Methods, apparatus, equipment, and media for transferring documents based on due diligence information

Abstract

This disclosure provides embodiments of a file handling method, apparatus, device, and medium based on due diligence information. One specific implementation of the method includes: generating account association graph information based on multi-dimensional profile data and graph construction technology; generating early warning feature information based on a pre-trained due diligence information generation model, account association graph information, and variable multi-dimensional profile data; generating due diligence information based on the pre-trained due diligence information generation model, account association graph information, preset due diligence template information, and early warning feature information; generating final due diligence information based on the due diligence information and graph construction technology; generating movement trajectory information based on the final due diligence information; and controlling a file handling robot to handle the target file based on the movement trajectory information. This implementation reduces the time required for file handling based on due diligence information and decreases the computational resources consumed.

Description

Technical Field

[0001] Embodiments of this disclosure relate to the field of computer technology, and more specifically to methods, apparatus, devices, and media for transferring files based on due diligence information. Background Technology

[0002] When a company formally transfers its business (for example, a securities company transfers the business of users in its branches (business departments) that meet the transfer conditions to the securities company headquarters; the transfer conditions could be users whose risk level has risen to a higher risk level), the user's original due diligence report becomes invalid (e.g., due to changes in customer risk). The branch needs to generate a new due diligence report to determine whether the user's business needs to be transferred to the securities company headquarters, thus achieving the transfer of the user's business and enabling centralized supervision of high-risk business users. Currently, when determining whether to move user files to achieve the transfer, the common approach is as follows: if the user's original due diligence report becomes invalid (e.g., due to changes in customer risk), a new due diligence report needs to be generated using a fixed rule engine and traditional data processing workflow, and then the user's file needs to be moved to complete the file transfer process.

[0003] However, in practice, when using the above method to transfer user files, the following technical problems often arise:

[0004] Traditional data processing workflows rely on rigid, fixed rule engine architectures that cannot dynamically adapt to the data (e.g., knowledge or relevant regulations) upon which due diligence reports are generated. This not only requires frequent adjustments to rule parameters but also lacks a unified standard for data processing workflows across different business scenarios. Consequently, a large amount of repetitive development work occurs during system iterations, and the accuracy of generating new due diligence information is low. This leads to the control of the file-moving robot mistakenly moving user files to the handover area, requiring repeated rework of the incorrectly moved files. The file-moving robot's movement is time-consuming and consumes a lot of operating resources.

[0005] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0007] Some embodiments of this disclosure provide methods, apparatus, electronic devices, and computer-readable media for transferring files based on due diligence information to address one or more of the technical problems mentioned in the background section above.

[0008] In a first aspect, some embodiments of this disclosure provide a file handling method based on due diligence information. The method includes: acquiring multi-dimensional profile data of a user corresponding to a target file; generating account association graph information corresponding to the user based on the multi-dimensional profile data and graph construction technology; generating early warning feature information corresponding to the user based on the graph construction technology, a pre-trained due diligence information generation model, the account association graph information, and variable multi-dimensional profile data; generating due diligence information corresponding to the user based on the pre-trained due diligence information generation model, the account association graph information, preset due diligence template information, the early warning feature information, and a preset report verification tool; generating final due diligence information corresponding to the user based on the due diligence information and the graph construction technology; generating movement trajectory information corresponding to the target file based on the final due diligence information; and controlling a file handling robot to handle the target file based on the movement trajectory information.

[0009] Secondly, some embodiments of this disclosure provide a file handling device based on due diligence information, comprising: an acquisition unit configured to acquire multi-dimensional profile data information of a user corresponding to a target file; a first generation unit configured to generate account association graph information corresponding to the user based on the multi-dimensional profile data information and graph construction technology; a second generation unit configured to generate early warning feature information corresponding to the user based on the graph construction technology, a pre-trained due diligence information generation model, the account association graph information, and variable multi-dimensional profile data information; a third generation unit configured to generate due diligence information corresponding to the user based on the pre-trained due diligence information generation model, the account association graph information, preset due diligence template information, the early warning feature information, and a preset report verification tool; a fourth generation unit configured to generate final due diligence information corresponding to the user based on the due diligence information and the graph construction technology; a fifth generation unit configured to generate movement trajectory information corresponding to the target file based on the final due diligence information; and a control unit configured to control a file handling robot to handle the target file based on the movement trajectory information.

[0010] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0011] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in any of the implementations of the first or second aspect.

[0012] The above-described embodiments of this disclosure have the following beneficial effects: Through a file transfer method based on due diligence information according to some embodiments of this disclosure, the file transfer time can be shortened, and the operating resources of the file transfer robot can be reduced. The reason for the long file transfer time and high operating resource consumption of the file transfer robot is that the fixed rule engine architecture relied upon by traditional data processing flows is rigid and cannot dynamically adapt to the data (e.g., knowledge or relevant regulations) on which the due diligence report is generated. This not only requires frequent adjustments to rule parameters, but also lacks a unified standard for data processing flows in different business scenarios. Consequently, a large amount of repetitive development work occurs during system iteration, and the accuracy of generating new due diligence information is low. This leads to the control of the file transfer robot to incorrectly transfer user files to the handover area, requiring repeated rework of the incorrectly transferred files. The file transfer robot's transfer time is long and its operating resources are high. Therefore, the file transfer method based on due diligence information according to some embodiments of this disclosure first obtains multi-dimensional profile data information of the user corresponding to the target file. Thus, the user's multi-dimensional profile data can be obtained. Then, based on the aforementioned multi-dimensional profile data and graph construction technology, account association graph information corresponding to the aforementioned user is generated. This allows for the processing of the multi-dimensional profile data to obtain the user's account association graph. Next, based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and the variable multi-dimensional profile data, early warning feature information corresponding to the aforementioned user is generated. This yields the user's early warning features. Next, based on the pre-trained due diligence information generation model, the account association graph information, the preset due diligence template information, the early warning feature information, and the preset report verification tool, due diligence information corresponding to the aforementioned user is generated. This yields the user's due diligence report. Then, based on the due diligence information and the graph construction technology, final due diligence information corresponding to the aforementioned user is generated. This allows for the optimization of the due diligence information to obtain the final due diligence report. Finally, based on the final due diligence information, the movement trajectory information of the target file is generated. This determines the movement trajectory of the file robot transporting the target file. Finally, based on the aforementioned movement trajectory information, the file-handling robot is controlled to move the target files. This allows for the movement of the target files and their categorized storage. Furthermore, because the due diligence information is not generated using a fixed rule engine and traditional data processing procedures, but rather by first integrating and processing the acquired multi-dimensional profile data using the aforementioned graph construction technology, account association graph information is obtained. Then, the pre-trained due diligence information generation model extracts and processes this account association graph information, thereby obtaining early warning feature information.Next, the aforementioned preset due diligence template information is used to assist in generating due diligence information. Finally, the aforementioned graph construction technology is used to optimize the due diligence information. Thus, final due diligence information can be obtained, and the accuracy of the generated due diligence report information can be improved through graph construction technology. Therefore, user files can be moved based on the highly accurate final due diligence information, thereby shortening the file moving time and reducing the operating resources consumed by the file moving robot. Attached Figure Description

[0013] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0014] Figure 1 This is a flowchart of some embodiments of a file transfer method based on due diligence information according to the present disclosure;

[0015] Figure 2 This is a schematic diagram of the structure of some embodiments of a file handling device based on due diligence information according to the present disclosure;

[0016] Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0017] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0018] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0019] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0020] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0021] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0022] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0023] Figure 1 A flow 100 of some embodiments of a document transfer method based on due diligence information according to this disclosure is shown. This document transfer method based on due diligence information includes the following steps:

[0024] Step 101: Obtain multi-dimensional profile data of the user corresponding to the target file.

[0025] In some embodiments, the entity executing the file transfer method based on due diligence information (e.g., a computing device) can acquire multi-dimensional profile data of the user corresponding to the target file. The user can represent a user whose risk level has risen to the target risk level and whose original due diligence report has become invalid, requiring a new due diligence investigation. The target file can represent the user's paper file. The specific setting of the target risk level is not limited; for example, the target risk level can be Level 1 (highest risk). The multi-dimensional profile data can represent data related to the user obtained from multiple data sources. The specific data included in the multi-dimensional profile data is not limited and can be set according to actual needs. For example, the multi-dimensional profile data may include the user's name, contact person, and value transfer information (e.g., transaction records). The specific types of the multiple data sources are not limited and can be determined according to actual needs. For example, multiple data sources may include, but are not limited to: CRM systems, pre-set blacklist systems, anti-money laundering systems, and abnormal value transfer systems. The specific type of the abnormal value transfer system is not limited; for example, the abnormal value transfer system may be an abnormal transaction system. In practice, the aforementioned executing entities can use ETL technology to extract relevant data of the aforementioned users from the aforementioned multiple data sources to obtain multi-dimensional profile data information of the aforementioned users.

[0026] Step 102: Based on multi-dimensional profile data and graph construction technology, generate account association graph information for the corresponding user.

[0027] In some embodiments, the aforementioned executing entity can generate account association graph information corresponding to the aforementioned user based on the aforementioned multi-dimensional profile data information and graph construction technology. The aforementioned graph construction technology can represent a knowledge graph analysis method. The aforementioned account association graph information can represent the aforementioned user's account group relationship graph, related transaction graph, same-origin flow graph, or convergent transaction graph. The aforementioned account group relationship graph can represent a knowledge graph of the association relationships between the user's various accounts (e.g., each of the user's accounts can have the aforementioned user as the customer name for each account). The aforementioned related transaction graph can represent a knowledge graph of value transfer records between various accounts (e.g., value transfer records can be transaction records). The aforementioned same-origin flow graph can represent a knowledge graph of the value transfer paths between the aforementioned recorded accounts (e.g., value transfer paths can be the flow paths of funds between the aforementioned accounts). The aforementioned convergent transaction graph can represent a knowledge graph of similar transaction behaviors existing between the aforementioned accounts. The aforementioned similar transaction behaviors can represent similar value transfer operations. For example, a value transfer operation can be a transfer.

[0028] In some optional implementations of certain embodiments, the aforementioned execution entity can generate account association graph information corresponding to the aforementioned user based on the aforementioned multi-dimensional profile data information and graph construction technology through the following steps:

[0029] The first step is to update the aforementioned multidimensional profile data to obtain updated multidimensional profile data as variable multidimensional profile data. This variable multidimensional profile data can represent the multidimensional profile data after data standardization. In practice, the implementing entity can use data cleaning methods to clean the multidimensional profile data to obtain variable multidimensional profile data. The specific type of data cleaning method is not limited here; for example, fuzzy deduplication can be used.

[0030] The second step is to generate account association graph information corresponding to the aforementioned users based on the graph construction technology and the variable multi-dimensional profile data information mentioned above.

[0031] In some optional implementations of certain embodiments, the aforementioned execution entity can generate account association graph information corresponding to the aforementioned user through the following steps based on the aforementioned graph construction technology and the aforementioned variable multi-dimensional profile data information:

[0032] The first step is to construct a unified data model based on the aforementioned graph construction techniques. This unified data model represents a data model that organizes, defines, and manages the dispersed and heterogeneous data of the aforementioned users according to a unified standard and structure. In practice, the aforementioned implementing entities can construct a unified data model using the aforementioned graph construction techniques.

[0033] The second step involves fusing the aforementioned variable multi-dimensional profile data based on a unified data model, using entity recognition and relation extraction to obtain account association graph information. In practice, firstly, the executing entity maps the variable multi-dimensional profile data to the entity and relation types defined in the unified data model, obtaining a structured intermediate representation of the variable multi-dimensional profile data. Then, a named entity recognition algorithm is used to identify entities in the structured intermediate representation, yielding the entities corresponding to the variable multi-dimensional profile data. Next, dependency parsing is used to extract relations from the structured intermediate representation, obtaining relation data between the entities in the variable multi-dimensional profile data. This relation data represents the relationships between various entities. Finally, based on the graph schema defined in the unified data model, the entities and relation data are imported into a graph database for storage and querying, resulting in the account association graph information. For example, the dependency parsing method could be a maximum spanning tree algorithm.

[0034] Step 103: Based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and the variable multi-dimensional profile data information, generate the warning feature information for the corresponding user.

[0035] In some embodiments, the aforementioned executing entity can generate warning feature information corresponding to the aforementioned user based on the aforementioned graph construction technology, the pre-trained due diligence information generation model, the aforementioned account association graph information, and the variable multi-dimensional profile data information. The aforementioned pre-trained due diligence information generation model can represent a large language model; for example, the pre-trained due diligence information generation model can be Qwen3. The aforementioned pre-trained due diligence information generation model can take the aforementioned account association graph information as input and potential warning information as output, take change regulatory document information as input and structured rule information as output, or take account association graph information as input and first due diligence report information as output. The aforementioned pre-trained due diligence information generation model can include an input embedding layer, an attention layer, a feedforward network layer, a normalization layer, and an output layer. The aforementioned warning feature information can represent the risk factors included in the aforementioned variable multi-dimensional profile data information. The aforementioned risk factors can include, but are not limited to, multiple of the following: abnormal value transfer operations exist in the various accounts of the aforementioned user (for example, abnormal value transfer operations can be abnormal fund flows or similar transaction behaviors exist between the various accounts of the aforementioned user). Here, no specific restrictions are placed on the above-mentioned abnormal value transfer operations; they can be set according to actual needs.

[0036] In some optional implementations of certain embodiments, the aforementioned execution entity can generate warning feature information corresponding to the aforementioned user through the following steps based on the aforementioned graph construction technology, the pre-trained due diligence information generation model, the aforementioned account association graph information, and the variable multi-dimensional profile data information:

[0037] The first step involves generating a warning feature generation framework for the aforementioned users based on the graph construction technology and account association graph information described above. This framework can represent risk indicators across different dimensions. The specific dimensions are not limited here; for example, they could include conducting due diligence on the users based on value transfer records or on the presence of abnormal value transfer operations. In practice, the implementing entity can use the graph construction technology to parse entity attributes, relationship strength, and abnormal association patterns from the account association graph information to obtain the warning feature generation framework.

[0038] The second step involves generating framework information based on the aforementioned warning characteristics and then performing the following steps:

[0039] First, based on the pre-trained due diligence information generation model and the aforementioned variable multi-dimensional profile data, potential early warning information corresponding to the aforementioned users is generated. This potential early warning information can characterize factors extracted from the variable multi-dimensional profile data that have a high probability of translating into actual losses or violations. The specific value of this "high probability" is not limited; for example, a high probability could be 80%. For example, actual losses could be losses that actually cause value (e.g., financial losses), and violations could be abnormal value transfer operations. In practice, the executing entity can input the aforementioned variable multi-dimensional profile data into the pre-trained due diligence information generation model to obtain the potential early warning information.

[0040] Then, based on the preset retrieval method information, external data information is obtained. The preset retrieval method information can represent the method of retrieval enhancement generation. The external data information can represent external knowledge. This external knowledge can include, but is not limited to, several of the following: the user's historical due diligence information, changes in laws and regulations, and industry benchmarks. Here, the specific content of the industry benchmarks is not limited. For example, an industry benchmark can be a basis for determining whether a value transfer operation is abnormal. The historical due diligence information can represent the user's due diligence reports prior to the current moment. In practice, the executing entity can obtain external data information from a vector database using the preset retrieval method information.

[0041] Finally, based on the aforementioned account association graph information, the potential early warning information and the aforementioned external data information are fused to obtain the early warning feature information corresponding to the aforementioned users. In practice, the aforementioned executing entity can use entity alignment technology to match the aforementioned account association graph information and the aforementioned potential early warning information based on the aforementioned external data information to obtain the early warning feature information.

[0042] Step 104: Generate due diligence information for the aforementioned users based on the pre-trained due diligence information generation model, account association graph information, preset due diligence template information, early warning feature information, and preset report verification tool.

[0043] In some embodiments, the aforementioned executing entity can generate due diligence information corresponding to the aforementioned user based on the aforementioned pre-trained due diligence information generation model, the aforementioned account association graph information, the preset due diligence template information, the aforementioned early warning feature information, and the preset report verification tool. The aforementioned preset due diligence template information can represent a pre-defined structured document template. For example, the document template includes customer name: xx, account name: xxx, value transfer record (e.g., transaction record): xx. The aforementioned preset report verification tool can represent a program for verifying and correcting the aforementioned user's due diligence information, or an API interface for verifying and correcting the aforementioned user's due diligence information. Here, the specific use of the aforementioned preset report verification tool is not limited and can be adjusted according to actual needs. The aforementioned due diligence information can represent the initially generated due diligence report corresponding to the aforementioned user.

[0044] In addressing the technical problems mentioned above by adopting technical solutions, and considering the application scenario—the formal handover of a company's business (e.g., a securities company transferring the business of users in its branches (business departments) that meet the handover conditions to the headquarters; the handover condition could be users whose risk level has increased to a higher risk level)—the following technical problem often arises: the original due diligence report of the user becomes invalid (e.g., due to changes in customer risk). The branch needs to regenerate a new due diligence report. Directly generating the user's due diligence report based on the model does not consider whether the model-generated user's due diligence report is supported by data, resulting in low accuracy of the model-generated user's due diligence report. This leads to the control of the file-moving robot incorrectly moving the user's files to the handover area, requiring repeated rework of the incorrectly moved files. The file-moving robot's movement is time-consuming and consumes a lot of operating resources. Considering the following requirements for this application scenario: adaptability to the handover of user business with changes in risk level, we decided to adopt the following solution:

[0045] In some optional implementations of certain embodiments, the aforementioned executing entity can generate due diligence information corresponding to the aforementioned user by following these steps: based on the aforementioned pre-trained due diligence information generation model, the aforementioned account association graph information, the preset due diligence template information, the aforementioned early warning feature information, and the preset report verification tool.

[0046] The first step is to generate a first due diligence report based on the pre-trained due diligence information generation model and the account association graph information. This first due diligence report represents the initial version of the due diligence report. In practice, the executing entity can input the account association graph information into the pre-trained due diligence information generation model to obtain the first due diligence report information, and record the correspondence between entity nodes in the account association graph information and entity names mentioned in the first due diligence report information as an entity mapping table. For example, the entity mapping table could be "Entity Name: Account Last Four Digits 8899; Entity Node: ENT_2005".

[0047] The second step involves deep semantic extraction and summarization of the information in the first due diligence report to obtain key conclusions. These key conclusions can characterize the conclusion statements in the first due diligence report that reflect suspicious value transfer operations (e.g., suspicious transactions). The specific content of these conclusion statements is not limited; for example, a conclusion statement could be, "This client's value transfers (e.g., transaction history) over the past six months have involved frequent large-amount transactions with unknown third parties, suspected of involving suspicious value transfer operations (e.g., high-risk money laundering)." In practice, firstly, the executing entity can use a semantic role labeling algorithm based on bidirectional long short-term memory networks to perform semantic framework parsing of the first due diligence report information to obtain structured events. These structured events can include, but are not limited to, any of the following: value transfer entity (e.g., transaction entity), operation type, value transfer amount (e.g., transaction amount), and value transfer time (e.g., transaction time). Then, the k-means algorithm is used to identify high-frequency value transfer operations within these structured events. Next, the Rete algorithm is used to match the aforementioned high-frequency value transfer operations with a preset knowledge base to obtain risk tags corresponding to these operations. These risk tags characterize the risk level of the corresponding high-frequency value transfer operation. The preset knowledge base can be a risk control knowledge base. Finally, a template filling algorithm is used to transform the aforementioned structured events, high-frequency value transfer operations, and risk tags into key conclusion information. The template can be "{Value Transfer Subject}, {Value Transfer Amount}, {Value Transfer Time}".

[0048] The third step involves annotating the evidence sources for the key conclusions based on the aforementioned account association graph information, resulting in a first annotated due diligence report. This first annotated due diligence report represents the first due diligence report after annotating the evidence sources for the key conclusions. The evidence sources represent the data extracted from the account association graph information to support the key conclusions. In practice, firstly, the executing entity can determine the entity nodes in the account association graph information corresponding to the entity names mentioned in the key conclusions based on the entity mapping table. Then, based on the entity nodes in the account association graph information, the relevant nodes, edges, and attributes are extracted from the account association graph information using a graph query language. Next, the extracted nodes, edges, and attributes are used as an evidence subgraph. Finally, through document object model operations, the key conclusions are annotated based on the evidence sources to obtain the first annotated due diligence report.

[0049] The fourth step involves optimizing the structure of the first-annotated due diligence report information based on the aforementioned preset due diligence template information to obtain the second due diligence report information. This second due diligence report information represents the optimized first-annotated due diligence report information. In practice, firstly, the implementing entity can adjust the chapter structure of the first-annotated due diligence report information using the longest common subsequence algorithm based on the preset due diligence template information. Then, using preset optimization methods, the first-annotated due diligence report information after adjusting the chapter structure is filled and optimized to obtain the second due diligence report information. These preset optimization methods may include, but are not limited to, the following operations: content filling, text refinement, and data formatting.

[0050] The fifth step involves performing contradiction detection processing on the second due diligence report information based on the aforementioned early warning feature information and the aforementioned preset report verification tool, thereby obtaining verification error information. This verification error information characterizes contradictions and unreasonable analyses within the second due diligence report information. The specific content of these contradictions and unreasonable analyses is not limited here; for example, a contradiction could be "Report content: The client has had no large transactions in the past year; Transaction history: 3 single transactions exceeding ten million yuan." An unreasonable analysis could be "Risk assessment: Multiple overdue records exist; Risk assessment: The client has excellent credit." It should be noted that the verification error information can be empty. In practice, the implementing entity can use the aforementioned preset report verification tool, based on the aforementioned early warning feature information, to perform contradiction detection processing on the second due diligence report information to obtain verification error information.

[0051] Step 6: Based on the aforementioned verification error information and the aforementioned preset report verification tool, generate correction suggestion information. This correction suggestion information can represent statements proposing corrections to the aforementioned verification error information. The specific content of the correction suggestion information is not limited here; for example, it could be "It is suggested to correct Section 4.1 of the report, supplementing the related party 'Company B' and its shareholding ratio information according to the equity chart." In practice, the executing entity can use the Drools rule engine to match the aforementioned verification error information according to preset matching rules to obtain the correction suggestion information. The preset matching rules can represent the correspondence between the aforementioned verification error information and the aforementioned correction suggestion information. The specific content of the preset matching rules is not limited here; for example, the preset matching rule could be "Verification error information: Section 4.1 of the report states 'The client has no related party transactions,' corresponding to an error in the account association chart information: The client holds a controlling stake in Company B (70% shareholding); Correction suggestion information: It is suggested to correct Section 4.1 of the report, supplementing the related party 'Company B' and its shareholding ratio information according to the equity chart."

[0052] Step 7: Based on the aforementioned verification error information, the aforementioned preset report verification tool, and the aforementioned correction suggestion information, update the aforementioned second due diligence report information to obtain the updated second due diligence report information as the due diligence information. Based on the aforementioned due diligence information, control the file-moving robot to move the aforementioned target file. In practice, the executing entity can use the aforementioned preset report verification tool to merge the aforementioned second due diligence report information based on the aforementioned verification error information and the aforementioned correction suggestion information to obtain the merged second due diligence report information as the due diligence information, thereby correcting the second due diligence report information. It should be noted that the specific implementation steps for controlling the file-moving robot to move the aforementioned target file based on the aforementioned due diligence information can be found in steps 105-107, and will not be repeated here.

[0053] The above-described technical solution, as an inventive point of this disclosure, solves technical problem two: "the long time required for the file handling robot to move files and the high consumption of the file handling robot's operating resources." The reasons for the long time required for the file handling robot to move files and the high consumption of the file handling robot's operating resources are as follows: the user's original due diligence report becomes invalid (e.g., customer risk changes), and the branch office needs to regenerate a new due diligence report. Directly generating the user's due diligence report based on the model does not consider whether the model-generated user's due diligence report has data support, resulting in low accuracy of the model-generated user's due diligence report. This causes the file handling robot to incorrectly move the user's files to the handover area, resulting in repeated rework of the incorrectly moved files, thus increasing the time required for the file handling robot to move files and consuming more operating resources. Solving the above factors can shorten the time required for the file handling robot to move files and reduce the consumption of the file handling robot's operating resources. To achieve this effect, the file handling method based on due diligence information disclosed herein first performs deep semantic extraction on the initially generated due diligence report information. Then, based on the key conclusions obtained, it determines the evidence sources corresponding to the initially generated due diligence report information (to achieve data support). Furthermore, based on the preset due diligence template information, it further optimizes the initial due diligence report. Finally, it corrects contradictions and unreasonable analyses in the optimized due diligence report information. This improves the accuracy of the generated due diligence information, thereby shortening the time spent by the file handling robot in handling files and reducing the operating resources consumed by the file handling robot.

[0054] In addressing the technical challenges mentioned above, and considering the specific application scenario—securities companies transferring files of clients with significantly different client types (e.g., the first file is of a listed company, the second of an individual)—due diligence needs to be generated based on different process types. This often leads to the following technical problem: the complexity of processing different types of users varies. Using the same processing flow results in long processing times for less complex user types and low accuracy for highly complex user types. This leads to long file transfer times for the file transfer robot, consuming significant resources for the robot. To meet the following requirements for this application scenario: adaptability to different process types of due diligence and adaptability to significantly different user types, we have decided to adopt the following solution:

[0055] Optionally, after step 104, the aforementioned executing entity may also perform the following steps:

[0056] The first step involves parsing the core parameter information based on the preset transmission interface information to obtain the corresponding process type information for the aforementioned user. This process type information can be of three types: Type 1, Type 2, or Type 3. Type 1 process type information represents the process of opening an account for a new customer. Type 2 process type information represents the process of regenerating early warning feature information. Type 3 process type information represents the process of monitoring abnormal value transfer operations. The preset transmission interface information represents a pre-defined API interface for receiving and parsing the user's core parameter information. The core parameter information represents basic data related to the aforementioned user. The specific content of the core parameter information is not limited; for example, it could be "Customer ID": "C202509001". The specific content of the basic data is also not limited; for example, it could represent the user's name and contact person. In practice, the executing entity can call the preset transmission interface information to parse the core parameter information and obtain the corresponding process type information for the aforementioned user.

[0057] The second step involves generating target template information corresponding to the aforementioned process type information based on the process type information and template mapping table. This target template information can represent either a first target template or a second target template. The first target template information can represent a template for generating a basic due diligence report. The second target template information can represent a template for generating a dynamic due diligence report. Basic due diligence can represent a method of conducting due diligence for low-risk scenarios, based on a rule engine and fixed scripts to generate conclusions. Fixed scripts generating conclusions can represent the scripts used to generate the due diligence report. For example, fixed scripts generating conclusions could include initial account opening. The specific content of the low-risk scenario is not limited here; for example, a low-risk scenario could be the user opening an account for the first time. Dynamic due diligence can represent a method of conducting due diligence for high-risk scenarios, based on multi-source data analysis methods and the pre-trained due diligence information generation model to generate conclusions. The specific content of the high-risk scenario is not limited here; for example, a high-risk scenario could be the user's transactions exhibiting anomalies. The template mapping table described above represents the correspondence between the process type information and the target template information. For example, the template mapping table could be "Process type information: New customer account opening process; Target template information: First target template information". In practice, the executing entity can query the template mapping table using a lookup table to obtain the target template information corresponding to the process type information.

[0058] Third, based on the target template information above, perform the following steps:

[0059] The first sub-step involves generating first-type due diligence information based on a first due diligence mode, in response to the detection that the aforementioned process type information is first-type process type information and that the aforementioned first-type process type information meets a first preset trigger condition. The first preset trigger condition may be that the aforementioned user is opening a securities account for the first time. The first due diligence mode may be conducting basic due diligence on the aforementioned user. The first-type due diligence information may represent a due diligence report corresponding to the aforementioned first-type process type information. In practice, firstly, the executing entity may conduct a first-type investigation on the aforementioned user according to the aforementioned first due diligence mode. Then, the results of the first-type investigation are input into the aforementioned first target template information to obtain the first-type due diligence information. Here, the specific content of the first-type investigation is not limited; for example, a first-type investigation may represent an investigation into the aforementioned user's account transaction summary.

[0060] The second sub-step, in response to the detection that the aforementioned process type information is a second type of process type information, and that the aforementioned second type of process type information meets the second preset trigger condition, generates second type of due diligence information according to the second due diligence mode. The aforementioned second preset trigger condition can be a change in the user's status, for example, a change in the user's occupation. The specific content of the aforementioned second preset trigger condition is not limited here; for example, the second preset trigger condition can be a change in the user's basic data. The aforementioned second due diligence mode can be dynamic due diligence on the aforementioned user. The aforementioned second type of due diligence information can represent a due diligence report corresponding to the aforementioned second type of process type information. In practice, firstly, the aforementioned executing entity can conduct a second type of investigation on the aforementioned user according to the aforementioned second due diligence mode. Then, the results of the aforementioned second type of investigation are input into the aforementioned second target template information to obtain the second type of due diligence information. Here, the specific content of the aforementioned second type of investigation is not limited; for example, the second type of investigation can represent an investigation of the aforementioned user's multi-dimensional data (such as credit score, associated accounts).

[0061] The third sub-step, in response to the detection that the aforementioned process type information is third-type process type information, and that the aforementioned third-type process type information meets the third preset trigger condition, generates third-type due diligence information according to the second due diligence mode. The aforementioned third preset trigger condition can be that the user's account transactions are suspicious, for example, the third preset trigger condition can be that the user received multiple transfers exceeding a preset threshold during the same period. The aforementioned preset threshold can be 1 million. The aforementioned third-type due diligence information can represent a due diligence report corresponding to the aforementioned third-type process type information. In practice, firstly, the aforementioned executing entity can conduct three types of investigations on the aforementioned user according to the aforementioned second due diligence mode. Then, the results of the aforementioned three types of investigations are input into the aforementioned third target template information to obtain the third-type due diligence information. Here, the specific content of the aforementioned three types of investigations is not limited; for example, the three types of investigations can represent an investigation into the transaction details of the aforementioned user's associated accounts.

[0062] The fourth step is to identify the aforementioned Category I, Category II, or Category III due diligence information as due diligence information. It should be noted that the aforementioned Category I, Category II, or Category III due diligence information can be empty.

[0063] The above-described technical solution, as an inventive point of this disclosure, solves technical problem three: "leading to long processing time and high resource consumption for the file-handling robot." The reasons for this are as follows: different types of users have varying levels of complexity. Using the same processing flow results in longer processing times for users with lower complexity and lower accuracy for users with extremely high complexity, leading to longer file-handling time and higher resource consumption for file transfer. Solving these factors can shorten the time spent on file handling and reduce resource consumption. To achieve this, the file-handling method based on due diligence information in this disclosure first determines three different types of process type information corresponding to the user based on the account association graph information. Then, based on the different types of process type information, the first, second, and third preset trigger conditions are introduced as trigger conditions for generating due diligence information. Next, based on the aforementioned triggering conditions and process type information, the due diligence mode corresponding to the aforementioned user is determined. This due diligence mode can represent both the first and second due diligence modes. Finally, based on the aforementioned due diligence modes, due diligence information is generated for users with different process types. Therefore, different due diligence processes can be conducted using the first and second due diligence modes for users with different needs, resulting in personalized due diligence information generation. Furthermore, the first due diligence mode can be used for users requiring simple due diligence, thereby reducing the complexity and time required to generate due diligence information. This, in turn, can shorten the time spent controlling the file handling robot to move files based on due diligence information and reduce the consumption of the file handling robot's operating resources.

[0064] Step 105: Based on the due diligence information and graph construction technology, generate the final due diligence information for the corresponding user.

[0065] In some embodiments, the aforementioned implementing entity may generate final due diligence information corresponding to the aforementioned user based on the aforementioned due diligence information and the aforementioned graph construction technology. The final due diligence information may characterize the final due diligence report generated for the aforementioned user.

[0066] In addressing the technical challenges mentioned above, the application scenario—securities companies needing to dynamically adapt to the data upon which due diligence reports are generated (e.g., regulations and regulatory documents related to securities compliance and anti-money laundering)—often presents the following technical problem: The data is frequently updated, and rule extraction from unstructured data (e.g., regulatory policy documents) is complex, leading to lengthy due diligence report generation and consequently, lengthy user file transfer. Furthermore, inconsistencies within the data make it impossible to determine data reliability, resulting in inaccurate due diligence reports and consequently, inaccurate user file transfer, necessitating repeated rework, which is time-consuming and consumes significant resources for the file transfer robot. Considering the specific requirements of this application scenario—adapting to dynamically updated and highly complex data—I have decided to adopt the following solution:

[0067] In some optional implementations of certain embodiments, the aforementioned execution entity may generate final due diligence information corresponding to the aforementioned user based on the aforementioned due diligence information and the aforementioned graph construction technology through the following steps: and control the file handling robot to handle the aforementioned target file based on the aforementioned final due diligence information:

[0068] The first step involves modifying the acquired regulatory document information to obtain modified regulatory document information. The acquired regulatory document information can represent policy documents related to a preset theme. The specific type of the preset theme is not limited; for example, it could be a theme prohibiting abnormal value transfer operations. The policy documents can include, but are not limited to, any of the following: administrative regulations, departmental rules, or normative documents. The modified regulatory document information represents the pre-processed regulatory document information. In practice, the implementing entity can first standardize the regulatory document information by calling a Python code library to obtain standardized regulatory document information. The specific content of the Python code library is not limited; for example, it could be `pdftotext`. Then, semantic denoising is performed on the standardized regulatory document information through format cleaning to obtain the modified regulatory document information. The standardization process can represent converting special format text from different sources into plain text format. The special format text can include, but is not limited to, PDF, Word, and web pages. The semantic denoising process can represent removing non-essential text from the document. The aforementioned non-essential text may include, but is not limited to: headers, footers, and chart labels.

[0069] The second step involves generating structured rule information based on the aforementioned changes to the regulatory documents. This structured rule information can represent triples containing structured entries. For example, a structured entry could be "Condition Item - Behavior Item - Penalty Item". The specific content of these structured entries is not limited; for instance, a structured entry could be "Condition Item: Cross-border daily limit exceeding 500,000 - Behavior Item: Requires reporting of source of funds - Penalty Item: Warning if not reported". In practice, the implementing entity can first input the aforementioned changes to the regulatory documents into the pre-trained due diligence information generation model to obtain the structured rule information.

[0070] The third step involves generating rule base information corresponding to the aforementioned structured rule information based on the structured rule information and the graph construction technology. This rule base information can include static and dynamic rule base information. Static rule base information represents rules based on long-term, unchanging regulations. Dynamic rule base information represents rules based on temporary policies or newly enacted regulations. In practice, the implementing entity can use the graph construction technology to classify and process the structured rule information to obtain the rule base information.

[0071] The fourth step is to determine the scenario tag information corresponding to the aforementioned due diligence information. This scenario tag information can represent the business scenario of the due diligence information. The specific content of the scenario tag information is not limited here; for example, the business scenario could be the scenario of opening a new user account. The scenario tag information could be "due diligence is conducted on users opening new accounts." In practice, the executing entity can use named entity recognition and relation extraction methods to identify and process the aforementioned due diligence information to obtain the corresponding scenario tag information.

[0072] The fifth step involves generating rule matching information based on the aforementioned scenario tag information, rule base information, and preset conflict handling methods. The preset conflict handling methods can represent priority based on timeliness, hierarchy, or specificity. Timeliness priority indicates that newly issued policies take precedence over older policies. Hierarchical priority indicates that national-level regulations take precedence over industry or local rules. Specificity priority indicates that specific rules for a particular scenario take precedence over general compliance rules. The specific types of these scenarios are not limited; for example, a specific scenario could be a newly opened account. The rule matching information represents the structured rule information and compliance judgment results that match the due diligence information with the rule base information. The specific content of the compliance judgment results is not limited; for example, the compliance judgment result could indicate omissions. In practice, the implementing entity can first select the structured rule information corresponding to the scenario tag information from the rule base information based on the scenario tag information. Then, using the aforementioned graph construction technology, the rule matching information is compared with the due diligence information, and multi-rule conflicts are handled according to the decision tree algorithm and the aforementioned preset conflict handling method to obtain the rule matching information.

[0073] Step 6: In response to the determination that the above rule matching information meets the preset constraints, the due diligence information is modified according to the first annotation method to obtain the modified due diligence information. The preset constraints can be that the above rule matching information conforms to the requirements in the above rule base information. The first annotation method can be a method of annotating updated content information in the above due diligence information. The updated content information can represent a descriptive statement conforming to the above rule base information and a link to the original regulatory document corresponding to the above rule matching information. Here, the specific content of the descriptive statement is not limited; for example, the descriptive statement can be "conforms to Article X of XX Regulation". In practice, the implementing entity can annotate the above due diligence information according to the first annotation method to obtain the due diligence information.

[0074] Step 7: In response to the determination that the above rule matching information does not meet the preset constraints, the due diligence information is modified according to the second annotation method to obtain the modified due diligence information as the final due diligence information. Based on the final due diligence information, the file-moving robot is controlled to move the target file. The second annotation method can be to annotate rectification suggestion statements in the due diligence information. The specific content of the rectification suggestion statements is not limited here; for example, the rectification suggestion statement could be "supplementary explanation of the source of funds." In practice, firstly, in response to the determination that the above rule matching information does not meet the preset constraints, the executing entity can use a subgraph isomorphic search algorithm to associate the violating entity in the account association graph information. The specific content of the violating entity is not limited here; for example, the violating entity could be an account with abnormal value transfer operations. Then, the due diligence information is annotated according to the second annotation method to obtain the due diligence information. It should be noted that, based on the final due diligence information mentioned above, the specific implementation steps for controlling the file handling robot to move the target files can be found in steps 106-107, and will not be repeated here.

[0075] The above-described technical solution, as an inventive point of this disclosure, solves technical problem four: "leading to long processing times and high resource consumption of the file-handling robot during file transfer." The reasons for the long processing times and high resource consumption of the file-handling robot during file transfer are as follows: the data used is updated quickly, and rule extraction from unstructured data (e.g., regulatory policy documents) is complex, resulting in long processing times for generating due diligence reports and consequently long processing times for transferring user files; when there are contradictions in the data used, the reliability of the data cannot be determined, leading to the generation of incorrect due diligence reports, resulting in low accuracy of the generated reports and low accuracy in transferring user files, requiring repeated rework, which further increases processing time and resource consumption of the file-handling robot. Solving these factors can shorten the processing time and reduce the resource consumption of the file-handling robot. To achieve this effect, the file transfer method based on due diligence information disclosed in this disclosure first performs structured parsing of the pre-processed regulatory document information according to the pre-trained due diligence information generation model. Thus, structured rule information can be obtained. Then, the structured rule information is classified using the aforementioned graph construction technology. This yields rule base information. Next, based on the scene tag information corresponding to the due diligence information, structured rule information corresponding to the scene tag information is filtered from the rule base information. Then, the aforementioned preset conflict handling method is introduced to handle multiple rule conflicts. This yields rule matching information. Finally, combining the aforementioned first annotation method and the aforementioned second annotation method, the due diligence information is modified. Because two types of rule bases (static rule base information and dynamic rule base information) are constructed, the rules within the rule base can be classified, and the preset conflict handling method is used to handle conflicting rules within the rule base. This improves the accuracy of the data used to generate the due diligence report information, thereby improving the accuracy of the generated due diligence report information. This improves the accuracy of the file handling robot in moving the target files, thereby reducing the time spent on repeated rework and reducing the operating resources consumed by the file handling robot.

[0076] Step 106: Based on the final due diligence information, generate the movement trajectory information of the corresponding target file.

[0077] In some embodiments, the aforementioned executing entity may generate movement trajectory information corresponding to the target file based on the aforementioned final due diligence information. This movement trajectory information can characterize the trajectory of the file-handling robot moving the target file. The file-handling robot can be a lurking lifting AGV. It should be noted that the aforementioned executing entity and the aforementioned file-handling robot can be connected by communication. It should be pointed out that the aforementioned communication connection may include, but is not limited to, 3G / 4G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultrawideband) connection, and other currently known or future developed communication methods.

[0078] In some optional implementations of certain embodiments, the aforementioned executing entity may generate movement trajectory information corresponding to the aforementioned target file based on the aforementioned final due diligence information through the following steps:

[0079] The first step is to determine the storage level information corresponding to the aforementioned final due diligence information. This storage level information characterizes the storage level of the target files corresponding to the final due diligence information. For example, the storage level information could be Level 1 storage. In practice, the executing entity can use the Analytic Hierarchy Process (AHP) to determine the storage level information corresponding to the aforementioned final due diligence information. This allows for the classification and placement of the files to be transferred.

[0080] The second step is to determine the placement location information corresponding to the aforementioned final due diligence information based on the storage level information. This placement location information represents the location (three-dimensional coordinates) where the file handling robot needs to place the target file to be transferred. In practice, the executing entity can use an optimal adaptation algorithm to determine the placement location information corresponding to the aforementioned final due diligence information based on the storage level information.

[0081] The third step is to generate the movement trajectory information of the target file based on the placement location information and the current location information of the target file. In practice, the executing entity can use the Dijkstra algorithm to generate the movement trajectory information of the target file based on the placement location information and the current location information of the target file.

[0082] Step 107: Based on the movement trajectory information, control the file handling robot to move the target file.

[0083] In some embodiments, the aforementioned execution entity may control the file handling robot to move the aforementioned target file based on the aforementioned movement trajectory information.

[0084] The above-described embodiments of this disclosure have the following beneficial effects: Through a file transfer method based on due diligence information according to some embodiments of this disclosure, the file transfer time can be shortened, and the operating resources of the file transfer robot can be reduced. The reason for the long file transfer time and high operating resource consumption of the file transfer robot is that the fixed rule engine architecture relied upon by traditional data processing flows is rigid and cannot dynamically adapt to the data (e.g., knowledge or relevant regulations) on which the due diligence report is generated. This not only requires frequent adjustments to rule parameters, but also lacks a unified standard for data processing flows in different business scenarios. Consequently, a large amount of repetitive development work occurs during system iteration, and the accuracy of generating new due diligence information is low. This leads to the control of the file transfer robot to incorrectly transfer user files to the handover area, requiring repeated rework of the incorrectly transferred files. The file transfer robot's transfer time is long and its operating resources are high. Therefore, the file transfer method based on due diligence information according to some embodiments of this disclosure first obtains multi-dimensional profile data information of the user corresponding to the target file. Thus, the user's multi-dimensional profile data can be obtained. Then, based on the aforementioned multi-dimensional profile data and graph construction technology, account association graph information corresponding to the aforementioned user is generated. This allows for the processing of the multi-dimensional profile data to obtain the user's account association graph. Next, based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and the variable multi-dimensional profile data, early warning feature information corresponding to the aforementioned user is generated. This yields the user's early warning features. Next, based on the pre-trained due diligence information generation model, the account association graph information, the preset due diligence template information, the early warning feature information, and the preset report verification tool, due diligence information corresponding to the aforementioned user is generated. This yields the user's due diligence report. Then, based on the due diligence information and the graph construction technology, final due diligence information corresponding to the aforementioned user is generated. This allows for the optimization of the due diligence information to obtain the final due diligence report. Finally, based on the final due diligence information, the movement trajectory information of the target file is generated. This determines the movement trajectory of the file robot transporting the target file. Finally, based on the aforementioned movement trajectory information, the file-handling robot is controlled to move the target files. This allows for the movement of the target files and their categorized storage. Furthermore, because the due diligence information is not generated using a fixed rule engine and traditional data processing procedures, but rather by first integrating and processing the acquired multi-dimensional profile data using the aforementioned graph construction technology, account association graph information is obtained. Then, the pre-trained due diligence information generation model extracts and processes this account association graph information, thereby obtaining early warning feature information.Next, the aforementioned preset due diligence template information is used to assist in generating due diligence information. Finally, the aforementioned graph construction technology is used to optimize the due diligence information. Thus, final due diligence information can be obtained, and the accuracy of the generated due diligence report information can be improved through graph construction technology. Therefore, user files can be moved based on the highly accurate final due diligence information, thereby shortening the file moving time and reducing the operating resources consumed by the file moving robot.

[0085] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a file transfer method based on due diligence information. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0086] like Figure 2 As shown, some embodiments of the file transfer device 200 based on due diligence information include: an acquisition unit 201, a first generation unit 202, a second generation unit 203, a third generation unit 204, a fourth generation unit 205, a fifth generation unit 206, and a control unit 207. The system comprises the following components: an acquisition unit 201, configured to acquire multi-dimensional profile data of the user corresponding to the target file; a first generation unit 202, configured to generate account association graph information corresponding to the user based on the multi-dimensional profile data and graph construction technology; a second generation unit 203, configured to generate early warning feature information corresponding to the user based on the graph construction technology, a pre-trained due diligence information generation model, the account association graph information, and variable multi-dimensional profile data; a third generation unit 204, configured to generate due diligence information corresponding to the user based on the pre-trained due diligence information generation model, the account association graph information, a preset due diligence template, the early warning feature information, and a preset report verification tool; a fourth generation unit 205, configured to generate final due diligence information corresponding to the user based on the due diligence information and the graph construction technology; a fifth generation unit 206, configured to generate movement trajectory information corresponding to the target file based on the final due diligence information; and a control unit 207, configured to control a file handling robot to move the target file based on the movement trajectory information.

[0087] It is understandable that the units recorded in the document handling device 200 based on due diligence information are related to the references. Figure 1The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to the document handling device 200 based on due diligence information and the units contained therein, and will not be repeated here.

[0088] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device 300 (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0089] like Figure 3 As shown, the electronic device 300 may include a processing unit 301 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0090] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0091] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0092] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0093] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0094] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently without being assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: acquire multi-dimensional profile data information of the user corresponding to the target file; generate account association graph information corresponding to the user based on the aforementioned multi-dimensional profile data information and graph construction technology; generate early warning feature information corresponding to the user based on the aforementioned graph construction technology, a pre-trained due diligence information generation model, the aforementioned account association graph information, and variable multi-dimensional profile data information; generate due diligence information corresponding to the user based on the aforementioned pre-trained due diligence information generation model, the aforementioned account association graph information, preset due diligence template information, the aforementioned early warning feature information, and preset report verification tools; generate final due diligence information corresponding to the user based on the aforementioned due diligence information and the aforementioned graph construction technology; generate movement trajectory information corresponding to the target file based on the aforementioned final due diligence information; and control a file-handling robot to move the target file based on the aforementioned movement trajectory information.

[0095] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0096] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0097] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor, and for example, can be described as: an acquisition unit, a first generation unit, a second generation unit, a third generation unit, a fourth generation unit, a fifth generation unit, and a control unit. The names of these units do not necessarily limit the unit itself; for example, the acquisition unit can also be described as "a unit that acquires multi-dimensional profile data information of the user corresponding to the target file."

[0098] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0099] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method of archive transfer based on due diligence information, characterized by, include: Obtain multi-dimensional profile data of users corresponding to the target profile; Based on the multi-dimensional profile data and graph construction technology, account association graph information corresponding to the user is generated, wherein generating the account association graph information corresponding to the user based on the multi-dimensional profile data and graph construction technology includes: The multidimensional portrait data information is updated to obtain the updated multidimensional portrait data information as the variable multidimensional portrait data information; Based on the graph construction technology and the variable multi-dimensional profile data, account association graph information corresponding to the user is generated, wherein the account association graph information is the user's account group relationship graph, related transaction graph, same source transaction graph, or convergent transaction graph; Based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and the variable multi-dimensional profile data, early warning feature information corresponding to the user is generated. The step of generating the early warning feature information corresponding to the user based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and the variable multi-dimensional profile data includes: Based on the graph construction technology and the account association graph information, generate early warning feature generation framework information corresponding to the user; Generate framework information based on the aforementioned warning features, and perform the following steps: Based on the pre-trained due diligence information generation model and the variable multi-dimensional profile data, potential early warning information corresponding to the user is generated; Based on the preset retrieval method information, external data information is obtained; Based on the account association graph information, the potential early warning information and the external data information are fused to obtain the early warning feature information corresponding to the user; Based on the pre-trained due diligence information generation model, the account association graph information, the preset due diligence template information, the early warning feature information, and the preset report verification tool, due diligence information corresponding to the user is generated, wherein generating due diligence information corresponding to the user further includes: Based on the preset transmission interface information, the core parameter information is parsed and processed to obtain the process type information corresponding to the user, wherein the process type information is a first type of process type information, a second type of process type information, or a third type of process type information; Based on the process type information and the template mapping table, generate target template information corresponding to the process type information; Based on the target template information, perform the following steps: In response to detecting that the process type information is a first type of process type information and that the first type of process type information meets a first preset triggering condition, a first type of due diligence information is generated according to a first due diligence mode. In response to detecting that the process type information is a second type of process type information and that the second type of process type information meets a second preset triggering condition, a second type of due diligence information is generated according to the second due diligence mode; In response to the detection that the process type information is a third type of process type information and the third type of process type information meets a third preset triggering condition, a third type of due diligence information is generated according to the second due diligence mode. The first type of due diligence information, the second type of due diligence information, or the third type of due diligence information are identified as due diligence information; Based on the due diligence information and the graph construction technology, generate the final due diligence information corresponding to the user; Based on the final due diligence information, movement trajectory information corresponding to the target file is generated, wherein generating the movement trajectory information corresponding to the target file based on the final due diligence information includes: Based on the final due diligence information, determine the storage level information corresponding to the final due diligence information; Based on the storage level information, determine the placement information of the corresponding final due diligence information; Based on the placement location information and the current location information of the target file, generate the movement trajectory information corresponding to the target file; Based on the movement trajectory information, the file handling robot is controlled to move the target file.

2. The method of claim 1, wherein, The step of generating account association graph information corresponding to the user based on graph construction technology and the variable multi-dimensional profile data information includes: Based on the aforementioned map construction technology, a unified data model is constructed; Based on a unified data model, entity recognition, and relationship extraction, the variable multi-dimensional profile data is fused to obtain account association graph information.

3. The method of claim 1, wherein, The first type of process information is the process for opening an account for a new customer; the second type of process information is the process for regenerating early warning feature information; and the third type of process information is the process for monitoring abnormal value transfer operations.

4. The method of claim 1, wherein, The first due diligence mode is to conduct basic due diligence on the user; the second due diligence mode is to conduct dynamic due diligence on the user; the basic due diligence is a due diligence method for low-risk scenarios, which generates conclusions based on a rule engine and fixed scripts; the dynamic due diligence is a due diligence method for high-risk scenarios, which generates conclusions based on multi-source data analysis methods and the pre-trained due diligence information generation model.

5. A file handling device based on due diligence information, characterized in that, include: The acquisition unit is configured to acquire multi-dimensional profile data of the user corresponding to the target file; The first generation unit is configured to generate account association graph information corresponding to the user based on the multi-dimensional profile data information and graph construction technology. The step of generating the account association graph information corresponding to the user based on the multi-dimensional profile data information and graph construction technology includes: updating the multi-dimensional profile data information to obtain updated multi-dimensional profile data information as variable multi-dimensional profile data information; and generating the account association graph information corresponding to the user based on the graph construction technology and the variable multi-dimensional profile data information. The account association graph information is the user's account group relationship graph, related transaction graph, same-origin transaction graph, or convergent transaction graph. The second generation unit is configured to generate warning feature information corresponding to the user based on the graph construction technology, a pre-trained due diligence information generation model, the account association graph information, and variable multi-dimensional profile data. The generation of warning feature information corresponding to the user based on the graph construction technology, the pre-trained due diligence information generation model, the account association graph information, and variable multi-dimensional profile data includes: generating a warning feature generation framework information corresponding to the user based on the graph construction technology and the account association graph information; and performing the following steps based on the warning feature generation framework information: generating potential warning information corresponding to the user based on the pre-trained due diligence information generation model and the variable multi-dimensional profile data; acquiring external data information based on a preset retrieval method; and fusing the potential warning information and the external data information based on the account association graph information to obtain the warning feature information corresponding to the user. The third generation unit is configured to generate due diligence information corresponding to the user based on the pre-trained due diligence information generation model, the account association graph information, the preset due diligence template information, the early warning feature information, and the preset report verification tool. The generation of due diligence information corresponding to the user further includes: parsing and processing core parameter information according to preset transmission interface information to obtain process type information corresponding to the user, wherein the process type information is a first type of process type information, a second type of process type information, or a third type of process type information; generating target template information corresponding to the process type information based on the process type information and a template mapping table; and performing the following steps based on the target template information: [Followed by a specific function / method]. Upon detecting that the process type information is a first type of process type information and the first type of process type information meets a first preset triggering condition, first type of due diligence information is generated according to a first due diligence mode; upon detecting that the process type information is a second type of process type information and the second type of process type information meets a second preset triggering condition, second type of due diligence information is generated according to a second due diligence mode; upon detecting that the process type information is a third type of process type information and the third type of process type information meets a third preset triggering condition, third type of due diligence information is generated according to a second due diligence mode; the first type of due diligence information, the second type of due diligence information, or the third type of due diligence information is determined as due diligence information; The fourth generation unit is configured to generate final due diligence information corresponding to the user based on the due diligence information and the graph construction technology. The fifth generation unit is configured to generate movement trajectory information corresponding to the target file based on the final due diligence information. The step of generating the movement trajectory information corresponding to the target file based on the final due diligence information includes: determining storage level information corresponding to the final due diligence information based on the final due diligence information; determining placement location information corresponding to the final due diligence information based on the storage level information; and generating movement trajectory information corresponding to the target file based on the placement location information and the current location information of the target file. The control unit is configured to control the file handling robot to move the target file based on the movement trajectory information.

6. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 4.

7. A computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 4.

Images