirement application information conforms to pre-defined dynamic criteria. Based on the verification results, all relevant processing data, including dynamic RULE update history and electronic document processing logs, is organized into a single data block. Data integrity is then ensured by assigning a digital signature and a cryptographic hash value. By applying a secure consensus algorithm to record the block on a blockchain network, it is permanently stored on an immutable distributed ledger. Finally, the electronic document of the requirement application information and processing history recorded on the blockchain network are automatically received to perform a final customs requirement verification procedure that reflects the verification results necessary for import and export customs clearance procedures.

[0029] In some examples, the at least one processor may be configured to perform preprocessing of the electronic document, calculation and updating of RULE mapping criteria, verification, and blockchain recording for each electronic document of the requirement application information in a cloud computing environment in a continuous and automatic manner.

[0030] In addition to this, other methods for implementing the present invention, other systems, and computer-readable recording media storing a computer program for executing said methods may be further provided.

[0031] Other aspects, features, and advantages other than those described above will become clear from the following drawings, claims, and detailed description of the invention. Effects of the invention

[0033] According to an embodiment of the present disclosure, by digitizing information exchange between servers, errors in the data transmission and storage process are significantly reduced compared to existing manual or distributed paper document methods. Furthermore, the digitized data is consistently managed in a digital format and can be updated and exchanged in real time, thereby improving the reliability of information and processing speed throughout the entire process of import / export customs clearance, requirement verification, and requirement exemption procedures. As a result, the complexity of customs-related tasks is alleviated, and the overall efficiency of trade transactions is enhanced.

[0034] According to an embodiment of the present disclosure, by providing jurisdictional tasks between servers based on an electronic document processing algorithm, data consistency and interoperability between agencies are enhanced. Furthermore, by automating the processes of data generation, verification, updating, and recording at each stage through the electronic document processing algorithm, and by supporting the smooth progress of import / export customs clearance, requirement verification, and requirement exemption procedures, the speed and accuracy of customs clearance operations are significantly improved, thereby reducing time delays and cost burdens in trade transactions and increasing the overall efficiency of business processes.

[0035] According to an embodiment of the present disclosure, by using an auto-completion processing function based on total frequency, individual frequency, and quarterly frequency when entering items, the optimal input value is recommended by analyzing word and number information that a user must input in real time when creating an electronic document, thereby significantly reducing errors that may occur during the input process, shortening the user's document creation time, and ensuring data consistency and accuracy, the effect of improving the efficiency and quality of customs clearance operations can be achieved.

[0036] According to an embodiment of the present disclosure, by integrally applying artificial intelligence-based dynamic RULE mapping criteria calculation and real-time updates, structural and semantic verification enhancement, blockchain-based integrity recording, and auto-completion processing functions to the electronic document processing process between servers, effects can be achieved in terms of improved accuracy, enhanced security, increased processing efficiency, and real-time response and adaptation.

[0037] According to an embodiment of the present disclosure, the application data for verification of requirements and customs clearance history within an electronic document are precisely preprocessed through processes such as quantitative feature extraction, standardization, and dimensionality reduction, and dynamic RULE mapping criteria are calculated through a deep learning or reinforcement learning model, thereby effectively reflecting changes in the customs clearance environment and diversity of data patterns, and the verification accuracy of the electronic document is greatly improved.

[0038] According to an embodiment of the present disclosure, as a result of verification, all relevant data including dynamic RULE update history and processing logs are assigned a digital signature and a cryptographic hash value, and then recorded on a blockchain network by applying a secure consensus algorithm, thereby preventing forgery of electronic documents and enabling transparent audit tracking through a distributed ledger.

[0039] According to an embodiment of the present disclosure, since information exchange between servers, electronic documentation, and automatic completion processing functions are automated and performed by a processor, requirements verification and customs clearance examination are carried out quickly without errors or delays caused by manual work, thereby significantly reducing the processing speed of the entire import and export customs clearance procedure and increasing work efficiency.

[0040] According to an embodiment of the present disclosure, through real-time updates of dynamic RULE mapping criteria and additional verification procedures, if abnormal signs or discrepancies occur, an automatic warning is quickly generated and an additional verification procedure is initiated, thereby detecting and responding to errors that may occur in customs clearance procedures at an early stage, and thus maintaining the stability and reliability of the system.

[0041] According to an embodiment of the present disclosure, by integrating a container and serverless scalable structure, reinforcement learning-based rule automation, blockchain integrity record, and ZeroTrust security through cloud-native RULE-based electronic document processing, it is possible to provide high-reliability and low-cost electronic document services uninterrupted even with large-scale traffic, and simultaneously ensure data integrity and audit traceability while complying with international regulations.

[0042] The effects of the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0044] The following drawings attached to this specification illustrate preferred embodiments of the present invention and serve to further enhance understanding of the technical concept of the present invention together with the detailed description of the invention provided below; therefore, the present invention should not be interpreted as being limited only to the matters described in such drawings. FIG. 1 is a schematic diagram illustrating an import / export requirement verification system through a RULE-based electronic document processing algorithm according to one embodiment of the present disclosure. Figure 2 is an example diagram visualizing the overall flow of import and export customs clearance procedures. FIG. 3 is a block diagram showing the server of an import / export requirement verification system in detail according to one embodiment of the present invention. FIG. 4 is a block diagram schematically showing an export / import requirement verification device according to one embodiment of the present invention. FIG. 5 is a flowchart illustrating a method for verifying import and export requirements according to one embodiment of the present invention. FIG. 6 is an exemplary diagram illustrating the configuration of an electronic document according to one embodiment of the present invention. FIGS. 7 and 8 are drawings for explaining the process of transmitting and receiving electronic documents and verifying between a Customs Service server and a requirement agency server in an import / export requirement verification device according to an embodiment of the present invention. FIG. 9a is a table showing the procedure in the import / export customs clearance process according to one embodiment of the present invention. FIG. 9b is an exemplary diagram illustrating the application procedure for either the verification of requirements or the exemption from requirements according to one embodiment of the present invention. Specific details for implementing the invention

[0045] The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described in detail together with the accompanying drawings.

[0046] However, the present invention is not limited to the embodiments presented below, but can be implemented in various different forms and should be understood to include all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention. The embodiments presented below are provided to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention. In describing the present invention, detailed descriptions of related known technologies are omitted if it is determined that such detailed descriptions may obscure the essence of the invention.

[0047] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, terms such as “comprising” or “having” are intended to indicate the presence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. Terms such as “first,” “second,” etc., may be used to describe various components, but the components should not be limited by these terms. These terms are used solely for the purpose of distinguishing one component from another.

[0048] Hereinafter, embodiments according to the present invention will be described in detail with reference to the attached drawings. In describing with reference to the attached drawings, identical or corresponding components are given the same reference numerals, and redundant descriptions thereof will be omitted.

[0049] FIG. 1 is a schematic diagram illustrating an import / export requirement verification system through a RULE-based electronic document processing algorithm according to one embodiment of the present disclosure.

[0050] Referring to FIG. 1, the import / export requirement verification system (1) may include an import / export requirement verification device (100), a user terminal (200), a server (300), and a network (400).

[0051] In some examples, the import / export requirement verification system (1) is an integrated system designed to process import / export customs clearance, requirement verification, and requirement exemption procedures quickly and accurately by digitizing information exchange between servers. The import / export requirement verification system (1) converts requirement verification application data received from the Korea Customs Service server into a predefined XML electronic document and inserts essential customs clearance information, including a cooperation protocol profile and cooperation protocol agreement information, into the electronic document. Subsequently, the application data included in the electronic document and related customs clearance history are aggregated, and dynamic RULE mapping criteria are calculated and updated in real time through artificial intelligence and machine learning algorithms, and the format, consistency, and integrity of the electronic document are automatically verified according to these criteria. In addition, the verification results and processing history are recorded on a blockchain network after being assigned digital signatures and cryptographic hash values, thereby preventing data tampering and ensuring transparency. Furthermore, by introducing an auto-completion function to minimize input errors when users create electronic documents, the efficiency and reliability of the entire import / export customs clearance process can be significantly improved.

[0052] First, regarding the import and export customs clearance procedure, Figure 2 is an example diagram visualizing the overall flow of the procedure. Each step is arranged in the form of a circular icon or box and connected by arrows, allowing the sequential progress to be checked at a glance. The procedure largely includes the following steps.

[0053] During the port entry / unloading stage, a vessel or aircraft arrives at its destination and unloads cargo. At this stage, the cargo is moved to the unloading location and subsequently transported to a bonded area during the bonded transport stage. Upon arrival at the bonded area, the goods are stored in a bonded warehouse or bonded storage facility through the bonded area entry stage. As such, the bonded transport and bonded area entry processes are procedures for moving foreign goods to an area managed under customs supervision; they are crucial steps to verify that the goods are being managed legally during subsequent customs duty assessments or clearance inspections.

[0054] Next, during the import / export declaration stage, the procedures for actually bringing goods into the country or exporting them abroad are carried out. Here, a declaration form prepared by a trader or customs broker is submitted to customs, accompanied by various documents (invoices, packing lists, certificates of origin, etc.). In the subsequent customs broker reception stage, the customs broker formally accepts the declaration documents and, if necessary, verifies the completeness of the documents or checks for errors.

[0055] Subsequently, during the goods inspection stage, customs examines the suitability of the declaration through various methods, such as physically inspecting the goods or assessing their validity solely through document review. Through this process, they verify the illegality of the goods, any violations of regulations, and the consistency between the declared details and the actual items. After undergoing the inspection process, at the import / export declaration acceptance stage, customs formally accepts the declaration if it determines that the declaration is lawful and free of issues. Once the declaration is accepted, the customs clearance procedure for the goods is effectively completed.

[0056] For repaired goods, customs duties and taxes are paid during the tax payment stage, and the final customs clearance procedure is completed when the goods are released from the bonded area during the goods release stage. This entire process involves various administrative procedures, such as document preparation, submission, and customs inspection; however, since these were previously carried out primarily manually or through limited electronic systems, issues such as missing or duplicated information between stages sometimes occurred.

[0057] Accordingly, the import / export requirement verification system (1) of the present invention can be designed to automatically collect, verify, and process data and documents generated at each stage by digitizing documents, particularly in the section from the import / export declaration to the acceptance of the import / export declaration. For example, the declaration information submitted at the import / export declaration stage is generated as an electronic document, automatically linking the document information required for the customs broker reception and goods inspection processes, and all records can be maintained consistently until the point at which the import / export declaration is finally accepted. This significantly reduces human errors occurring during the document preparation and verification process, and allows customs clearance procedures to proceed without data omission between stages, thereby significantly improving the efficiency of trade operations. In addition, through automatic verification and log recording functions, the processing status of each stage can be tracked, which has the advantage of increasing work transparency and accuracy.

[0058] Meanwhile, in one embodiment, users may access an application or website implemented on a user terminal (200) to request membership (registration) and import / export requirement verification services, check the results of the service request, and interact through the provided interface. Additionally, in one embodiment, users may access an application or website implemented on a user terminal (200) to perform a support process for providing import / export requirement verification services.

[0059] In some examples, the user terminal (200) may include a first user terminal (210), a second user terminal (220), and a third user terminal (230). In some examples, the first user terminal (210) may be a terminal of an individual who uses and requests the import / export requirements verification service, namely a trader or a customs broker acting on behalf of a trader in filing declarations. In some examples, the second user terminal (220) may be a terminal of a manager or person in charge who manages the import / export customs clearance process through the import / export requirements verification service. And the third user terminal (230) may be a terminal of a service support manager who provides support for and offers the import / export requirements verification service.

[0060] However, in one embodiment, user terminals are described separately for convenience of explanation, but are not limited thereto. That is, in one embodiment, the user terminal (200) may refer to different devices such as the first user terminal (210), the second user terminal (220), and the third user terminal (230), but may also refer to the same terminal.

[0061] Such user terminals (200) can receive an import / export requirement verification service through an authentication process after accessing an application or website. The authentication process may include, but is not limited to, authentication of user information such as membership registration and authentication of the user terminal, and the authentication process may be performed simply by accessing a link transmitted from the import / export requirement verification device (100) and / or server (300).

[0062] These user terminals (200) may be, but are not limited to, desktop computers, smartphones, laptops, tablet PCs, smart TVs, mobile phones, PDAs (personal digital assistants), laptops, media players, micro servers, GPS (global positioning system) devices, e-book readers, digital broadcasting terminals, navigation systems, kiosks, MP3 players, digital cameras, home appliances, and other mobile or non-mobile computing devices operated by the user.

[0063] Additionally, the user terminal (200) may be a wearable terminal such as a watch, glasses, a hair band, and a ring equipped with communication functions and data processing functions. The user terminal (200) is not limited to the above description, and any terminal capable of web browsing may be used without restriction.

[0064] In some examples, the import / export requirement verification system (1) may be implemented by an import / export requirement verification device (100) and / or a server (300). In some examples, the server (300) may be a server for operating the import / export requirement verification system (1) which includes the import / export requirement verification device (100).

[0065] That is, the server (300) may be a server that controls the operation of the import / export requirement verification device (100) for the overall process of providing import / export requirement verification services.

[0066] Additionally, the server (300) may be a database server that provides data for operating the import / export requirement verification device (100). Furthermore, the server (300) may include a web server or an application server.

[0067] And the server (300) may include a big data server and an AI server necessary for applying various artificial intelligence algorithms, and a computation server that performs computations of various algorithms. In some examples, the server (300) may include the aforementioned servers or network with such servers, but is not limited thereto. That is, in some examples, the server (300) may include the aforementioned web server and AI server or network with such servers.

[0068] FIG. 3 is a block diagram showing in detail a server (300) of an import / export requirement verification system (1) according to an embodiment of the present invention. Referring to FIG. 3, the server (300) may include or be referred to as a Customs Service server (310) and a requirement agency server (320). In some examples, at least one of the Customs Service server (310) and the requirement agency server (320) may be a server included within the server (300) or a separate external server. Additionally, servers other than the Customs Service server (310) and the requirement agency server (320) may be included.

[0069] In some examples, the Customs Service server (310) may include or be referred to as a civil complaint server (311) and a customs administration system (312). Also, in some examples, the requirements agency server (320) may include or be referred to as an electronic document distribution server (321) and a requirements verification server (322).

[0070] In some examples, when the civil complaint server (311) receives a requirement verification application before the import / export declaration at the requirement verification stage, it transmits the requirement application information as an electronic document to the electronic document distribution server (321), and when it receives approval / rejection of the requirement examination from the requirement verification server (322) at the import / export declaration stage, it enters the approved number into the import / export requirement number field at the time of import / export declaration and then applies for the import / export declaration.

[0071] When the requirement verification number of the civil complaint server (311) is entered during the import / export declaration examination stage, the customs administration server (312) examines the import / export declaration, examines / rejects it, checks whether the requirements are approved, and notifies the electronic document distribution server (321) of the examination result during the import / export declaration approval stage.

[0072] The electronic document distribution server (321) receives a requirement application from the civil complaint server (311) during the requirement verification stage, processes XML (eXtensible Markup Language), ebXML (electronic business XML), electronic signature, encryption / decryption, SOAP (Simple Object Access Protocol), HTTP (HyperText Transfer Protocol), and RULE mapping to create an electronic document, and during the import / export declaration approval stage, processes the examination results of the customs administration server (312) using XML, ebXML, electronic signature, encryption / decryption, SOAP, HTTP, and RULE mapping to create an electronic document and transmits them to the requirement verification server (322).

[0073] The requirement verification server (322) reviews the requirements of the electronic document of the electronic document distribution server (321) during the requirement verification stage, notifies the civil complaint server (311) of approval / rejection, and maps the requirements application information and customs clearance results of the electronic document of the electronic document distribution server (321) during the import / export declaration approval stage.

[0074] In some examples, the import / export requirement verification device (100), user terminal (200), and server (300) in the import / export requirement verification system (1) may be connected by a network (400). Such a network (400) may include wired networks such as LANs (local area networks), WANs (wide area networks), MANs (metropolitan area networks), and ISDNs (integrated service digital networks), or wireless networks such as wireless LANs, CDMA, Bluetooth, and satellite communication, but the scope of the present invention is not limited thereto. Additionally, the network (400) may transmit and receive information using short-range communication and / or long-range communication.

[0075] Additionally, the network (400) may include connections of network elements such as hubs, bridges, routers, switches, and gateways. The network (400) may include one or more connected networks, such as a multi-network environment, including a public network such as the Internet and a private network such as a secure corporate private network. Access to the network (400) may be provided through one or more wired or wireless access networks. Furthermore, the network (400) may support an Internet of Things (IoT) network and / or 5G communication that exchanges and processes information between distributed components, such as objects.

[0076] FIG. 4 is a block diagram schematically showing an export / import requirement verification device (100) according to one embodiment of the present invention.

[0077] Referring to FIG. 4, the export / import requirement verification device (100) of the present embodiment may include a communication interface (110), a user interface (120), a memory (130), and a processor (140).

[0078] The communication interface (110) may provide a communication interface necessary to provide transmission and reception signals between external devices in the form of packet data in conjunction with the network (400). Additionally, the communication interface (110) may be a device including hardware and software necessary to transmit and receive signals, such as control signals or data signals, through a wired or wireless connection with another network device.

[0079] That is, the processor (140) can receive various data or information from an external device connected through the communication interface (110), and can also transmit various data or information to the external device.

[0080] The user interface (120) may include an input interface into which user requests and commands are entered to control the operation of the import / export requirement verification device (100) (e.g., application for import / export requirement verification and result verification, application for import / export examination and result verification, setting of electronic document processing algorithms, creation of various algorithms and learning models, change of settings, setting of parameters, etc.).

[0081] In addition, the user interface (120) may include an output interface that outputs service outputs, responses, etc., in response to a request for a service to verify import / export requirements. That is, the user interface (120) can output results in response to user requests and commands. In some examples, the input interface and the output interface of the user interface (120) may be implemented as separate interfaces or implemented in the same interface. In some examples, the user interface (120) may be formed integrally with the user terminal (200) or formed separately. Furthermore, the user interface (120) may be formed integrally with the user terminal (200) and simultaneously provided separately.

[0082] The memory (130) can store various information required for the operation of the import / export requirement verification device (100) and / or the control (operation) of the server (200), and can store control software, and may include a volatile or non-volatile recording medium.

[0083] The memory (130) is connected to one or more processors (140) and can store codes that cause the processors (140) to control the import / export requirement verification device (100) and / or server (300) when executed by the processors (140).

[0084] Here, the memory (130) may include a magnetic storage medium or a flash storage medium, but the scope of the present invention is not limited thereto. The memory (130) may include an internal memory and / or an external memory, and may include a volatile memory such as DRAM, SRAM, or SDRAM, a non-volatile memory such as OTPROM (one time programmable ROM), PROM, EPROM, EEPROM, mask ROM, flash ROM, NAND flash memory, or NOR flash memory, a flash drive such as an SSD, CF (compact flash) card, SD card, Micro-SD card, Mini-SD card, Xd card, or memory stick, or a storage device such as an HDD. In addition, information related to an algorithm for performing learning according to the present disclosure may be stored in the memory (130). In addition, various information necessary within the scope of achieving the purpose of the present disclosure may be stored in the memory (130), and the information stored in the memory (130) may be updated as it is received from a server or external device or input by a user.

[0085] Such a processor (140) can control the overall operation of the import / export requirement verification device (100). Specifically, the processor (140) is connected to the import / export requirement verification device (100) including a memory (130) and can control the overall operation of the import / export requirement verification device (100) by executing at least one command stored in the memory (130).

[0086] The processor (140) can be implemented in various ways. For example, the processor (140) can be implemented as at least one of an Application Specific Integrated Circuit (ASIC), an embedded processor, a microprocessor, hardware control logic, a hardware finite state machine (FSM), and a digital signal processor (DSP).

[0087] The processor (140) is a type of central processing unit that can control the operation of the entire import / export requirement verification device (100) by running control software loaded in memory (130). The processor (140) may include all types of devices capable of processing data. Here, 'processor' may refer to a data processing device embedded in hardware, having a physically structured circuit to perform a function expressed by code or instructions included in a program, for example.

[0088] FIG. 5 is a flowchart illustrating a method for verifying import and export requirements according to one embodiment of the present invention.

[0089] Referring to FIG. 5, the process for verifying the import and export requirements of the processor (140) is described.

[0090] In some examples, the method for verifying import / export requirements is a method for verifying import / export requirements through a RULE-based electronic document processing algorithm in which at least part of each step is performed by a processor (140) to process any one of the procedures of import / export clearance, verification of requirements, and exemption from requirements by electronically documenting information exchange with a customs server (310) through a RULE-based electronic document processing algorithm.

[0091] In step S10, when a request for verification of requirements for import / export products is received by the processor (140) through the customs server (310), the processor receives requirement application information data from the customs server (310) using the SOAP standard communication protocol, converts the received raw data into a predefined XML electronic document format, and generates an electronic document of requirement application information that inserts Collaboration Protocol Profile (CPP) information, Collaboration Protocol Agreement (CPA) information, and essential customs clearance information into the electronic document.

[0092] That is, when a request for verification of requirements for import / export products is received by the processor (140) through the customs server (310), the processor (140) can securely receive requirement application information data from the customs server (310) using the SOAP standard communication protocol. The processor (140) can parse the received raw data using an XML parser and convert the data according to a predefined XML electronic document format to generate an electronic document of requirement application information that includes Collaboration Protocol Profile (CPP) information, Collaboration Protocol Agreement (CPA) information, and essential customs clearance information such as applicant, product, and transaction.

[0093] In some examples, the processor (140) parses raw data received from the Customs Service server (310) using the SOAP standard communication protocol through an XML parser. For example, it generates an XML electronic document by mapping each field (applicant name, product name, transaction conditions, etc.) included in the received data to a corresponding tag according to a predefined XML schema. When generating the electronic document, the processor (140) can control the inclusion of standardized document supplementary information, such as a Collaboration Protocol Profile (CPP) and a Collaboration Protocol Agreement (CPA), so that each electronic document contains all essential information required for customs clearance operations. The electronic documents generated in this way can be shared in a consistent data format between servers, such as the Customs Service server (310), and the standardization of data is ensured.

[0094] FIG. 6 is an exemplary diagram illustrating the configuration of an electronic document according to an embodiment of the present invention. Referring to FIG. 6, the processor (140) first analyzes raw data received from the Customs Service server (310) via the SOAP standard communication protocol in order to generate an electronic document of application information for requirements. At this time, the processor (140) can interpret the message structure using an XML parser and control each field (e.g., applicant name, product name, transaction conditions, etc.) to be mapped to a tag corresponding to a predefined XML schema. For example, " <msgid> 001< / msgid> The XML tag " contains the identifier of the requirements application, " <prdtcd> A04< / prdtcd> The received raw data is reconstructed into XML format by placing product codes in tags.

[0095] In this process, the processor (140) refers to the RULE file to determine which field to place in which XML tag, and if necessary, can perform field-to-field conversion logic (DB2GOVXML or GOVXML2DB). For example, if the column name used in the DB within the Customs Service server (310) is "USER_TEL", in the XML electronic document, " <telno>< / telno>A rule file is defined to express it in the same form, and the processor (140) can automatically map field names according to this definition. At this time, if CPP (Cooperation Protocol Profile) or CPA (Cooperation Protocol Agreement) information is required, the information is also inserted as a separate tag within the XML document, so that the cooperation rules and procedures required for customs clearance are recorded together in the document.

[0096] XML schema (XSD) and a verification step utilizing the schema can also be performed simultaneously. The processor (140) checks the order, data format, and value range of each tag according to the rules specified in the XSD, and if a required tag is missing or the data is outside the allowed range, it detects this as an error, records it in a log, and requests correction. When the XML electronic document generated through this process is shared between servers, such as the Customs Service server (310), it can maintain a consistent data format, and can facilitate mapping between DB and XML through modules such as GOVXML2DB and DB2GOVXML shown in FIG. 6.

[0097] That is, the processor (140) structures the raw data transmitted via SOAP message through an XML parser and completes the electronic document by inserting supplementary information such as CPP / CPA according to the mapping rules defined in the rule file. Through this, the electronic document includes all items essential for customs clearance without omission, and the standardized format is maintained even when the document is exchanged between servers such as the Customs Service server (310), thereby greatly improving work efficiency and data reliability.

[0098] Meanwhile, FIGS. 7 and 8 are presented to specifically show, from different perspectives, how electronic documents are transmitted, received, and verified between the Customs Service server (310) and the requirement agency server (320).

[0099] FIG. 7 illustrates the overall structure of the Customs Service server (310) and the Requirement Agency server (320), including the GSG (electronic document server, 314) and the linkage server (323), and explains in a broad sense which modules each server uses to exchange electronic documents, convert them into a DB, or reverse-convert them into XML, etc. On the Customs Service server (310) side, the AP server (313) and the GSG (314) are positioned to generate, verify, and store electronic documents. When necessary, they are transmitted to the linkage server (323) on the Requirement Agency server (320) side in the form of SOAP or ebXML messages. The linkage server (323) then stores the received documents back into a DB or reverse-converts them and passes them on to the AP server (324) responsible for requirement approval tasks. As such, FIG. 7 explains the process through which each server handles electronic documents, the order in which SOAP messages are transmitted, and when DB mapping (DB2XML, XML2DB, etc.) takes place.

[0100] FIG. 8 focuses on explaining how the ebXML structure is combined with a SOAP message and how MIME parts are attached and transmitted within it. For example, it schematically illustrates that the components such as SOAP-ENV:Envelope, SOAP-ENV:Header, and Body are divided into MIME-Part formats, and that document integrity and authentication are guaranteed through GPKI-based digital signatures and encryption. In other words, when the Customs Service server (310) or the requirement agency server (320) transmits and receives electronic documents, they use the SOAP protocol to transmit the ebXML document, and specifically show what information is included in each MIME part and where and how CPP / CPA (Collaboration Protocol Profile and Collaboration Protocol Agreement) verification is performed.

[0101] That is, referring to FIG. 7, on the side of the Customs Service server (310), the AP server (313) can create application documents or notification information for requirements and transmit or receive electronic documents to the GSG (314) through an internal process such as batch processing. The GSG (314) is a core server that mediates the exchange of electronic documents between the Customs Service server and the requirements agency server. It is designed to verify the received documents using SOAP or ebXML protocols, store them in a DB, and then transmit them back to the linkage server (323) of the requirements agency server (320) when necessary. Meanwhile, on the side of the requirements agency server (320), the linkage server (323) receives the electronic documents, converts them into an interface DB and stores them, and is structured to link with the AP server (324), which internally performs requirements approval tasks, to return the results back to the Customs Service server.

[0102] Referring to FIG. 7, the Customs Service server (310) may further include or be referred to as an AP (Application) server (313) and a GSG (Electronic Document Server) (314). Additionally, the Required Agency server (320) may further include or be referred to as a linkage server (323) and an AP server (324).

[0103] The AP server (313) of the Customs Service server (310) prepares the application for requirements, transmits the electronic document to the GSG (314), receives the electronic document from the GSG (314), performs batch processing, stores the processing status, and stores notification information and application information. The GSG (314) receives the electronic document from the AP server (313) of the Customs Service server (310) and the linkage server (323) of the requirements agency server (320), verifies it, stores it in the DB, and transmits the electronic document to the linkage server (323) of the requirements agency server (320). The linkage server (323) of the requirements agency server (320) converts the business information of the requirements approval work into data, verifies the electronic document, transmits the verified electronic document to the GSG (314) of the Customs Service server (310), receives and verifies the electronic document from the GSG (314), converts the data, and stores the application information in the DB. The AP server (324) of the requirement agency server (320) links with the linkage server (323) and notifies the application.

[0104] Based on the configuration shown in FIG. 7, the processor (140) can precisely control the exchange and processing of electronic documents between the Customs Service server and the requirement agency server. The Customs Service server (310) includes an AP server (313) and a GSG (electronic document server) (314), and these components are managed integrally by the processor (140). The processor (140) can first control the process in which the AP server (313) of the Customs Service server creates a requirement application, generates the corresponding electronic document in a predefined XML format, and transmits it to the GSG (314). The AP server (313) can batch process the generated electronic document and systematically store processing status, notification information, and application information, thereby providing base data necessary for subsequent data verification and transmission operations. Meanwhile, the GSG (314) receives electronic documents from the AP server (313) of the Customs Service server and the linkage server (323) of the requirement agency server (320), verifies the accuracy of the format and content of the documents using the embedded XML schema and ontology specifications, stores them in a database, and then transmits them to the linkage server (323) of the requirement agency server (320) if necessary. The requirement agency server (320) is composed of the linkage server (323) and the AP server (324). The linkage server (323) can verify the received electronic documents once again through data conversion and precise verification, and transmit the results to the GSG (314) of the Customs Service server (310) to ensure the reliability of the stored data. The AP server (324) can perform the role of determining whether the electronic documents are suitable for customs clearance and requirement verification procedures by closely linking with the linkage server (323) to finally perform application notification. In this way, the processor (140) can monitor and control data in real time at all stages, from the creation, verification, storage, and transmission of electronic documents between the Customs Service server and the requirement agency server to the final application notification, thereby maximizing the efficiency and reliability of the entire system.

[0105] Referring to FIG. 8, the receiving unit (410) and the transmitting unit (420) of the electronic document distribution server specifically show the order in which XML is mapped to the DB, CPP / CPA information is verified, and SOAP messages or ebXML messages with a MIME part structure are exchanged. For example, the part "SOAP with Attachments MIME envelope" shows how the MIME format is used when transmitting an electronic document divided into several parts, and implies that the integrity and reliability of the electronic document are guaranteed through GPKI (certified public certificate)-based encryption and digital signature verification processes. In addition, the procedure in which the receiving unit (410) examines CPP / CPA information, SOAP Envelope, Body, etc., and converts XML into a DB, and the procedure in which the transmitting unit (420) processes encryption, digital signature, and notification information, etc., and transmits the electronic document to the requirement agency server (300), are explained in steps. At this time, the Customs Service server (310) transmits the requirement application or notification information via SOAP or ebXML, and the requirement agency server (300) examines the electronic document in stages S301 to S303 and then sends the result back to the Customs Service, thereby proceeding with the entire customs clearance flow.

[0106] In FIG. 8, the Customs Service server (310) processes the application for requirements regarding import / export products using the SOAP protocol (S101). Subsequently, the receiving unit (410) of the server (300) receives the electronic document of the received information and then precisely verifies the Collaboration Protocol Profile (CPP) and Collaboration Protocol Agreement (CPA) information included in the electronic document (S411). After verification, the received electronic document is decrypted and processing information and application information are stored (S412). Based on this information, the requirements agency server (320) accepts the application for requirements (S301), proceeds with the examination procedure, and finally approves it (S302, S303). Subsequently, the server (300) encrypts the approved electronic document and stores notification information (S422). Then, the transmitting unit (420) converts this notification information into an electronic document and transmits it (S421), and the Customs Service server (310) finally notifies the approval result, thereby completing the entire procedure (S102).

[0107] In summary, when combining FIGS. 7 and 8, FIGS. 7 presents the overall structure and flow of how AP servers, GSGs, and linkage servers deployed at the Customs Service server (310) and the requirement agency server (320), respectively, exchange electronic documents through DB conversion processes, while FIGS. 8 shows in detail what MIME part structures SOAP and ebXML messages actually have and how they are received and transmitted after undergoing encryption, digital signatures, and CPP / CPA information verification. In other words, FIGS. 7 focuses on the "big picture of electronic document exchange between servers and the DB conversion process," whereas FIGS. 8 specifically reveals the "internal structure of SOAP / ebXML messages and transmission and verification procedures." Through FIGS. 7 and 8, the entire system can systematically implement the order in which documents are created, verified, transmitted, and received, and at which stage CPP / CPA verification or DB mapping takes place.

[0108] In step S11, the processor (140) verifies the format, required items, data consistency, and accuracy of the generated electronic document of requirements application information according to a predefined schema (XSD) and related specifications, aggregates the requirements confirmation application data and related customs clearance history included in the electronic document of requirements application information, quantitatively extracts features from each data item, minimizes noise in the input data by applying at least one of statistical outlier removal, Z-score normalization, and Min-Max scaling, and preprocesses the input data into an optimal form that a machine learning model can learn through Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor Embedding (t-SNE) techniques.

[0109] In some examples, the processor (140) parses the electronic document of the requirements application information using an XML parser and automatically compares the existence of elements specified in the schema, the order of elements, the data format of each data item, and the maximum and minimum limits of the defined values ​​through an XSD verification engine, while precisely checking whether each element is designated as a required element, whether the elements are arranged in the defined order, and whether the data value exists within the range of values ​​specified in the schema, and can automatically detect errors such as omission, duplication, order mismatch, format error, or exceeding the value range.

[0110] Additionally, the processor (140) can calculate quantitative evaluation indicators for each data item, such as the frequency of error occurrence, the degree of data inconsistency, and the consistency of cross-reference relationships between elements, based on the XSD verification results, and compare these with a predefined threshold to determine whether the consistency and accuracy of the data within the electronic document of the requirements application information have been ensured.

[0111] Additionally, after the above-mentioned decision step, if a verification error or inconsistency is detected, the processor (140) can automatically generate a verification result log including details such as the error code of the error, the time of occurrence, a specific description of the error, and the location of the data item where the error occurred.

[0112] In some examples, the processor (140) may apply a predefined transformation function to the features extracted from the aggregated data items to quantify them, perform statistical analysis on each quantified feature value to identify extreme values ​​of the data distribution, and apply a predefined outlier removal algorithm to remove abnormal values.

[0113] Additionally, the processor (140) can apply a Z-score normalization formula to each feature value to convert it into a normal distribution with a mean of 0 and a standard deviation of 1, or apply a Min-Max scaling technique to linearly transform each feature value and standardize it by adjusting it to a fixed range of 0 and 1.

[0114] Additionally, the processor (140) can apply a principal component analysis (PCA) algorithm to a standardized multidimensional data matrix to calculate the variance contribution of each feature, select principal components that exceed a pre-set variance contribution criterion to reduce the data dimension, or apply a t-distributed Stochastic Neighbor Embedding (t-SNE) algorithm to generate low-dimensional embeddings, remove unnecessary noise from the input data, and convert it into an optimal input data form that preserves only the core features.

[0115] That is, the processor (140) can parse the generated electronic document once again using an XML parser and compare and analyze in detail whether each element is arranged in the appropriate order and format according to a predefined XML schema (XSD) and whether the data value is outside the allowed range. For example, the XSD rule may specify that the 'amount' item must be entered as real number data and is valid only between the minimum and maximum values, and the schema may set a constraint that the 'processing result' item must be expressed only as 0 or 1. Based on these rules, the processor (140) checks all tags of the electronic document and automatically detects omissions, duplicates, order discrepancies, format errors, value range exceedances, etc., and supports subsequent corrective measures by recording detailed information such as the location and time of the error and the type of error in a log format.

[0116] For the verified electronic document, the processor (140) performs a preprocessing process by combining it with customs clearance history data. At this time, key items such as 'time', 'amount', and 'processing result' are converted into numerical values ​​through a predefined conversion function. The 'time' item is unified into a timestamp expressed in the ISO 8601 standard, the 'amount' item is converted into a floating-point value in the corresponding currency unit, and the 'processing result' item is coded as 0 or 1 to be changed into a form that is easy for a machine learning model to process. The processor (140) statistically analyzes these quantified features to identify extreme values ​​(outliers) and reduces noise by removing abnormal data through a predefined outlier removal algorithm. Subsequently, each feature value is standardized by applying Z-score normalization or Min-Max scaling techniques. For example, Z-score normalization sets the mean to 0 and the standard deviation to 1 to make the data distribution uniform, and Min-Max scaling sets the minimum value to 0 and the maximum value to 1 so that all values ​​can be compared within the same range. Through this process, the processor (140) provides an optimal form that allows the machine learning model to learn the data without bias.

[0117] The processor (140) applies a PCA (Principal Component Analysis) algorithm or a t-SNE (t-distributed Stochastic Neighbor Embedding) algorithm to remove unnecessary features from high-dimensional data and efficiently represent only key features in a low-dimensional space. In the case of PCA, only the principal components are selected by calculating how much each feature contributes to the total variance, and t-SNE creates embeddings that are better distinguished visually and in terms of model training by reflecting the similarity between data points based on probability distributions. The processor (140) supports deep learning or reinforcement learning models in achieving higher accuracy and efficiency by finally generating an input matrix in which noise between data is removed and only important features are preserved in the reduced dimension.

[0118] In step S12, the processor (140) trains a deep learning or reinforcement learning model using the preprocessed data to calculate an optimal RULE mapping standard that accurately reflects changes in the customs environment and data patterns by comparing it with a statically set RULE mapping standard, and updates a pre-set RULE mapping table in real time based on the calculated learning result.

[0119] In some examples, the processor (140) can take the preprocessed data as input, build a multilayer perceptron-based deep learning network using a mean squared error loss function, calculate the gradient of the loss function for the weights and biases of the current model for each learning iteration using an error backpropagation algorithm, and repeat the learning process of updating model parameters (weights and biases) with the learning rate set constant using stochastic gradient descent, and perform learning until a predetermined number of epochs is completed or the change in the loss value is negligible and the convergence condition is satisfied.

[0120] Additionally, after the learning is completed, the processor (140) calculates a prediction result for the input data using the learned model and compares and analyzes the prediction result with the result according to a pre-set static RULE mapping standard, wherein the comparison is performed by quantitatively evaluating the degree of agreement between the actual customs clearance history and the prediction result using accuracy and F1 score, and through an optimization algorithm that determines the optimal RULE mapping standard based on a pre-set threshold value calculated as an evaluation indicator, the optimal RULE mapping standard that precisely reflects changes in the customs clearance environment and data patterns can be calculated.

[0121] In some examples, the processor (140) can numerically evaluate the need to update each item by directly comparing the optimal RULE mapping criteria calculated by the learned model with each item of an existing static RULE mapping table.

[0122] Additionally, the processor (140) can replace the corresponding item in the existing RULE mapping table by applying the calculated optimal RULE mapping criteria to the item requiring update according to the evaluation result, and generate an update log including the change history of the replaced item, the update time, and the applied RULE mapping criteria.

[0123] In addition, the processor (140) can automatically verify the consistency between the generated update log and the updated RULE mapping table, and then monitor the update results in real time to immediately reflect them in the customs data processing system.

[0124] That is, the processor (140) receives data prepared through a preprocessing process, constructs a multi-layer perceptron-based deep learning network or reinforcement learning algorithm, and performs model training through backpropagation. At this time, the processor (140) applies a stochastic gradient descent algorithm to calculate a gradient to minimize a loss function (e.g., mean squared error (MSE) or cross-entropy, etc.) for each training iteration, and updates the model's internal parameters (weights and biases) according to the calculated gradient. This update is performed based on a pre-set learning rate, and the processor (140) can continue training until the number of epochs is repeated a certain number of times or until the loss value converges to below a certain threshold. When training is complete, the processor (140) derives a predicted value for the preprocessed input data using the trained model and compares it with the result calculated based on the existing static RULE mapping criteria.

[0125] Specifically, the processor (140) takes actual customs clearance history data as ground truth and calculates quantitative evaluation indicators such as accuracy and F1 score to determine how closely it matches the prediction results produced by the trained model, and precisely evaluates the performance difference between the two criteria (dynamic RULE criteria based on the trained model and static RULE criteria) based on these indicators. If the trained model shows high performance by more accurately reflecting changes in the customs clearance environment and data patterns, the processor (140) derives the optimal RULE mapping criteria based on the results and updates the existing RULE mapping table in real time to control the entire system to automatically adapt to the changed conditions. Through this, the processor (140) can dynamically respond even when new customs clearance situations or exceptional data patterns occur, and can continuously adjust the customs clearance procedure in a way that minimizes errors and increases efficiency.

[0126] In step S13, the processor (140) applies the updated RULE mapping criteria to the electronic document of the requirement application information to automatically extract, compare, and analyze the format, consistency, and alignment of CPP / CPA information and other customs-related data included in the electronic document of the requirement application information, and verifies whether the electronic document of the requirement application information meets the predefined dynamic criteria based on the comparison and analysis results.

[0127] That is, the processor (140) applies the updated RULE mapping criteria to the electronic document of the requirement application information, automatically extracts whether the electronic document correctly contains standardized customs clearance-related data such as Collaboration Protocol Profile (CPP) and Collaboration Protocol Agreement (CPA) information, and can precisely compare and analyze based on predefined dynamic criteria (dynamic RULE mapping criteria and ontology specifications). At this time, the processor (140) uses an XML parser or RDF / OWL-based semantic web technology to search for each tag or attribute within the electronic document and checks the format, data type, cross-reference relationship, etc. required by the dynamic RULE mapping criteria. For example, " <cpp>If there is a rule that the collaboration protocol profile listed on the tag must match the corresponding protocol agreement (CPA), the processor (140) refers to the mutual association rule specified in the ontology specification to convert the relationship between the CPP and CPA tags into a quantitative indicator (e.g., a match rate) and evaluates the consistency of the electronic document by calculating the mismatch rate or missing items.

[0128] Based on the results, the processor (140) quantifies the degree of consistency and the probability of error occurrence for each data item, and then determines whether the electronic document meets dynamic criteria by comparing it with a preset threshold. If the evaluation is below the threshold and an anomaly or discrepancy is detected, the processor (140) immediately generates a verification result log including an error code, time of occurrence, detailed description, and the specific location of the data item where the error occurred. This log can be transmitted to the customs officer or system administrator by issuing an automatic warning signal, and simultaneously, additional verification procedures (e.g., execution of a re-analysis module, request for manual review, etc.) can be initiated to control the problem and resolve it quickly. In this process, the processor (140) permanently stores the error history through a log DB or blockchain record module, and can use it as a reference when similar errors occur in the future, thereby further enhancing the reliability and accuracy of the overall customs clearance procedure.

[0129] In step S14, the processor (140) forms all relevant processing data, including the dynamic RULE update history and electronic document processing log, into a single data block based on the verification results, then ensures the integrity of the data by assigning a digital signature and a cryptographic hash value, and permanently stores the block in a distributed ledger that cannot be tampered with by applying a secure consensus algorithm and recording it on a blockchain network.

[0130] In some examples, the processor (140) can serialize the entire contents of the processing data in bytes and apply a defined cryptographic hash function (SHA-256) to produce a fixed-length hash value.

[0131] Additionally, the processor (140) can combine the calculated hash value and the serialized processing data, generate a digital signature according to the RSA algorithm using a previously registered private key, and attach the digital signature and hash value to the corresponding data block to ensure the integrity of the data.

[0132] Additionally, the processor (140) executes a secure consensus algorithm (applying a Proof-of-Work algorithm) on the data block to calculate a Proof-of-Work value that satisfies the difficulty goal based on the cryptographic hash value of the data block, transmits the data block containing the calculated Proof-of-Work value to the network so that all participating nodes on the blockchain network reach a consensus, and through the consensus process, it can be confirmed that the data block is permanently recorded in the distributed ledger without tampering.

[0133] That is, the processor (140) first serializes each data in byte units to integrate all collected related data, such as verification results, dynamic RULE update history, and electronic document processing logs, into a single continuous data stream. In this serialization process, a fixed encoding rule is applied so that the structure and order of the raw data are maintained, thereby allowing the data to be converted into a continuous array of bytes without losing its original form and content. Subsequently, the processor (140) applies a SHA-256 cryptographic hash function to the serialized data to produce a fixed-length hash value. The SHA-256 algorithm is a standard encryption technique that generates a hash capable of guaranteeing the integrity of the input data, and the generated hash value has the characteristic of producing a completely different value if the data is modified even slightly.

[0134] The processor (140) combines the calculated hash value and the serialized data to form a combined data structure. For this combined data, a digital signature is generated according to the RSA algorithm using a previously registered private key. The RSA algorithm is a representative method of public-key cryptography, and this digital signature serves to guarantee the origin and integrity of the data, and makes it easy to detect whether the data block with the signature attached has been altered or forged externally.

[0135] Subsequently, the processor (140) applies Proof-of-Work (PoW) as a secure consensus algorithm. In this process, a Proof-of-Work value that satisfies a pre-set difficulty target is calculated based on the cryptographic hash value of the data block. The calculation of the Proof-of-Work value is performed through iterative hash calculations, and is carried out by trying various nonce values ​​until the target difficulty is satisfied. The data block containing the calculated Proof-of-Work value is then transmitted by the processor (140) to the blockchain network so that all participating nodes on the network reach a consensus on the validity of the block. Once the consensus process is complete, the data block is permanently recorded in the distributed ledger without tampering, thereby ensuring that all related information, such as electronic document processing results, update history, and verification logs, can be reliably stored.

[0136] In step S15, the processor (140) automatically receives the electronic document of the requirement application information and the processing history recorded in the blockchain network and performs a final customs requirement verification procedure that reflects the result of verifying the requirements necessary for the import / export customs clearance procedure.

[0137] That is, the processor (140) performs the function of periodically monitoring and automatically receiving electronic documents and processing history recorded in the blockchain network. At this time, the processor (140) utilizes an algorithm (e.g., an event-based receiving module) that detects newly recorded blocks in the distributed ledger to retrieve electronic document data and related processing logs contained in each block in real time. The received data is first verified for integrity and forgery through cryptographic hash values ​​and digital signatures, and this process is carried out by applying the RSA algorithm and the SHA-256 hash function. Subsequently, the processor (140) compares and analyzes the received electronic documents and processing history with an internal database to finally evaluate the format, content, consistency, and integrity of the data within the electronic documents according to predefined verification criteria (dynamic RULE mapping criteria and ontology specifications, etc.). In this evaluation process, the processor (140) can derive a verification result by analyzing the cross-reference relationships between data elements and calculating whether each element conforms to the specified specifications using quantitative indicators (e.g., matching rate, error rate). The derived final verification result is immediately transmitted to the Customs Service server and the relevant customs clearance system, thereby supporting the smooth and error-free progress of actual import and export customs clearance procedures. Through this, the processor (140) can significantly improve the reliability and efficiency of the entire customs clearance system by ensuring that all electronic document processing history recorded on the blockchain is safely stored, while also rapidly reflecting the latest verification information during the final customs clearance requirement verification process.

[0138] In step S16, the processor (140), after the final customs clearance requirement verification step, automatically initiates and performs an import / export declaration examination procedure for import / export products for which the requirement examination has been completed through the customs server, based on relevant electronic documents and data including requirement application information and examination history.

[0139] And the processor (140) can generate an electronic document of the examination results including the examination results, related processing history, and examination details by analyzing the received export / import declaration examination result data after receiving the results from the customs server (310).

[0140] Additionally, the processor (140) can verify the consistency between the requirements application information and the final customs clearance result for the import / export products by executing an algorithm that automatically maps the requirements application information electronic document and the import / export declaration review result electronic document based on the electronic document of the review result.

[0141] That is, after the final customs clearance requirement verification procedure is completed, the processor (140) automatically receives relevant electronic documents and data, including requirement application information and examination history for import / export products for which the requirements examination has been completed, through the Customs Service server. In this process, the processor (140) detects the electronic documents transmitted from the Customs Service server (310) in real time via a network interface and verifies the integrity of the received data using a cryptographic verification algorithm. Subsequently, based on the received electronic documents and related data, the processor (140) automatically initiates the import / export declaration examination procedure. At this time, the processor (140) uses a predefined algorithm to extract information included in each item within the electronic document (e.g., applicant qualifications, product details, transaction conditions, and customs clearance history), refines the data, and performs a process of comparing and analyzing it with past examination results stored in an internal database.

[0142] For this comparative analysis, the processor (140) can quantitatively calculate the degree of agreement and discrepancy between the actual customs clearance history and the information recorded in the electronic document by utilizing statistical and machine learning-based algorithms, and in this process, the reliability of the examination results is determined using evaluation indicators such as accuracy and F1 score. The export / import declaration examination result data received from the Customs Service server (310) is parsed by the processor (140) through a high-precision parsing module and compared with the prior examination history stored in the database to identify errors or omissions that occurred at each examination stage. Based on this, the processor (140) generates an examination result electronic document that includes the examination results, related processing history, and examination details in detail, and the format and consistency of the data in this electronic document are ensured through XML format and XSD schema verification.

[0143] The generated electronic document of the examination result is combined with the original electronic document of the requirement application information by an automatic mapping algorithm. This mapping algorithm is designed to precisely compare the two documents based on common key fields or data tags within the electronic documents and to verify mutual consistency between the requirement application information for import / export products and the final customs clearance result. Through this automatic mapping process, the processor (140) can immediately generate an error notification and initiate a supplementary verification procedure if a discrepancy or error is found, thereby ultimately guaranteeing the accuracy and reliability of the entire customs clearance procedure.

[0144] Meanwhile, the processor (140) can automatically recommend the optimal input value based on a predefined frequency setting for the word or number entered by the user when receiving an application for verification of requirements for import / export products through the Customs Service server (310) for inputting items in an electronic document.

[0145] In some examples, the processor (140) can quantitatively calculate the usage pattern of each item by aggregating the total usage frequency, the input frequency by a specific user, and the quarterly input frequency of word and number data recorded in the electronic document of the requirement verification application information received from the Customs Service server (310).

[0146] Additionally, the processor (140) can calculate the weight of each item by applying a predefined algorithm based on the aggregated frequency data, and determine the priority of each item by considering the ratio of the total frequency, the frequency per user, and the input frequency per quarter.

[0147] Additionally, when the user inputs some characters into an item of an electronic document, the processor (140) sorts words and number items similar to the input characters in real time based on the input characters and the calculated weight values, and displays a high-priority recommendation list in the autocomplete window. At this time, the recommendation list may be provided in a differentiated manner according to the user-specific input frequency and quarterly input pattern of the data recorded in the electronic document of the requirement verification application information received from the Customs Service server (310).

[0148] That is, the processor (140) can perform the function of automatically recommending optimal input values ​​for items entered into the user interface based on word and number data recorded in the electronic document of the requirement verification application information received from the Customs Service server (310). In this process, the processor (140) first precisely calculates the total usage frequency, the input frequency by a specific user, and the quarterly input frequency of all word and number data included in the electronic document through a total data aggregation algorithm, thereby quantitatively calculating the usage pattern of each item. The calculated frequency values ​​are converted into weights representing the relative importance of each item by a subsequent algorithm, and these weights are calculated according to a strict mathematical formula based on the ratio of the total frequency, the frequency by user, and the input frequency by quarter. Then, when a user enters some characters into a specific item of the electronic document, the processor (140) executes an algorithm that compares the input string in real time with the pre-calculated weight values ​​of each item, and sorts the word and number items that show the highest degree of agreement with the input string through similarity measurement (e.g., using cosine similarity or edit distance). These sorted results are arranged in order from highest priority to lowest and displayed in the autocomplete window, allowing the user to shorten input time and reduce errors by selecting them. Additionally, the processor (140) can provide more precisely personalized recommendation results by utilizing predefined statistical and machine learning models in the process of generating the recommendation list to comprehensively consider the user's past input patterns and current input situation. As a result, the recommendation list appears as a differentiated result that goes beyond simple string matching and reflects the user's input habits and the results of the overall data aggregation. Furthermore, since the user-specific input frequency and quarterly input patterns of the data recorded in the electronic documents received from the Customs Service server (310) are strictly analyzed and provided, the accuracy and consistency of the data are guaranteed when creating electronic documents, and the efficiency and reliability of the entire customs clearance process can be increased.

[0149] FIG. 9a is a table showing the procedure in the import / export customs clearance process according to one embodiment of the present invention, and FIG. 9b is an example diagram showing the procedure for applying for either the verification of requirements or the exemption from requirements according to one embodiment of the present invention. With reference to FIG. 9a and FIG. 9b, the process according to the import / export customs clearance procedure of the processor (140) will be described.

[0150] Based on FIG. 9a, the processor (140) can control the various steps occurring in the import / export customs clearance and requirements verification procedures in sequence. FIG. 9a clearly indicates which system plays which role in each step by arranging the business procedure and applied technology on the horizontal axis and the Customs Service’s civil complaint server (e.g., UNIPASS), customs administration server (system), electronic document distribution server (system, RULE-based linkage server), requirements verification server (system), etc., on the vertical axis. The procedure shown in FIG. 9a can generally be summarized as follows.

[0151] First, in Step 1 (requirement verification), when a user applies for requirement verification through the civil complaint server (311) of the Customs Service server (310), the processor (140) can digitize the document through the electronic document distribution server (321) and transmit it to the requirement verification server (322) using SOAP or HTTP communication protocols. At this time, the electronic document in XML or JSON format is configured to include CPP / CPA information or other customs clearance information, and the processor (140) can apply encryption / decryption and electronic signature procedures if necessary.

[0152] Next, in step 2 (import / export declaration), the user completes the import / export declaration form, and the processor (140) receives it and generates an electronic document in the form of XML or ebXML, which can then be transmitted to the customs administration server (312). During this process, the user input auto-completion function operates, and the processor (140) recommends input values ​​based on existing frequency or quarterly patterns, thereby reducing errors and increasing convenience.

[0153] Subsequently, when the customs administration server (312) examines the import / export declaration in the third stage (export / import declaration examination), the processor (140) receives the result to the electronic document distribution server (321), converts the examination result into an electronic document, and transmits it to the customs office server (310) and the requirements verification server (322), thereby supporting the re-verification of the examination result or the supplementation of additional documents. If necessary, the processor (140) can quantitatively analyze the data within the electronic document using an XML parser or semantic web technology, and apply dynamic RULE mapping criteria to verify the format, content, and consistency of the electronic document.

[0154] Finally, in step 4 (approval of import / export declaration), when the customs administration server (312) decides on final approval or rejection, the processor (140) can digitize this result by linking it with the import / export declaration form and transmit the information to the civil complaint server (311) and the requirement verification server (322) so that the final customs clearance procedure can be carried out. With this, the import / export procedure is completed, and if necessary, the processor (140) can record the electronic document and processing log on the blockchain to ensure data integrity.

[0155] As a result, the processor (140) can generate, verify, and transmit electronic documents at each step presented in the image, and by applying various technologies such as XML parser, encryption and decryption, auto-completion, RULE mapping, and blockchain recording, it can support the entire import / export customs clearance and requirements verification process to proceed efficiently and reliably.

[0156] Referring to FIG. 9b, the processor (140) can sequentially control a series of processes that take place between the user (applicant) and the person in charge of the requirements agency during the requirements verification procedure. In FIG. 9b, the box at the top left indicates the process in which the applicant accesses the civil complaint system (601), logs in (602), and then applies for either an exemption from requirements or a requirements verification (603). Afterwards, the work continues in the order in which the applicant checks the processing status (604), verifies (605) if there is a recommendation letter issued by the person in charge of the requirements agency, and then proceeds with import and export declarations (606).

[0157] At the same time, the bottom box of FIG. 9b shows a flow in which a person in charge of the requirements agency accesses the system (701), views the application that has already been submitted (702), checks the security requirements if necessary, and takes measures such as requesting supplementation, accepting, rejecting, or approving (703, 704). Finally, at the stage of issuing a recommendation letter (705), the processor (140) can generate the result, such as supplementation or approval, in the form of an electronic document and transmit it back to the civil complaint system or customs administration system.

[0158] The processor (140) automatically generates and verifies electronic documents (e.g., XML, JSON, etc.) at each of these steps and can check the format, consistency, and integrity of the information entered by the applicant according to RULE mapping criteria or XML schemas (XSD). For example, when an applicant requests a requirement verification, the processor (140) parses SOAP communication messages between the civil complaint system and the requirement agency server, extracts necessary fields, and checks whether Collaboration Protocol Profile (CPP) or Collaboration Protocol Agreement (CPA) information is missing. In steps requiring security, the processor (140) applies encryption / decryption algorithms to securely transmit the application and supplementary materials, and when the requirement agency official decides on approval or rejection, the result is generated as an electronic document and reflected in the civil complaint system, thereby allowing the entire requirement verification process to proceed automatically and efficiently.

[0159] Meanwhile, furthermore, the processor (140) can implement an embodiment that calculates the risk level of import / export goods in real time by integrating additional artificial intelligence-based predictive analysis functions. In this case, the processor (140) first collects detailed data included in past customs clearance history, product classification information, and application documents, and then performs learning by using this as input to a neural network model such as a multilayer perceptron. For example, data regarding cases where the declared price in past transactions significantly deviates from the average range, or data regarding goods where past illegal transaction patterns have repeatedly appeared, are used as training data so that the trained model can distinguish between normal transaction patterns and outlier patterns. At the same time, the processor (140) monitors the statistical deviation of the input data in real time by applying an outlier detection algorithm, such as variance-based statistical analysis or robust scaling techniques, and can detect cases where the declared price of a specific import / export item is significantly higher or lower compared to past transactions. When the model detects a risk signal in this way, the processor (140) immediately generates a warning and controls the automatic initiation of an additional verification procedure. In this additional verification step, the processor (140) can refer to the processing history recorded on the blockchain network to precisely compare and analyze the processing paths and results of past similar cases, and based on this, calculate a risk score and deliver an immediate notification to the customs officer.

[0160] Additionally, the processor (140) can implement encryption and digital signature procedures using quantum-resistant cryptographic techniques to ensure safe operation even in a quantum computing environment during the process of supplementing or transmitting electronic documents of application information. In this case, the processor (140) encrypts data using a predefined quantum-resistant cryptographic algorithm (e.g., lattice-based cryptographic technique) and supplements the security provided by traditional RSA-based digital signatures. Furthermore, for sensitive information, the processor (140) can utilize zero-knowledge proof techniques to support the verification of the validity of the information without directly exposing the applicant's personal information or other sensitive data. This procedure ensures that the customs authority can trust the necessary information while adhering to the principle of least disclosure.

[0161] Additionally, the processor (140) adopts a container-based microservices architecture, allowing each functional module, such as requirements verification, electronic document conversion, blockchain recording, and artificial intelligence verification, to be independently deployed and operated. This architecture enables the entire system to operate stably even with sudden increases in traffic or changes in throughput by automatically scaling out (horizontally expanding) or scaling in (horizontally contracting) only the module when the load is concentrated on that specific module. In this way, by integrating various algorithms, security technologies, and a scalable distributed processing architecture, the processor (140) can predict risk factors in import and export customs clearance procedures in advance and support efficient customs clearance processing while maintaining data integrity and security.

[0162] Meanwhile, the RULE-based electronic document processing algorithm according to the present invention can be implemented based on a cloud-native architecture. The system can be designed to perform the entire pipeline, including preprocessing, verification, quantification, dimensionality reduction, RULE mapping, and blockchain recording of electronic documents, in a multi-distributed processing environment. Here, a multi-distributed processing environment may refer to a cloud infrastructure configured such that multiple virtualization resources are logically bundled to operate as one large cluster.

[0163] The aforementioned cluster can be configured by utilizing large-scale data centers such as public cloud services like AWS, Azure, and GCP, or through a private cloud installed within the enterprise. Kubernetes, a widely adopted container orchestration platform, deploys each module as a Pod, automatically manages the Pod lifecycle, and can automatically increase the number of instances when the load increases through Horizontal Pod Autoscaling. Since this orchestration platform supports desired-state management, operators simply need to describe the target state in a YAML manifest, and the cluster will attempt to maintain that state on its own.

[0164] In some examples, the electronic document preprocessing module can load hundreds of gigabytes or more of JSON, XML, or PDF data into Apache Spark-based distributed dataframes and perform schema checking and XSD-based validation in parallel. During this process, feature extraction may include tokenization by natural language processing models, naming entity recognition, and keyword frequency analysis, while outlier removal can exclude statistical anomalies through Inter-Quartile Range (IQR) or DBSCAN clustering. In the normalization process, the input distribution can be compressed to a specific range by selectively applying Z-score normalization or Min-Max scaling. In the dimensionality reduction step, PCA can selectively remove principal components with high variance, or t-SNE can project high-dimensional non-linear patterns onto lower dimensions using a neighbor-preserving method.

[0165] In some examples, machine learning pipelines that generate rule mapping criteria can be configured based on AWS SageMaker Pipelines or Vertex AI Pipelines. Within the pipeline, a Feature Store stores preprocessed features by version, and during the training phase, a deep learning model such as a Transformer Encoder or a reinforcement learning algorithm such as Proximal Policy Optimization (PPO) can generate rule candidates from the input data. The trained model is registered in the Model Registry and can then be applied to only a portion of real-time traffic via Canary Deployment to verify performance. In this process, AutoML Ops schedules retraining cycles based on pipeline status, training logs, and metrics, and can automatically retrain the model using the latest data if performance degradation is detected.

[0166] In some examples, the blockchain record module can be implemented using Hyperledger Fabric, a permissioned blockchain framework, as a managed service. Since Hyperledger Fabric provides a pluggable consensus structure and channel-based data separation capabilities, sensitive electronic document metadata can be isolated and stored by channel. The block header includes a timestamp, the previous block hash, and the Merkle Root, while the payload may include the electronic document hash, a RULE update JSON patch, and a validation log hash. The Practical Byzantine Fault Tolerance (PBFT) consensus algorithm can guarantee data integrity by requiring that at least two-thirds of the network's participating nodes agree on the same result for a block to be finalized.

[0167] In some examples, the entire system may adopt a Zero Trust architecture for security. Zero Trust treats all traffic, including internal networks, as a potential threat and can perform identity verification and granular authorization for every request. Specifically, an Identity-Aware Proxy can issue JWT tokens based on OAuth 2.0 and OpenID Connect, and the service mesh Istio can encrypt inter-service communication using mTLS. Data at rest can be encrypted using the AES-256-GCM algorithm with a Customer Managed Key (CMK) that is periodically rotated from a Key Management Service (KMS).

[0168] In some examples, systems can employ a multi-region active-active architecture for operations and disaster recovery. A Global Load Balancer can connect users to the optimal region by performing geographic round-robin and health checks. Litmus, a chaos engineering tool, can measure the system's Mean-Time-To-Recovery (MTTR) by periodically injecting scenarios such as Pod outages, network partitions, and node failures. Infrastructure is defined by Terraform scripts and managed as a code state, while GitOps workflows can automatically detect and recover from state deviations.

[0169] Compared to traditional on-premises infrastructure, this configuration can handle large-scale traffic without service interruption and optimize costs through a usage-based billing model. Furthermore, reinforcement learning-based rule updates reflecting real-time logs and changes in electronic document structure can shorten the policy establishment cycle from days to hours, and the history of electronic document processing can be permanently and transparently preserved through a blockchain-based distributed ledger. Multi-region deployment and automatic failover guarantee availability of over 99.99%, and the Zero Trust security model can enhance data confidentiality and integrity while complying with international regulations. Therefore, the cloud integration implementation of the present invention simultaneously resolves issues of performance limitations, security vulnerabilities, and operational costs, enabling uninterrupted and highly reliable electronic document processing even in a rapidly changing international trade environment.

[0170] The embodiments according to the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program may be recorded on a computer-readable medium. In this case, the medium may include a magnetic medium such as a hard disk, a floppy disk, and a magnetic tape, an optical recording medium such as a CD-ROM and a DVD, a magneto-optical medium such as a floptical disk, and a hardware device specifically configured to store and execute program instructions, such as a ROM, RAM, or flash memory.

[0171] Meanwhile, the above-mentioned computer program may be one specifically designed and configured for the present invention, or one known and available to those skilled in the art of computer software. Examples of computer programs may include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0172] In the specification of the present invention (particularly in the claims), the use of the term "above" and similar descriptive terms may be in both singular and plural. Furthermore, where a range is described in the present invention, it is to include an invention to which individual values ​​belonging to said range are applied (unless otherwise stated), and this is equivalent to describing each individual value constituting said range in the detailed description of the invention.

[0173] Unless explicitly stated or contrary to the order of the steps constituting the method according to the present invention, said steps may be performed in a suitable order. The present invention is not necessarily limited by the order in which said steps are described. The use of all examples or exemplary terms (e.g., etc.) in the present invention is merely for the purpose of describing the present invention in detail, and the scope of the present invention is not limited by said examples or exemplary terms unless limited by the claims. Furthermore, those skilled in the art will understand that various modifications, combinations, and changes may be made according to design conditions and factors within the scope of the claims or equivalents to which they are added.

[0174] Accordingly, the scope of the present invention should not be limited to the embodiments described above, and all scopes equivalent to or equivalently modified from the claims set forth below, as well as the claims set forth below, shall be considered to fall within the scope of the concept of the present invention. Explanation of the symbols

[0178] 1: Import / Export Requirements Verification System 100: Export / Import Requirement Verification Device 110: Communication interface 120: User Interface 130: Memory 140: Processor 200: User terminal 300: Server 400: Network< / cpp>

Claims

Claim 1 delete Claim 2 A method for verifying import / export requirements through a rule-based electronic document processing algorithm, wherein at least a portion of each step is performed by a processor to process any one of the procedures of import / export customs clearance, verification of requirements, and exemption from requirements by digitizing information exchange with a Customs Service server through a rule-based electronic document processing algorithm, wherein when an application for verification of requirements for import / export products is received through the Customs Service server, requirement application information data is received from the Customs Service server using a SOAP standard communication protocol, and the received raw data is converted into a predefined XML electronic document format to generate a requirement application information electronic document by inserting Collaboration Protocol Profile (CPP) information, Collaboration Protocol Agreement (CPA) information, and essential customs clearance information into the electronic document; a step of verifying the format, essential items, data consistency, and accuracy of the generated requirement application information electronic document according to a predefined schema (XSD) and related standards, aggregating the requirement verification application data and related customs clearance history included in the requirement application information electronic document, quantitatively extracting features from each data item, and a statistical outlier removal algorithm A preprocessing step for verifying electronic documents of requirements application information, which removes extreme values ​​of the data distribution identified by statistical analysis, standardizes the data by adjusting it to a fixed range of 0 and 1 through Z-score normalization or Min-Max scaling, and preprocesses the data into a form suitable for machine learning models to learn by preserving only key features through Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor Embedding (t-SNE) techniques;A dynamic RULE mapping standard calculation and update step, wherein a deep learning or reinforcement learning model is trained using the above-mentioned preprocessed data to calculate and reflect changes in the customs clearance environment and data patterns by comparing them with pre-set static RULE mapping standards using the gradient of the mean squared error loss function via a backpropagation algorithm, and a prediction result for the input data is calculated using the trained model; a RULE mapping standard is calculated by quantitatively evaluating the prediction result and the result based on the pre-set static RULE mapping standards to determine the calculated value based on a pre-set threshold, and a pre-set RULE mapping table is updated in real time based on the calculated training result; an electronic document verification step, wherein the updated RULE mapping standard is applied to the electronic document of the requirement application information to automatically extract, compare, and analyze the format, consistency, and alignment of CPP / CPA information and other customs clearance-related data included in the electronic document of the requirement application information, and verify whether the electronic document of the requirement application information conforms to pre-defined dynamic standards based on the comparison and analysis results; and, as a result of the verification, all related processing data, including dynamic RULE update history and electronic document processing logs, are combined into a single data A blockchain recording step, wherein, after forming a block, data integrity is ensured by assigning a digital signature according to the RSA algorithm and a SHA-256 cryptographic hash value using a pre-registered private key, and the block is permanently stored in an immutable distributed ledger by recording it on a blockchain network by applying a Proof-of-Work consensus algorithm;The method includes a final customs clearance requirement verification step that automatically receives the electronic document of the requirement application information and processing history recorded on the blockchain network and performs a final customs clearance requirement verification procedure that reflects the results of the requirement verification required for the import / export customs clearance procedure; the preprocessing step for verifying the electronic document of the requirement application information includes: parsing the electronic document of the requirement application information using an XML parser; automatically comparing the existence of elements specified in the schema, the order of elements, the data format of each data item, and the maximum and minimum limits of defined values ​​through an XSD verification engine, verifying whether each element is designated as a required element, whether the elements are arranged in the defined order, and whether the data value exists within the range of values ​​specified in the schema; and automatically detecting errors such as omission, duplication, order mismatch, format error, or exceeding the value range; calculating quantitative evaluation indicators for each data item, such as the frequency of error occurrence, the degree of data inconsistency, and the consistency of cross-reference relationships between elements, based on the XSD verification results, and comparing these with a predefined threshold to determine whether the consistency and accuracy of the data within the electronic document of the requirement application information have been secured; and after the determining step, if verification errors or inconsistencies A step of automatically generating a verification result log including detailed information such as the error code of the error, the time of occurrence, a specific description of the error, and the location of the data item where the error occurred, when detected; a step of applying a predefined transformation function to features extracted from the aggregated data item to quantify them, performing statistical analysis on each quantified feature value to identify extreme values ​​of the data distribution, and applying a predefined outlier removal algorithm to remove abnormal values;A step of applying a Z-score normalization formula to each feature value to convert it into a normal distribution with a mean of 0 and a standard deviation of 1, or applying a Min-Max scaling technique to linearly transform each feature value and standardize it by adjusting it to a fixed range of 0 and 1; The method includes the step of applying a Principal Component Analysis (PCA) algorithm to a standardized multidimensional data matrix to calculate the variance contribution of each feature, selecting principal components that exceed a pre-set variance contribution criterion to reduce the data dimensionality, or applying a t-distributed Stochastic Neighbor Embedding (t-SNE) algorithm to generate low-dimensional embeddings, and converting the input data into a form that removes unnecessary noise from the input data and preserves only core features; the dynamic RULE mapping criterion calculation and update step comprises the step of constructing a multilayer perceptron-based deep learning network using a mean squared error loss function with respect to the preprocessed data as input, calculating the gradient of the loss function for the weights and biases of the current model for each learning iteration using a backpropagation algorithm, and repeating a learning process of updating model parameters (weights and biases) while setting the learning rate to a constant value using stochastic gradient descent, until a predetermined number of epochs is completed or the change in loss value becomes negligible and the convergence condition is satisfied.And after the completion of the above learning, a prediction result for the input data is calculated using the learned model, and the prediction result is compared and analyzed with the result according to a pre-set static RULE mapping standard, wherein the comparison is performed by quantitatively evaluating the degree of agreement between the actual customs clearance history and the prediction result, and includes a step of calculating a RULE mapping standard that reflects changes in the customs clearance environment and data patterns through an algorithm that determines the RULE mapping standard based on a pre-set threshold value calculated as an evaluation indicator; and the blockchain recording step comprises: a step of serializing the entire content of the processing data in byte units and calculating a fixed-length hash value by applying a defined cryptographic hash function; a step of combining the calculated hash value and the serialized processing data, generating a digital signature according to the RSA algorithm using a pre-registered private key, and attaching the digital signature and hash value to the corresponding data block to ensure the integrity of the data; A method for verifying import / export requirements, comprising the step of executing a secure consensus algorithm on the data block to calculate a proof-of-work value that satisfies a difficulty objective based on the cryptographic hash value of the data block, transmitting the data block containing the calculated proof-of-work value to a network so that all participating nodes on the blockchain network reach consensus, and confirming that the data block is permanently recorded in a distributed ledger without tampering through the consensus process. Claim 3 A method for verifying import and export requirements according to paragraph 2, further comprising: a step of automatically initiating and performing an import / export declaration examination procedure based on relevant electronic documents and data including requirement application information and examination history for import / export products for which the requirement examination has been completed through the Korea Customs Service server after the final customs clearance requirement verification step; a step of receiving the import / export declaration examination result from the Korea Customs Service server, analyzing the received import / export declaration examination result data to generate an electronic examination result document including the examination result, related processing history, and examination details; and a step of executing an algorithm that automatically maps the requirement application information electronic document and the import / export declaration examination result electronic document based on the electronic examination result document to verify consistency between the requirement application information and the final customs clearance result for the import / export products. Claim 4 delete Claim 5 delete Claim 6 delete Claim 7 In paragraph 2, the step of calculating and updating dynamic RULE mapping criteria comprises: a step of numerically evaluating the necessity of updating each item by directly comparing the RULE mapping criteria calculated by the learned model with each item of an existing static RULE mapping table; a step of replacing the corresponding item of the existing RULE mapping table by applying the calculated RULE mapping criteria to the item requiring update according to the evaluation result, and generating an update log including the change history of the replaced item, the update time, and the applied RULE mapping criteria; and a step of automatically verifying the consistency between the generated update log and the updated RULE mapping table by a processor, and then monitoring the update results in real time and immediately reflecting them in a customs data processing system. Claim 8 delete

Images