Bank flow analysis system and method based on BI dialogue model

Abstract

The invention provides a bank statement analysis system and method based on a BI dialogue model, and belongs to the technical field of data processing, and the system comprises a user interaction module which receives a bank statement, receives a natural language, and displays the obtained bank statement and a bank statement analysis result; the database construction module analyzes and processes the bank flow and then imports the bank flow into a database; the SQL query statement generation module is used for preprocessing a natural language and then calling a Haiyu BI dialogue model interface to generate an SQL query statement; the bank statement query module is connected with the database to execute the SQL query statement to obtain the required bank statement; and the bank flow analysis processing module responds to the analysis request of the user to summarize and analyze the required bank flow. According to the method, automatic analysis and display of the bank flow are realized, the analysis efficiency and accuracy are high, the error rate of manual operation is reduced, integration analysis of inter-bank data is supported, query result display and bank flow analysis are diversified, and the requirements of complex business scenes are met.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a bank flow analysis system and method based on a BI dialogue model. Background Art

[0002] As a representation of the business status of enterprises and personal financial status, the demand for bank statement analysis is increasing. Traditional bank statement analysis methods often rely on manual operations, which have problems such as large workload, low efficiency and easy errors. An intelligent automatic analysis method is urgently needed.

[0003] First, traditional bank flow analysis mainly relies on manual work, which takes a lot of time and manpower, and is easily limited by personal experience and judgment, making it difficult to ensure the accuracy and reliability of the analysis results. Secondly, bank flow data usually comes from different banks and financial institutions, and their formats and contents vary greatly, which brings great difficulties to the unified processing and analysis of data. Thirdly, existing bank flow analysis tools can often only provide simple data query and statistical functions, lack intelligent analysis methods, and cannot meet the needs of banks for complex business scenarios.

[0004] In summary, the traditional bank flow analysis method uses manual operation, which is not only labor-intensive, but also affects the analysis quality. The inconsistency of data formats increases the difficulty of preprocessing and hinders the integrated analysis of cross-bank data. The single function of existing tools limits their use in complex application scenarios and can no longer meet the current needs for efficiency, accuracy, and intelligence. Therefore, it is particularly important to develop an intelligent bank flow analysis system. Summary of the invention

[0005] In a first aspect, an embodiment of the present application provides a bank flow analysis system based on a BI dialogue model, including: The user interaction module receives the bank statements input by the user and provides them to the database construction module, receives the natural language input of the user's query and provides it to the back-end bank statement acquisition module, and displays the acquired bank statements and bank statement analysis results; The database construction module parses and processes the received bank statements and imports them into the database; The SQL query statement generation module pre-processes the received natural language and calls the Hairuo BI dialogue model interface to transmit it, and then receives the returned SQL query statement; The bank statement query module connects to the database to execute SQL query statements to obtain the required bank statements and convert them into a recognizable format; The bank statement analysis and processing module responds to the user's analysis request and performs summary analysis on the required bank statements.

[0006] Furthermore, the user interaction module includes: The bank statement import unit responds to the user's operation of inputting the bank statement file stream and provides it to the database construction module; The dialogue unit responds to the natural language of the user's query and provides it to the back-end bank statement query module through HTTP request; The bank statement display unit displays the bank statements queried by the bank statement query module in a preset format; The analysis request acquisition unit responds to the user's request for bank statement analysis and provides it to the back-end bank statement analysis module; The bank statement analysis result display unit displays the bank statement analysis result generated according to the user's analysis request in a preset format.

[0007] Furthermore, the database construction module includes: The bank statement parsing unit recognizes the bank statements in the input CVS file, XLS file and PDF file format, and uses the bank statement template or general template of the corresponding bank to parse and identify the content of each field; The bank statement processing unit removes invalid and duplicate records from the parsed bank statements; The bank statement import unit encrypts the processed bank statements and imports them into the database in the form of single files or batch files, and backs up the database regularly.

[0008] Furthermore, the SQL query statement generation module includes: A natural language preprocessing unit removes irrelevant characters from the natural language obtained by the user interaction module and then performs word segmentation to complete preprocessing; The request data construction unit obtains the session ID, fills the pre-processed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; The BI dialogue model calling unit obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, packages them with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; The SQL query statement acquisition unit receives and parses the response data returned by the Hairuo BI dialogue model interface to extract the SQL query statement.

[0009] Furthermore, the bank statement query module includes: The SQL query statement execution unit uses JDBC to connect to the database to execute SQL query statements in the database and generate query results; The bank statement tabular processing unit processes the query results according to the preset format and generates a tabular form, which is returned to the user interaction module for display; The bank flow graphical processing unit generates a chart of the corresponding style in response to the chart type selected by the user from the query results, and then returns to the user interaction module for ECHARTS chart display.

[0010] Furthermore, the bank flow analysis and processing module includes: An analysis request type identification unit identifies the analysis request type of the bank statement, wherein the analysis request type includes bill analysis, bill detail analysis, fund flow analysis, common transaction analysis, and relationship network analysis; The bill analysis unit summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the user interaction module for display; The bill details analysis unit summarizes and analyzes the bill details in the queried bank flow according to the transaction date, transaction amount, cardholder name, card number, account opening bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the user interaction module for display; The fund flow analysis unit takes the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the inflow transaction objects, and the right binary tree nodes with the outflow transaction objects, and expands the flow according to the set level, and then provides the summary results to the user interaction module for display; The common transaction analysis unit counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the user interaction module for display; The relationship network analysis unit summarizes the relationship types between the transferor and the transferee in the queried bank flow, and provides them to the user interaction module for differentiated display according to the relationship type.

[0011] In a second aspect, the embodiment of the present application further provides a bank flow analysis method based on the BI dialogue model, comprising the following steps: S100. Receive the bank statement input by the user and parse and process it and then pour it into the database; S200. After receiving the natural language input of the user's query and preprocessing it, call the Hairuo BI dialogue model interface for transmission, then receive the returned SQL query statement, and then connect to the database to execute the SQL query statement to obtain the required bank statement and convert it into a recognizable format; S300. Summarize and analyze the acquired bank statements according to the user's analysis request.

[0012] Furthermore, the specific steps of step S100 are as follows: S101. In the front-end interface, respond to the user's input according to the bank statement import template, obtain the case, cardholder name, certificate type, certificate number, bank name and bank statement to be imported; S102. Identify the format of the input bank statement; If the file format is CVS, the file is read according to the CVS format and the process goes to step S106; When the file format is XLS, the file is read according to the XLS format and the process goes to step S106; When the file format is PDF, go to step S103; S103. Identify whether the input bank statement PDF file is a scanned file or an editable PDF / A format; If it is in editable PDF / A format, go to step S105; If the document is not scanned, proceed to step S104; S104. Use the OCR tool to read the file content and proceed to step S106; S105. Use pdfbox to read the file content; S106. Identify whether there is a bank statement template; If yes, the bank statement template of the account opening bank is used to parse the input bank statement, and the process goes to step S107; If not, use the common template to parse the input bank statement; S107. Remove the data with empty account number from the parsed bank statements, remove one of the two identical data, and filter out a record with completely identical content except for the primary key; S108. After the processed bank statements are encrypted, they are imported into the database individually or in batches in the manner selected by the user, and the database is backed up regularly to facilitate recovery in case of failure.

[0013] Furthermore, the specific steps of step S200 are as follows: S201. Respond to the natural language of the user's query on the front-end interface and provide it to the back-end processor via HTTP request; S202. The backend processor removes irrelevant characters from the acquired natural language, performs word segmentation, completes preprocessing, obtains the session ID, fills the preprocessed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; S203. The back-end processor obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, and packages it with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; S204. The back-end processor receives and parses the response data returned by the Hairuo BI dialogue model interface, extracts the SQL query statement, and then uses JDBC to connect to the database to execute the SQL query statement in the database to generate the query result; S205. The backend processor processes the query results in a preset format and generates a table, which is returned to the front-end interface for display to the user; S206. The front-end interface identifies whether the user selects a chart to display the query results; If yes, go to step S207; If not, proceed to step S300; S207. The front-end interface obtains the chart type selected by the user; If it is a line chart or a bar chart, go to step S208; If it is a pie chart, go to step S209; S208. The front-end interface obtains the X-axis field and the Y-axis field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS chart to display a line chart or a bar chart according to the corresponding data amplitude, and enters step S300; S209. The front-end interface obtains the key field and value field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS pie chart according to the corresponding data ratio.

[0014] Further, step S300 includes the following steps: S301. The front-end interface obtains and identifies the user's request for query result analysis; When the analysis request is for bill analysis, proceed to step S302; When the analysis request is for bill details analysis, proceed to step S303; When the analysis request is for capital flow analysis, proceed to step S304; When the analysis request is for joint transaction analysis, proceed to step S305; When the analysis request is a relationship network analysis, proceed to step S306; S302. The back-end processor summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the front-end interface for display; S303. The back-end processor analyzes the details of the bank statements in the query according to the transaction date, transaction amount, cardholder name, card number, bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the front-end interface for display; S304. The back-end processor uses the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the incoming transaction objects, and constructs the right binary tree nodes with the outgoing transaction objects, and expands the flow according to the set level, and then provides the summary results to the front-end interface for display; S305. The back-end processor counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the front-end interface for display; S306. The relationship types between the transferor and the transferee in the queried bank flow are summarized by the back-end processor, and provided to the front-end interface for differentiated display according to the relationship type.

[0015] It can be seen from the above technical solutions that the present invention has the following advantages: The bank flow analysis system and method based on the BI dialogue model provided by this application realizes the automatic analysis and display of bank flow in an intelligent way, improves the analysis efficiency and accuracy, and reduces the error rate of manual operation. At the same time, this application supports the integrated analysis of cross-bank data, provides a variety of query result display methods and comprehensive bank flow analysis functions, meets the needs of complex business scenarios, and provides strong decision-making support for banks and enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1It is a schematic diagram of an embodiment of a bank flow analysis system based on a BI dialogue model of the present invention.

[0018] Figure 2 It is a schematic diagram of another embodiment of the bank flow analysis system based on the BI dialogue model of the present invention.

[0019] Figure 3 It is a flow chart of the bank flow analysis method based on the BI dialogue model of the present invention.

[0020] Figure 4 This is a schematic diagram of the bank statement import template of the present invention.

[0021] Figure 5 It is a schematic diagram showing a line graph of the query result of the present invention. DETAILED DESCRIPTION

[0022] In the following, a bank flow analysis system based on the BI dialogue model will be described in detail, and various embodiments of the present disclosure will be described more comprehensively. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0023] For example, with the growing demand for bank flow analysis, the traditional bank flow analysis model relies heavily on manual work, which is not only time-consuming and inefficient, but also has a significant impact on the accuracy and credibility of the analysis results due to the subjectivity of personal experience and judgment. Furthermore, due to the diverse formats and complex content of the flow data provided by various banks and financial institutions, this undoubtedly sets up many obstacles for the standardized processing and comprehensive analysis of data. In addition, most of the current analysis tools on the market only have basic data query and statistical functions, lack the ability of deep intelligent analysis, and are unable to cope with the increasingly complex business needs of the banking industry.

[0024] In summary, the traditional manual analysis method is not only labor-intensive and time-consuming, but also affects the quality of the analysis results; the inconsistency of data formats increases the complexity of pre-processing work and hinders the integration and deep insight of cross-institutional data; and the functional limitations of existing tools restrict their use in diversified and deep application scenarios. Therefore, it has become a top priority to develop a system that can efficiently, accurately and intelligently analyze bank flow.

[0025] In response to the above problems, this embodiment provides a bank statement analysis system based on the BI dialogue model. Through the user interaction module, database construction module, SQL query statement generation module, bank statement query module and bank statement analysis and processing module, it realizes the intelligent analysis of bank statements, improves the analysis efficiency and accuracy, and meets the needs of complex business scenarios.

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] See also Figure 1 The figure is a schematic diagram of a bank flow analysis system based on a BI dialogue model in a specific embodiment, and the system includes: The user interaction module receives the bank statements input by the user and provides them to the database construction module, receives the natural language input of the user's query and provides it to the back-end bank statement acquisition module, and displays the acquired bank statements and bank statement analysis results; It should be noted that the user interaction module improves the user interactivity and usability of the system, allowing users to easily input bank statements, query and analyze results; The database construction module parses and processes the received bank statements and imports them into the database; It should be noted that the database construction module ensures the accuracy and completeness of bank flow data and improves the efficiency of data processing; The SQL query statement generation module pre-processes the received natural language and calls the Hairuo BI dialogue model interface to transmit it, and then receives the returned SQL query statement; It should be noted that the SQL query statement generation module realizes the automatic conversion of natural language to SQL query statements, which improves the convenience of query; The bank statement query module connects to the database to execute SQL query statements to obtain the required bank statements and convert them into a recognizable format; It should be noted that the bank statement query module provides a variety of query result display methods, which enhances the visualization effect of the system; The bank statement analysis and processing module responds to the user's analysis request and performs summary analysis on the required bank statements; It should be noted that the bank statement analysis and processing module provides comprehensive bank statement analysis functions to meet the needs of different users.

[0028] This embodiment automatically identifies and processes bank statement files, ensures the accuracy and completeness of data, provides personalized data analysis and suggestions, ensures the security of user data, provides data support for financial institutions, enterprises and individual users, and promotes business development.

[0029] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another bank flow analysis system based on the BI dialogue model is provided, such as Figure 2 As shown, the system includes: The user interaction module receives the bank statements input by the user and provides them to the database construction module, receives the natural language input by the user and provides it to the back-end bank statement acquisition module, and displays the acquired bank statements and bank statement analysis results; the user interaction module includes: The bank statement import unit responds to the user's operation of inputting the bank statement file stream and provides it to the database construction module; The dialogue unit responds to the natural language of the user's query and provides it to the back-end bank statement query module through HTTP request; The bank statement display unit displays the bank statements queried by the bank statement query module in a preset format; The analysis request acquisition unit responds to the user's request for bank statement analysis and provides it to the back-end bank statement analysis module; The bank statement analysis result display unit displays the bank statement analysis result generated according to the user's analysis request in a preset format; It should be noted that the user interaction module's bank statement import, natural language query, bank statement display, analysis request acquisition and bank statement analysis result display functions enhance the system's user interactivity and ease of use; The database construction module parses and processes the received bank statements and imports them into the database; the database construction module includes: The bank statement parsing unit recognizes the bank statements in the input CVS file, XLS file and PDF file format, and uses the bank statement template or general template of the corresponding bank to parse and identify the content of each field; The bank statement processing unit removes invalid and duplicate records from the parsed bank statements; The bank statement import unit encrypts the processed bank statements and imports them into the database in single or batch file form, and backs up the database regularly; It should be noted that the database construction module that parses, processes and imports bank statements ensures the accuracy and integrity of the data and improves the efficiency of data processing; The SQL query statement generation module pre-processes the received natural language and calls the Hairuo BI dialogue model interface to transmit it, and then receives the returned SQL query statement; the SQL query statement generation module includes: A natural language preprocessing unit removes irrelevant characters from the natural language obtained by the user interaction module and then performs word segmentation to complete preprocessing; The request data construction unit obtains the session ID, fills the pre-processed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; The BI dialogue model calling unit obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, packages them with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; SQL query statement acquisition unit, receiving and parsing the response data returned by the Hairuo BI dialogue model interface, and extracting the SQL query statement; It should be noted that the SQL query statement generation module realizes the automatic conversion from natural language to SQL query statement through natural language preprocessing, request data construction, BI dialogue model calling and SQL query statement acquisition, which improves the convenience of query; The bank statement query module connects to the database to execute SQL query statements to obtain the required bank statements and convert them into a recognizable format; the bank statement query module includes: The SQL query statement execution unit uses JDBC to connect to the database to execute SQL query statements in the database and generate query results; The bank statement tabular processing unit processes the query results according to the preset format and generates a tabular form, which is returned to the user interaction module for display; The bank flow graphical processing unit generates a corresponding style chart in response to the chart type selected by the user from the query results, and then returns to the user interaction module for ECHARTS chart display; It should be noted that the bank statement query module, which executes SQL query statements, tabulates bank statements, and processes them in graphics, provides a variety of query result display methods, which enhances the visualization effect of the system; The bank flow analysis and processing module responds to the user's analysis request and summarizes and analyzes the required bank flow; the bank flow analysis and processing module includes: An analysis request type identification unit identifies the analysis request type of the bank statement, wherein the analysis request type includes bill analysis, bill detail analysis, fund flow analysis, common transaction analysis, and relationship network analysis; The bill analysis unit summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the user interaction module for display; The bill details analysis unit summarizes and analyzes the bill details in the queried bank flow according to the transaction date, transaction amount, cardholder name, card number, account opening bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the user interaction module for display; The fund flow analysis unit takes the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the inflow transaction objects, and the right binary tree nodes with the outflow transaction objects, and expands the flow according to the set level, and then provides the summary results to the user interaction module for display; The common transaction analysis unit counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the user interaction module for display; The relationship network analysis unit summarizes the relationship types between the transferor and the transferee in the queried bank flow, and provides them to the user interaction module for differentiated display according to the relationship type; It should be noted that the bank statement analysis processing module provides comprehensive bank statement analysis by analyzing request type identification and bill analysis unit, bill detail analysis unit, and fund flow analysis unit to meet the needs of different users.

[0030] like Figure 3 As shown, the following is an embodiment of the bank flow analysis method based on the BI dialogue model provided in an embodiment of the present disclosure. This method and the bank flow analysis system based on the BI dialogue model in the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the bank flow analysis method based on the BI dialogue model, please refer to the embodiment of the bank flow analysis system based on the BI dialogue model.

[0031] The method comprises the following steps: S100. Receive the bank statement input by the user and parse and process it before importing it into the database; It should be noted that the accuracy and security of bank statement data are ensured by parsing and importing bank statements, providing a reliable basis for subsequent analysis; S200. After receiving the natural language input of the user's query and preprocessing it, call the Hairuo BI dialogue model interface for transmission, then receive the returned SQL query statement, and then connect to the database to execute the SQL query statement to obtain the required bank statement and convert it into a recognizable format; It should be noted that by converting natural language into SQL query statements, automatic conversion and display of natural language to query results are achieved, which improves the convenience and visualization of queries; S300. Summarize and analyze the acquired bank statements according to the user's analysis request; It should be noted that through comprehensive bank statement analysis, the needs of different users are met and the accuracy and depth of the analysis are improved.

[0032] This embodiment realizes intelligent analysis and display of bank statements by receiving bank statements, natural language queries and analysis requests input by users, thereby improving analysis efficiency and accuracy.

[0033] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process in this embodiment, another bank flow analysis method based on the BI dialogue model is provided, and the method includes the following steps: S100. Receive the bank statement input by the user and parse and process it and then pour it into the database; the specific steps of step S100 are as follows: S101. In the front-end interface, respond to the user's input according to the bank statement import template, obtain the case, cardholder name, certificate type, certificate number, bank name and bank statement to be imported; For example, by Figure 4 The bank statement import template shown is used to import bank statements; S102. Identify the format of the input bank statement; If the file format is CVS, the file is read according to the CVS format and the process goes to step S106; When the file format is XLS, the file is read according to the XLS format and the process goes to step S106; When the file format is PDF, go to step S103; S103. Identify whether the input bank statement PDF file is a scanned file or an editable PDF / A format; If it is in editable PDF / A format, go to step S105; If the document is not scanned, proceed to step S104; S104. Use the OCR tool to read the file content and proceed to step S106; S105. Use pdfbox to read the file content; S106. Identify whether there is a bank statement template; If yes, the bank statement template of the account opening bank is used to parse the input bank statement, and the process goes to step S107; If not, use the common template to parse the input bank statement; It should be noted that the export format of each bank is fixed, but the fields in the bank statement table exported by different banks have different file formats. Parse according to the specific export style and finally generate a unified data format; S107. Remove the data with empty account number from the parsed bank statements, remove one of the two identical data, and filter out a record with completely identical content except for the primary key; S108. After the processed bank flow is encrypted, it is imported into the database individually or in batches according to the method selected by the user, and the database is backed up regularly to facilitate recovery in case of failure; It should be noted that the steps of bank statement import and processing through format recognition, file reading, template use, data filtering and encrypted import ensure the accuracy and security of bank statement data; S200. After receiving the natural language input of the user's query and preprocessing it, call the Hairuo BI dialogue model interface for transmission, then receive the returned SQL query statement, and then connect to the database to execute the SQL query statement to obtain the required bank statement and convert it into a recognizable format; the specific steps of step S200 are as follows: S201. Respond to the natural language of the user's query on the front-end interface and provide it to the back-end processor via HTTP request; For example, you need to query all transaction records of a person 10 days before and after a transaction; S202. The backend processor removes irrelevant characters from the acquired natural language, performs word segmentation, completes preprocessing, obtains the session ID, fills the preprocessed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; S203. The back-end processor obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, and packages it with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; It should be noted that the Hairuo model needs to be trained in advance with bank statements to obtain the Hairuo BI dialogue model, so that the SQL query statement used as the query purpose can be obtained based on the bank statement-related terms; S204. The back-end processor receives and parses the response data returned by the Hairuo BI dialogue model interface, extracts the SQL query statement, and then uses JDBC to connect to the database to execute the SQL query statement in the database to generate the query result; S205. The backend processor processes the query results in a preset format and generates a table, which is returned to the front-end interface for display to the user; S206. The front-end interface identifies whether the user selects a chart to display the query results; If yes, go to step S207; If not, proceed to step S300; S207. The front-end interface obtains the chart type selected by the user; If it is a line chart or a bar chart, go to step S208; If it is a pie chart, go to step S209; S208. The front-end interface obtains the X-axis field and the Y-axis field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS chart to display a line chart or a bar chart according to the corresponding data amplitude, and enters step S300; For example, Figure 5 The X-axis fields shown can be selected from transaction date, account number, creation time, account balance, transaction amount, counterparty account name, IP address, relationship, bank, transaction type and summary label, counterparty account number, transaction time, customer name or transaction institution name, and the Y-axis fields can be selected from account balance, transaction amount and IP address; S209. The front-end interface obtains the key field and value field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS chart pie chart according to the corresponding data ratio; It should be noted that through the natural language query steps of preprocessing, data package construction, BI dialogue model calling, SQL query statement execution and result display, the automatic conversion and display of natural language to query results are realized, which improves the convenience and visualization effect of query; S300. The obtained bank statements are summarized and analyzed according to the user's analysis request; step S300 includes the following steps: S301. The front-end interface obtains and identifies the user's request for query result analysis; When the analysis request is for bill analysis, proceed to step S302; When the analysis request is for bill details analysis, proceed to step S303; When the analysis request is for capital flow analysis, proceed to step S304; When the analysis request is for joint transaction analysis, proceed to step S305; When the analysis request is a relationship network analysis, proceed to step S306; S302. The back-end processor summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the front-end interface for display; S303. The back-end processor analyzes the details of the bank statements in the query according to the transaction date, transaction amount, cardholder name, card number, bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the front-end interface for display; S304. The back-end processor uses the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the incoming transaction objects, and constructs the right binary tree nodes with the outgoing transaction objects, and expands the flow according to the set level, and then provides the summary results to the front-end interface for display; It should be noted that through the fund flow analysis, users can see at a glance the fund flow between the principal case holder and other transaction objects, whether it is the inflow or outflow of funds; users can easily track the flow of funds, starting from the principal case holder, and check the movement path of funds step by step; compared with manually flipping through bank statements, the binary tree structure makes the analysis process more efficient, and users can find transaction patterns or anomalies more quickly; users can not only analyze the transactions of the principal case holder, but also further check the transaction records of the transaction objects to achieve in-depth mining; S305. The back-end processor counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the front-end interface for display; It should be noted that the common transaction analysis chart can clearly show the transaction relationship, facilitate quick understanding, easily identify transaction patterns, frequency and scale, support in-depth analysis, quickly obtain key information, reduce decision-making time and risks, reveal potential business relationships or cooperation, and discover abnormal transactions; S306. The back-end processor summarizes the relationship types between the transferor and the transferee in the queried bank flow, and provides it to the front-end interface for display according to the relationship type; It should be noted that through relationship network analysis, users can intuitively see the complex relationship network between the transferor and the transferee, as well as the transaction activities between them based on the relationship chart, without having to read and analyze large amounts of data in depth; through the division of six relationship types (other, family, work, cooperation, society and unmaintained), users can more clearly understand the specific relationship between the transferor and the transferee; through the combination of relationship charts and data lists, users can easily group, filter and analyze data; automated data query and chart generation significantly reduce the time and energy of users in manually processing and analyzing data, thereby improving work efficiency.

[0034] It should be noted that the steps of bank statement analysis and processing are implemented through analysis request identification and various analysis functions, providing comprehensive bank statement analysis functions, meeting the needs of different users and improving the accuracy and depth of analysis.

[0035] This embodiment can not only calculate the annual / monthly income and expenditure of the cardholder, but also analyze in detail the key information such as the flow of transaction funds, frequency, location distribution, object type, and proportion of large transactions; through these analyses, it can help the case management personnel such as the Commission for Discipline Inspection to confirm whether the current cardholder has any illegal transactions. At the same time, users can also ask questions related to bank statements based on the Hairuo BI dialogue model to analyze and obtain the required data.

[0036] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0037] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A bank flow analysis system based on BI dialogue model, characterized in that: include: The user interaction module receives the bank statements input by the user and provides them to the database construction module, receives the natural language input of the user's query and provides it to the back-end bank statement acquisition module, and displays the acquired bank statements and bank statement analysis results; The database construction module parses and processes the received bank statements and imports them into the database; The SQL query statement generation module pre-processes the received natural language and calls the Hairuo BI dialogue model interface to transmit it, and then receives the returned SQL query statement; The bank statement query module connects to the database to execute SQL query statements to obtain the required bank statements and convert them into a recognizable format; The bank statement analysis and processing module responds to the user's analysis request and performs summary analysis on the required bank statements.

2. The bank flow analysis system based on the BI dialogue model according to claim 1 is characterized in that: The user interaction module includes: The bank statement import unit responds to the user's operation of inputting the bank statement file stream and provides it to the database construction module; The dialogue unit responds to the natural language of the user's query and provides it to the back-end bank statement query module through HTTP request; The bank statement display unit displays the bank statements queried by the bank statement query module in a preset format; The analysis request acquisition unit responds to the user's request for bank statement analysis and provides it to the back-end bank statement analysis module; The bank statement analysis result display unit displays the bank statement analysis result generated according to the user's analysis request in a preset format.

3. The bank flow analysis system based on the BI dialogue model according to claim 2 is characterized in that: The database building blocks include: The bank statement parsing unit recognizes the bank statements in the input CVS file, XLS file and PDF file format, and uses the bank statement template or general template of the corresponding bank to parse and identify the content of each field; The bank statement processing unit removes invalid and duplicate records from the parsed bank statements; The bank statement import unit encrypts the processed bank statements and imports them into the database in the form of single files or batch files, and backs up the database regularly.

4. The bank flow analysis system based on the BI dialogue model according to claim 3 is characterized in that: The SQL query statement generation module includes: A natural language preprocessing unit removes irrelevant characters from the natural language obtained by the user interaction module and then performs word segmentation to complete preprocessing; The request data construction unit obtains the session ID, fills the pre-processed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; The BI dialogue model calling unit obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, packages them with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; The SQL query statement acquisition unit receives and parses the response data returned by the Hairuo BI dialogue model interface to extract the SQL query statement.

5. The bank flow analysis system based on the BI dialogue model according to claim 4 is characterized in that: The bank statement query module includes: The SQL query statement execution unit uses JDBC to connect to the database to execute SQL query statements in the database and generate query results; The bank statement tabular processing unit processes the query results according to the preset format and generates a tabular form, which is returned to the user interaction module for display; The bank flow graphical processing unit generates a chart of the corresponding style in response to the chart type selected by the user from the query results, and then returns to the user interaction module for ECHARTS chart display.

6. The bank flow analysis system based on the BI dialogue model according to claim 5 is characterized in that: The bank flow analysis and processing module includes: An analysis request type identification unit identifies the analysis request type of the bank statement, wherein the analysis request type includes bill analysis, bill detail analysis, fund flow analysis, common transaction analysis, and relationship network analysis; The bill analysis unit summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the user interaction module for display; The bill details analysis unit summarizes and analyzes the bill details in the queried bank flow according to the transaction date, transaction amount, cardholder name, card number, account opening bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the user interaction module for display; The fund flow analysis unit takes the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the inflow transaction objects, and the right binary tree nodes with the outflow transaction objects, and expands the flow according to the set level, and then provides the summary results to the user interaction module for display; The common transaction analysis unit counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the user interaction module for display; The relationship network analysis unit summarizes the relationship types between the transferor and the transferee in the queried bank flow, and provides them to the user interaction module for differentiated display according to the relationship type.

7. A bank flow analysis method based on BI dialogue model, characterized in that: The steps include: S100. Receive the bank statement input by the user and parse and process it and then pour it into the database; S200. After receiving the natural language input of the user's query and preprocessing it, call the Hairuo BI dialogue model interface for transmission, then receive the returned SQL query statement, and then connect to the database to execute the SQL query statement to obtain the required bank statement and convert it into a recognizable format; S300. Summarize and analyze the acquired bank statements according to the user's analysis request.

8. The bank flow analysis method based on the BI dialogue model according to claim 7 is characterized in that: The specific steps of step S100 are as follows: S101. In the front-end interface, respond to the user's input according to the bank statement import template, obtain the case, cardholder name, certificate type, certificate number, bank name and bank statement to be imported; S102. Identify the format of the input bank statement; If the file format is CVS, the file is read according to the CVS format and the process goes to step S106; When the file format is XLS, the file is read according to the XLS format and the process goes to step S106; When the file format is PDF, go to step S103; S103. Identify whether the input bank statement PDF file is a scanned file or an editable PDF / A format; If it is in editable PDF / A format, go to step S105; If the document is not scanned, proceed to step S104; S104. Use the OCR tool to read the file content and proceed to step S106; S105. Use pdfbox to read the file content; S106. Identify whether there is a bank statement template; If yes, the bank statement template of the account opening bank is used to parse the input bank statement, and the process goes to step S107; If not, use the common template to parse the input bank statement; S107. Remove the data with empty account number from the parsed bank statements, remove one of the two identical data, and filter out a record with completely identical content except for the primary key; S108. After the processed bank statements are encrypted, they are imported into the database individually or in batches in the manner selected by the user, and the database is backed up regularly to facilitate recovery in case of failure.

9. The bank flow analysis method based on the BI dialogue model according to claim 8 is characterized in that: The specific steps of step S200 are as follows: S201. Respond to the natural language of the user's query on the front-end interface and provide it to the back-end processor via HTTP request; S202. The backend processor removes irrelevant characters from the acquired natural language, performs word segmentation, completes preprocessing, obtains the session ID, fills the preprocessed natural language and session ID into the preset request data template, and converts it into a format recognizable by the Hairuo dialogue model to obtain a data packet; S203. The back-end processor obtains the interface address, access key, and request timeout parameter of the Hairuo BI dialogue model, and packages it with the data packet to generate a data message, and then uses the HTTP client to send it to the Hairuo BI dialogue model interface; S204. The back-end processor receives and parses the response data returned by the Hairuo BI dialogue model interface, extracts the SQL query statement, and then uses JDBC to connect to the database to execute the SQL query statement in the database to generate the query result; S205. The backend processor processes the query results in a preset format and generates a table, which is returned to the front-end interface for display to the user; S206. The front-end interface identifies whether the user selects a chart to display the query results; If yes, go to step S207; If not, proceed to step S300; S207. The front-end interface obtains the chart type selected by the user; If it is a line chart or a bar chart, go to step S208; If it is a pie chart, go to step S209; S208. The front-end interface obtains the X-axis field and the Y-axis field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS chart to display a line chart or a bar chart according to the corresponding data amplitude, and enters step S300; S209. The front-end interface obtains the key field and value field selected by the user, provides them to the back-end processor for statistics, and uses the ECHARTS pie chart according to the corresponding data ratio.

10. The bank flow analysis method based on BI dialogue model according to claim 9 is characterized in that: Step S300 includes the following steps: S301. The front-end interface obtains and identifies the user's request for query result analysis; When the analysis request is for bill analysis, proceed to step S302; When the analysis request is for bill details analysis, proceed to step S303; When the analysis request is for capital flow analysis, proceed to step S304; When the analysis request is for joint transaction analysis, proceed to step S305; When the analysis request is a relationship network analysis, proceed to step S306; S302. The back-end processor summarizes the number of relevant personnel, the number of relevant bank cards, the number of bill records, the basic information of the survey subjects, the transaction analysis of annual / monthly income and expenditure, the flow analysis of the top N transactions, the frequency distribution analysis of the top N transactions, the distribution pattern analysis of the top N transaction locations, the transaction object type analysis, and the proportion of large transactions, and provides the summary results to the front-end interface for display; S303. The back-end processor analyzes the details of the bank statements in the query according to the transaction date, transaction amount, cardholder name, card number, bank name, counterparty name, relationship type, counterparty card number, transaction institution and transaction type, and provides the summary results to the front-end interface for display; S304. The back-end processor uses the main case person as the central node for the queried bank flow, and constructs the left binary tree child nodes with the incoming transaction objects, and constructs the right binary tree nodes with the outgoing transaction objects, and expands the flow according to the set level, and then provides the summary results to the front-end interface for display; S305. The back-end processor counts the number of transactions between the transferor and the bank card used by the transferee in the queried bank flow, and provides the number of transactions to the front-end interface for display; S306. The relationship types between the transferor and the transferee in the queried bank flow are summarized by the back-end processor, and provided to the front-end interface for differentiated display according to the relationship type.

Images