Intelligent reconciliation method and system for procurement subject

Abstract

The invention discloses an intelligent reconciliation method and system for a procurement subject, and the method comprises the steps: firstly obtaining a target procurement subject archive library and an initial reconciliation request, and then respectively loading the request and corresponding first and second reconciliation prompt words to corresponding models, and obtaining the missing necessary first information item and the ambiguous second information item in the initial request. According to the technical scheme, the multiple pieces of undetermined confirmation information are determined according to the second information item and the archiving library, then the target account checking request is obtained by optimizing the initial request according to the supplementary content of the first information item and the confirmation instruction of the undetermined confirmation information, query is executed in the archiving library accordingly, intelligent and accurate account checking operation is achieved, and account checking efficiency and quality are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular, to an intelligent reconciliation method and system for procurement entities. Background Art

[0002] In procurement operations, the reconciliation work is crucial. Traditional reconciliation methods for procurement entities rely on manual checking of a large amount of procurement data, which is inefficient and error-prone. At the same time, initial reconciliation requests often have problems of missing or ambiguous information, resulting in inaccurate or delayed reconciliations. Facing a vast amount of procurement contracts and transaction records, existing simple query methods are difficult to meet complex and changing reconciliation requirements and cannot quickly and accurately obtain the required reconciliation information. There is an urgent need for an intelligent reconciliation method to improve the efficiency and accuracy of reconciliation. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent reconciliation method and system for procurement entities.

[0004] In a first aspect, an embodiment of the present invention provides an intelligent reconciliation method for a procurement entity, including:

[0005] Obtaining an archive library of a target procurement entity and an initial reconciliation request for querying the archive library of the target procurement entity;

[0006] Loading the initial reconciliation request and a first reconciliation prompt word corresponding to the initial reconciliation request into a first information item requirement parsing model to obtain a first information item, where the first information item is a required information item missing in the initial reconciliation request;

[0007] Loading the initial reconciliation request and a second reconciliation prompt word corresponding to the initial reconciliation request into a second information item requirement parsing model to obtain a second information item, where the second information item is an information item with ambiguity in the initial reconciliation request;

[0008] Determining a plurality of pending confirmation information corresponding to the second information item according to the second information item and the archive library of the target procurement entity;

[0009] Optimizing the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the plurality of pending confirmation information to obtain a target reconciliation request, and performing a query in the archive library of the target procurement entity according to the target reconciliation request.

[0010] In a possible implementation manner, before loading the initial reconciliation request and a first reconciliation prompt word corresponding to the initial reconciliation request into a first information item requirement parsing model to obtain a first information item, the method further includes:

[0011] Query the target procurement entity's archive database according to the initial reconciliation request to obtain an initial reconciliation result;

[0012] Load the initial reconciliation request, the initial reconciliation result, and a third reconciliation prompt word corresponding to the initial reconciliation request and the initial reconciliation result into a reconciliation complexity assessment model to obtain an inferred reconciliation complexity assessment result;

[0013] If the reconciliation complexity assessment result indicates that the initial reconciliation request does not require complex reconciliation processing, terminate the query according to the initial reconciliation request.

[0014] In a possible implementation manner, the step of querying the target procurement entity's archive database according to the initial reconciliation request to obtain an initial reconciliation result includes:

[0015] Obtain a contract element extraction rule;

[0016] Extract contract elements from the initial reconciliation request according to the contract element extraction rule;

[0017] Determine the clause coverage of the associated transaction instance with respect to the contract element according to the number of times the clause in each associated transaction instance in the target procurement entity's archive database matches the contract element;

[0018] Determine the clause constraint strength of the associated transaction instance with respect to the contract element according to the number of associated transaction instances containing the contract element in the target procurement entity's archive database;

[0019] Determine a write-off candidate set in each associated transaction instance in the target procurement entity's archive database according to the clause coverage and the clause constraint strength as the initial reconciliation result.

[0020] In a possible implementation manner, the third reconciliation prompt word is obtained through the following process, including:

[0021] Obtain a first business rule preset condition for a complex reconciliation processing determination condition;

[0022] Construct a first logical determination component according to the first business rule preset condition, and the first logical determination component is used to configure whether the initial reconciliation request requires complex reconciliation processing according to the first business rule preset condition;

[0023] Construct an execution strategy component and an archiving strategy component. The execution strategy component configures the processing when the initial reconciliation request requires complex reconciliation processing, and the archiving strategy component configures the processing when the initial reconciliation request does not require complex reconciliation processing;

[0024] Construct the third reconciliation prompt word according to the first logic determination component, the execution policy component, and the archiving policy component;

[0025] The first reconciliation prompt word is obtained through the following process, including:

[0026] Obtain the first contract clause specification and the first clause performance instance of each first information item;

[0027] Construct a second logic determination component according to the first contract clause specification and the first clause performance instance, where the second logic determination component is used to configure to determine whether the initial reconciliation request includes the first information item according to the first contract clause specification and the first clause performance instance;

[0028] Construct a clause matching prompt component and a clause missing prompt component. The clause matching prompt component configures the contract clause association prompt when the initial reconciliation request includes the first information item, and the clause missing prompt component configures the warning of contract element missing when the initial reconciliation request does not include the first information item;

[0029] Construct the first reconciliation prompt word according to the second logic determination component, the clause matching prompt component, and the clause missing prompt component;

[0030] The construction of the clause matching prompt component and the clause missing prompt component includes:

[0031] Obtain the contract clause parsing process, the clause matching status code, and the clause performance description. The contract clause parsing process is the parsing process when the initial reconciliation request includes the first information item, the clause matching status code represents that the initial reconciliation request includes the first information item, and the clause performance description is the text information when the initial reconciliation request includes the first information item;

[0032] Construct the clause matching prompt component according to the contract clause parsing process, the clause matching status code, and the clause performance description;

[0033] Obtain the contract write-off exception analysis process, the element missing identifier, and the element missing reason description. The contract write-off exception analysis process is the analysis process when the initial reconciliation request does not include the first information item, the element missing identifier represents that the initial reconciliation request does not include the first information item, and the element missing reason description represents the missing first information item in the initial reconciliation request;

[0034] Construct the clause missing prompt component according to the contract write-off exception analysis process, the element missing identifier, and the element missing reason description;

[0035] The second reconciliation prompt word is obtained through the following process, including:

[0036] Obtain the second contract clause specifications and second clause performance examples of each second information item;

[0037] Construct a third logical determination component according to the second contract clause specification and the second clause performance example, where the third logical determination component is configured to extract the second information item from the initial reconciliation request according to the second contract clause specification and the second clause performance example;

[0038] Construct a fourth logical determination component according to the write-off element priority rule for extracting the second information item; the write-off element priority rule includes the mandatory parsing principle of the amount element, the conflict handling principle of the time limit element, and the signature validity verification principle;

[0039] Construct the second reconciliation prompt word according to the third logical determination component and the fourth logical determination component.

[0040] In a possible implementation manner, the target procurement entity archive library includes a procurement contract clause knowledge graph, and the procurement contract clause knowledge graph includes a plurality of contract clause nodes;

[0041] The determination of a plurality of pending confirmation information corresponding to the second information item according to the second information item and the target procurement entity archive library includes:

[0042] For each second information item, query a first associated clause instance corresponding to the second information item among the plurality of contract clause nodes;

[0043] If multiple first associated clause instances are queried, add the multiple first associated clause instances to the procurement term feature library;

[0044] With reference to the procurement term feature library, perform contract element structured parsing on the second information item to obtain performance element atomic items;

[0045] Determine a plurality of the pending confirmation information corresponding to the performance element atomic items.

[0046] In a possible implementation manner, the procurement contract clause knowledge graph further includes a plurality of clause element constraint conditions corresponding to the contract clause nodes;

[0047] The determination of a plurality of the pending confirmation information corresponding to the performance element atomic items includes:

[0048] Query the procurement contract clause knowledge graph according to the performance element atomic items;

[0049] If a second associated clause instance corresponding to the atomic item of the performance element is obtained by querying the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions, then according to the multiple second clause element constraint conditions, the multiple to-be-confirmed information is constructed.

[0050] In a possible implementation manner, the step of if a second associated clause instance corresponding to the atomic item of the performance element is obtained by querying the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions, then according to the multiple second clause element constraint conditions, the multiple to-be-confirmed information is constructed, includes:

[0051] Determine that a second associated clause instance corresponding to the atomic item of the performance element is obtained by querying the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions;

[0052] Load the atomic item of the performance element and the fifth reconciliation prompt word corresponding to the atomic item of the performance element into the clause element constraint condition inference model, and obtain at least one supplementary clause element constraint condition inferred corresponding to the atomic item of the performance element;

[0053] Construct the multiple to-be-confirmed information according to the multiple second clause element constraint conditions and the at least one supplementary clause element constraint condition;

[0054] After querying the procurement contract clause knowledge graph according to the atomic item of the performance element, the method further includes:

[0055] If a second associated clause instance corresponding to the atomic item of the performance element is not obtained by querying the procurement contract clause knowledge graph, load the atomic item of the performance element and the fourth reconciliation prompt word corresponding to the atomic item of the performance element into the contract element compliance verification model;

[0056] If the contract element compliance verification model identifies that the atomic item of the performance element is not a standard contract atomic item, construct a contract element update instruction, where the contract element update instruction is used to indicate to update the atomic item of the performance element;

[0057] The fourth reconciliation prompt word is obtained by the following process, including:

[0058] According to the contract clause specification of the standard contract atomic item, construct a fifth logical determination component, where the fifth logical determination component is configured to determine whether the atomic item of the performance element is the standard contract atomic item according to the contract clause specification of the standard contract atomic item;

[0059] Construct a compliance verification rule determination component according to the terms compliance verification rules of the standard contract atomic items;

[0060] Construct the fourth reconciliation prompt word according to the fifth logic determination component and the compliance verification rule determination component.

[0061] In a possible implementation manner, the optimizing the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the multiple pending confirmation information to obtain a target reconciliation request includes:

[0062] Receive the required supplementary content for the first information item loaded by the current administrator and the confirmation instruction for selecting a target confirmation information from the multiple pending confirmation information;

[0063] Supplement the first information item in the initial reconciliation request according to the required supplementary content;

[0064] Optimize the second information item in the initial reconciliation request according to the target confirmation information, and use the initial reconciliation request after supplementing the first information item and optimizing the second information item as the target reconciliation request.

[0065] In a possible implementation manner, the optimizing the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the multiple pending confirmation information to obtain a target reconciliation request includes:

[0066] Obtain the first contract classification identifier and the first transaction scenario feature of the initial reconciliation request;

[0067] Obtain the set of past reconciliation requests of the current administrator, where the current administrator is the administrator who initiated the initial reconciliation request;

[0068] For each past reconciliation request in the set of past reconciliation requests, obtain the second contract classification identifier and the second transaction scenario feature of the past reconciliation request;

[0069] If the second contract classification identifier of a past reconciliation request in the set of past reconciliation requests matches the first contract classification identifier, and the second transaction scenario feature of the past reconciliation request matches the first transaction scenario feature, then obtain the first past information item and the second past information item in the past reconciliation request;

[0070] Construct the required supplementary content according to the first past information item, and construct the confirmation instruction according to the second past information item;

[0071] Optimize the initial reconciliation request according to the required supplementary content and the confirmation instruction to obtain the target reconciliation request.

[0072] In a second aspect, an embodiment of the present invention provides a server system, including a server, where the server is used to execute the method described in the first aspect.

[0073] Compared with the prior art, the beneficial effects provided by the present invention include: By using an intelligent reconciliation method and system for a procurement entity disclosed in the present invention, after obtaining the target procurement entity archive library and the initial reconciliation request, the request and the corresponding first and second reconciliation prompt words are respectively loaded into the corresponding models to obtain the missing first information items that are required and the ambiguous second information items in the initial request. Determine multiple pending confirmation information based on the second information item and the archive library, and then optimize the initial request according to the supplementary content of the first information item and the confirmation instruction for the pending confirmation information to obtain the target reconciliation request, and perform a query in the archive library accordingly to achieve intelligent and accurate reconciliation operations, improving the efficiency and quality of reconciliation. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without creative efforts.

[0075] Figure 1 It is a schematic flowchart of the steps of the intelligent reconciliation method for a procurement entity provided by an embodiment of the present invention;

[0076] Figure 2 It is a schematic block diagram of the structure of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0077] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0078] The following will describe the specific embodiments of the present invention in detail with reference to the drawings.

[0079] To solve the technical problems in the foregoing background art, Figure 1It is a schematic flowchart of the intelligent reconciliation method for the purchasing entity provided by the embodiments of the present disclosure. The intelligent reconciliation method for the purchasing entity will be introduced in detail below.

[0080] Step S201: Obtain the target purchasing entity archive library and an initial reconciliation request for querying the target purchasing entity archive library.

[0081] Step S202: Load the initial reconciliation request and a first reconciliation prompt word corresponding to the initial reconciliation request into a first information item requirement parsing model to obtain a first information item, where the first information item is a required information item missing in the initial reconciliation request.

[0082] Step S203: Load the initial reconciliation request and a second reconciliation prompt word corresponding to the initial reconciliation request into a second information item requirement parsing model to obtain a second information item, where the second information item is an information item with ambiguity in the initial reconciliation request.

[0083] Step S204: Determine multiple pending confirmation information corresponding to the second information item according to the second information item and the target purchasing entity archive library.

[0084] Step S205: Optimize the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the multiple pending confirmation information to obtain a target reconciliation request, and perform a query in the target purchasing entity archive library according to the target reconciliation request.

[0085] In an embodiment of the present invention, exemplarily, in today's digital business environment, enterprise procurement activities are frequent, involving a large amount of contract and transaction data, and the reconciliation work has become extremely complex. This intelligent reconciliation method for procurement entities can efficiently process the reconciliation process and improve the accuracy and efficiency of work with the help of advanced technical means. The following takes the server as the execution entity to give a detailed scenario example of each step of this intelligent reconciliation method. Suppose the server serves a large chain supermarket group, which has business dealings with many suppliers. The server stores relevant data of each procurement entity (i.e., suppliers), and these data are integrated to form a target procurement entity archive. This archive contains a vast amount of information, such as the detailed terms of procurement contracts, previous transaction records, delivery information, payment details, etc. One day, the staff responsible for procurement reconciliation in the group found that there were doubts about the accounts with a certain supplier and hoped to check the accounts of a recent batch of food purchases. So, the staff sent an initial reconciliation request to the server through the group's internal reconciliation system. This request might be expressed as: "Query the reconciliation information regarding recent food purchases with [supplier name]". After receiving this request, the server simultaneously obtains the corresponding target procurement entity archive, which here is all the procurement data records related to this supplier, and is ready to start the subsequent reconciliation processing flow. In the scenario of the above supermarket group, the generation of the first reconciliation prompt word has its specific process. The server first obtains the first contract clause specifications and the first clause performance instances corresponding to each possible missing first information item (such as the procurement contract number, specific procurement date range, etc.). For example, the first contract clause specification for the procurement contract number specifies that it should have a specific format, consisting of a combination of numbers and letters, and be prominently presented at the beginning of the contract; the first clause performance instance is the actual presentation style of this number in numerous past contracts. Based on these, the server constructs a second logical determination component to determine whether the initial reconciliation request contains a certain first information item. At the same time, a clause matching prompt component and a clause missing prompt component are constructed. The clause matching prompt component is configured to give a contract clause association prompt when the request contains a certain first information item, such as providing the specific clause location and related explanations of this information item in the contract; the clause missing prompt component gives a warning of missing contract elements when the request does not contain a certain first information item, such as prompting "The key procurement contract number is missing, which may affect accurate reconciliation". The server loads the initial reconciliation request "Query the reconciliation information regarding recent food purchases with [supplier name]" and the corresponding first reconciliation prompt word into the first information item requirement analysis model. After analysis and judgment by the model, it is found that the required procurement contract number information item is missing in this request, so the first information item output is "Procurement contract number". Similarly, taking the supermarket group as an example, the acquisition of the second reconciliation prompt word also follows a specific process. The server obtains the second contract clause specifications and the second clause performance instances of each possible ambiguous second information item (such as the specific time range of "recent", the specific category definition of "food", etc.).For example, regarding the definition of the "recent" time range, the second contract clause specification may stipulate that it is within 30 days from the contract signing date; the second clause performance instance shows the actual time span corresponding to the expression "recent" in past contracts. Based on this, the server constructs a third logical determination component for extracting the second information item from the initial reconciliation request. At the same time, a fourth logical determination component is constructed according to the write-off element priority rules for extracting the second information item (including the mandatory parsing principle of amount elements, the conflict handling principle of time elements, and the signature validity verification principle, etc.). For example, the mandatory parsing principle of amount elements requires that the ambiguous information related to the amount be clarified first; the conflict handling principle of time elements stipulates the handling method when there are conflicts in time definition. The server loads the initial reconciliation request and the second reconciliation prompt word into the second information item requirement analysis model. After model analysis, it is found that the expression "recent" is ambiguous because the specific time range is not clear, so the second information item output is "the time range of'recent'". Still taking the supermarket group as an example, the target procurement entity archive contains a knowledge graph of procurement contract terms, which has multiple contract clause nodes. For the second information item of "the time range of'recent'", the server queries the corresponding first associated clause instance among multiple contract clause nodes. Suppose the server queries multiple first associated clause instances regarding the definition of the "recent" time range. For example, some contracts stipulate that recent is within 15 days after the contract signing, and some stipulate it is within 30 days, etc. The server adds these instances to the procurement term feature library. Then, the server refers to the procurement term feature library and conducts a structural analysis of the contract elements for "the time range of'recent'", obtaining the atomic items of performance elements, such as the minimum units of time definition like "within 15 days" and "within 30 days". Then, according to the multiple clause element constraint conditions corresponding to the contract clause nodes in the procurement contract terms knowledge graph, if the atomic items of performance elements related to the "recent" time range correspond to multiple clause element constraint conditions, such as different restrictive conditions for the "recent" time range in different contract scenarios, the server constructs multiple pending confirmation messages, which may be "within 15 days (applicable to fresh food procurement scenario)" and "within 30 days (applicable to ambient temperature food procurement scenario)". In the supermarket group scenario, the staff responsible for reconciliation, as the current administrator, sees that the first information item is missing the procurement contract number in the system prompt, so fills in the required supplementary content, such as "[specific contract number]" in the system interface. At the same time, for multiple pending confirmation messages, such as "within 15 days (applicable to fresh food procurement scenario)" and "within 30 days (applicable to ambient temperature food procurement scenario)", the staff selects "within 30 days (applicable to ambient temperature food procurement scenario)" as the confirmation instruction according to the fact that this reconciliation involves the procurement of ambient temperature food.After the server receives this content, it supplements the first information item in the initial reconciliation request according to the required supplementary content, optimizes the second information item in the initial reconciliation request according to the target confirmation information, and uses the supplemented and optimized initial reconciliation request as the target reconciliation request, which may become "query the reconciliation information regarding the purchase of ambient temperature food within the last 30 days for the contract with the contract number [specific contract number] and [supplier name]". Then, based on this target reconciliation request, the server performs a query in the target procurement entity archive, and accurately filters out the reconciliation information that meets the conditions from the massive data and feeds it back to the staff. The server obtains the first contract classification identifier (such as the category identifier of "food procurement contract") and the first transaction scenario feature (such as the feature of "ambient temperature food procurement scenario") of the initial reconciliation request. At the same time, it obtains the set of past reconciliation requests of the staff (the current administrator) who initiated this initial reconciliation request. In the set of past reconciliation requests, for each past request, it obtains its second contract classification identifier and second transaction scenario feature. Suppose the server finds that the second contract classification identifier of one of the past reconciliation requests and the first contract classification identifier of the current request are both the category identifier of "food procurement contract", and the second transaction scenario feature and the first transaction scenario feature are both the feature of "ambient temperature food procurement scenario". Then, the server obtains the first past information item (such as the purchase contract number filled in at that time) and the second past information item (such as the confirmation information for the selection of the time range of "recent" at that time) in this past reconciliation request. It constructs the required supplementary content according to the first past information item, such as "[past contract number]"; it constructs the confirmation instruction according to the second past information item, such as selecting the same "within 30 days (applicable to the ambient temperature food procurement scenario)" as in the past. The server optimizes the initial reconciliation request based on these constructed contents to obtain the target reconciliation request, such as "query the reconciliation information regarding the purchase of ambient temperature food within the last 30 days for the contract with the contract number [past contract number] and [supplier name]", and performs a query in the target procurement entity archive to provide accurate reconciliation data for the staff.

[0086] In the embodiment of the present invention, before loading the initial reconciliation request and the first reconciliation prompt word corresponding to the initial reconciliation request into the first information item requirement parsing model to obtain the first information item, the embodiment of the present invention provides the following implementation manners.

[0087] According to the initial reconciliation request, query the target procurement entity archive to obtain the initial reconciliation result;

[0088] Load the initial reconciliation request, the initial reconciliation result, and the third reconciliation prompt word corresponding to the initial reconciliation request and the initial reconciliation result into the reconciliation complexity assessment model to obtain the inferred reconciliation complexity assessment result;

[0089] If the reconciliation complexity evaluation result indicates that the initial reconciliation request does not require complex reconciliation processing, terminate the query according to the initial reconciliation request.

[0090] In an embodiment of the present invention, by way of example, first, the server receives an initial reconciliation request sent by a staff member, such as "query the reconciliation information regarding the recent food purchases with [supplier name]", and then queries the target procurement entity archive according to this request. The server obtains the contract element extraction rule, for example, it is stipulated to extract key information such as the supplier name, procurement category, and time from the request as contract elements. According to this rule, contract elements such as "[supplier name]", "food", and "recent" are extracted from the initial reconciliation request. Next, the server counts the number of times the terms of these contract elements match in each associated transaction instance in the target procurement entity archive to determine the term coverage of the associated transaction instance with respect to the contract elements. At the same time, according to the number of associated transaction instances containing these contract elements, the term constraint intensity is determined. For example, if most of the associated transaction instances mention the relevant content of the supplier's food purchases, the term coverage and constraint intensity are relatively high. Based on these two indicators, the server determines the write-off candidate set as the initial reconciliation result, and may draw a preliminary conclusion that there are a relatively large number of recent food purchase transactions involving this supplier, and there are some transaction records whose time range needs to be further clarified. After that, the server obtains the first business rule preset condition for determining the complex reconciliation processing condition, such as it is stipulated that if the transaction involves a large amount, a long time span, or complex contract terms, complex reconciliation processing is required. Based on this, a first logical determination component is constructed to determine whether the initial reconciliation request requires complex reconciliation processing. At the same time, an execution strategy component and an archiving strategy component are constructed. The former configures the specific process of complex reconciliation processing, and the latter sets the operations when no complex reconciliation processing is required. Combining these components, a third reconciliation prompt word is generated. The server loads the initial reconciliation request, the initial reconciliation result, and the third reconciliation prompt word into the reconciliation complexity evaluation model. After analysis by the model, if the reconciliation complexity evaluation result shows that this initial reconciliation request does not require complex reconciliation processing, for example, this query only involves regular food purchases, and factors such as amount and time are within a simple range, the server terminates the query according to this initial reconciliation request, does not perform subsequent complex processes, and directly feeds back the existing results to the staff, thus improving the reconciliation efficiency.

[0091] In an embodiment of the present invention, the querying of the target procurement entity archive according to the initial reconciliation request to obtain the initial reconciliation result can be implemented through the following examples.

[0092] Obtain the contract element extraction rule;

[0093] Extract contract elements from the initial reconciliation request according to the contract element extraction rule;

[0094] Load the contract elements into the target procurement entity archiving library, and use the target query algorithm to obtain the initial reconciliation result.

[0095] In an embodiment of the present invention, exemplarily, the server receives an initial reconciliation request sent by the procurement department, with the content "Query the reconciliation information regarding the procurement of leisure snacks from the Huimei Food Supplier in October 2024". First, the server obtains the contract element extraction rules. These rules are preset. For example, it is stipulated to extract the supplier name, procurement time, and procurement category from the request as key contract elements. Then, the server parses the initial reconciliation request according to the contract element extraction rules. From "Query the reconciliation information regarding the procurement of leisure snacks from the Huimei Food Supplier in October 2024", it successfully extracts "Huimei Food Supplier" as the supplier name element, "October 2024" as the procurement time element, and "leisure snacks" as the procurement category element. After that, the server loads the extracted contract elements into the target procurement entity archiving library. This archiving library stores a large number of procurement contracts and transaction records between the supermarket and numerous suppliers. The server uses the target query algorithm to search in the archiving library. The target query algorithm will perform the retrieval according to a certain logic and order. For example, first locate all the records related to the Huimei Food Supplier based on the supplier name, then filter out the records in October 2024 from these records, and finally determine the content related to the procurement of leisure snacks in these records. After the retrieval by the target query algorithm, the server obtains the initial reconciliation result. For example, the initial reconciliation result shows that in October 2024, the supermarket had three procurement transactions of leisure snacks with the Huimei Food Supplier. Among them, one transaction amount was 5000 yuan, and the types of snacks purchased included potato chips, nuts, etc.; the other two transactions also had corresponding detailed information, such as delivery time, payment status, etc. These results provide basic data for subsequent further reconciliation operations.

[0096] In an embodiment of the present invention, the step of loading the contract elements into the target procurement entity archiving library and using the target query algorithm to obtain the initial reconciliation result can be implemented through the following example.

[0097] Determine the clause coverage of the associated transaction instance with respect to the contract element according to the number of times the clause in each associated transaction instance in the target procurement entity archiving library matches the contract element;

[0098] Determine the clause constraint intensity of the associated transaction instance with respect to the contract element according to the number of associated transaction instances containing the contract element in the target procurement entity archiving library;

[0099] Determine a write-off candidate set from each associated transaction instance in the target procurement entity's archive database according to the clause coverage and the clause constraint strength, as the initial reconciliation result.

[0100] In an embodiment of the present invention, by way of example, take the server of a large supermarket chain group processing a reconciliation request with the "Meiweiyuan Food Factory" as a supplier. The initial reconciliation request is "Query the reconciliation information regarding the purchase of biscuits from the Meiweiyuan Food Factory in the second half of 2024". The server has extracted contract elements such as "Meiweiyuan Food Factory", "the second half of 2024", and "biscuits" from it. The server loads these contract elements into the target procurement entity's archive database and starts calculating the clause coverage. There are numerous associated transaction instances in the archive database. For example, there are 100 transaction instances involving the Meiweiyuan Food Factory. For the contract element of "biscuits", the server traverses the clauses of each associated transaction instance. If the purchase-related clauses of "biscuits" are clearly mentioned in 80 associated transaction instances, and the details of these clauses are different, some only mention the category, and some also include specific biscuit brands, specifications, etc. The server assigns different weights according to the detail level and matching degree of the clauses. After calculation, the score corresponding to the number of clause matches of the "biscuits" contract element in these associated transaction instances is obtained, and then the clause coverage of the "biscuits" element is determined to be 80% (assuming the result after score conversion). Similarly, the clause coverages corresponding to "Meiweiyuan Food Factory" and "the second half of 2024" are calculated. Next, the server determines the clause constraint strength. For the "biscuits" contract element, in the target procurement entity's archive database, there are a total of 60 associated transaction instances that contain the purchase-related content of "biscuits", and the total number of associated transaction instances is 500. The server calculates the proportion of the number of associated transaction instances containing the "biscuits" element in the total number of instances, and obtains the clause constraint strength of the "biscuits" element as 12% (60÷500). In the same way, the clause constraint strengths of "Meiweiyuan Food Factory" and "the second half of 2024" are calculated. Finally, the server determines the write-off candidate set based on the clause coverage and the clause constraint strength. The server sets certain screening criteria. For example, associated transaction instances with a clause coverage reaching 60% and a clause constraint strength reaching 10% can enter the write-off candidate set. According to this standard, screening is performed among each associated transaction instance in the target procurement entity's archive database, and the associated transaction instances that meet the standard are determined as the write-off candidate set, which is the initial reconciliation result. For example, 30 associated transaction instances are selected. They detail the biscuit purchase transactions of the Meiweiyuan Food Factory in the second half of 2024, including information such as purchase quantity, price, delivery date, etc. These instances constitute the initial reconciliation result and provide a basis for further reconciliation in the future.

[0101] In an embodiment of the present invention, the third reconciliation prompt word is obtained through the following process and can be implemented through the following example.

[0102] Obtain the first business rule preset condition for the determination condition of complex reconciliation processing;

[0103] Construct a first logical determination component according to the first business rule preset condition, where the first logical determination component is used to configure whether the initial reconciliation request requires complex reconciliation processing according to the first business rule preset condition;

[0104] Construct an execution policy component and an archiving policy component, where the execution policy component configures the processing when the initial reconciliation request requires complex reconciliation processing, and the archiving policy component configures the processing when the initial reconciliation request does not require complex reconciliation processing;

[0105] Construct the third reconciliation prompt word according to the first logical determination component, the execution policy component and the archiving policy component.

[0106] In an embodiment of the present invention, by way of example, still taking the server of a large chain supermarket group as an example. When the server receives an initial reconciliation request such as "Query the reconciliation information regarding the purchase of imported fruits from [Supplier Name] in November 2024" sent by the purchasing department, it begins to obtain the third reconciliation prompt word. First, the server obtains the first business rule preset conditions for complex reconciliation processing determination conditions. For example, the preset conditions stipulate that when the purchase involves goods with multiple different tax rates, the purchase amount exceeds 500,000 yuan, or the goods involve multiple different origins, complex reconciliation processing is required. Then, based on these first business rule preset conditions, the server constructs the first logical determination component. This component analyzes the initial reconciliation request according to the above preset conditions. For example, for the request "Query the reconciliation information regarding the purchase of imported fruits from [Supplier Name] in November 2024", it will look in the target procurement entity archive to find out whether the purchase of imported fruits this time involves goods with multiple different tax rates, whether the purchase amount exceeds 500,000 yuan, and whether the goods involve multiple different origins, so as to determine whether the initial reconciliation request requires complex reconciliation processing. Then, the server constructs an execution strategy component and an archiving strategy component. The execution strategy component configures the specific processing methods when the initial reconciliation request requires complex reconciliation processing. For example, if it is determined that complex reconciliation is needed, the execution strategy component will arrange a dedicated financial team to conduct multiple rounds of detailed accounting, require the supplier to provide more detailed cost composition and pricing basis, and conduct sampling inspections on the purchased goods, etc. The archiving strategy component, on the other hand, configures the processing methods when the initial reconciliation request does not require complex reconciliation processing. For example, if it is determined that no complex reconciliation is needed, the archiving strategy component will directly organize and file the data related to this reconciliation according to the regular process for subsequent simple review. Finally, the server constructs the third reconciliation prompt word based on the first logical determination component, the execution strategy component, and the archiving strategy component. This prompt word combines the logics and processing methods of the above components and can clearly indicate how the subsequent process should proceed. For example, the third reconciliation prompt word may be expressed as: "If the purchase of imported fruits this time involves goods with multiple different tax rates, the purchase amount exceeds 500,000 yuan, or the goods involve multiple different origins, then start the complex reconciliation process, and a dedicated financial team will conduct detailed accounting and other operations; if the above situations are not involved, then organize and file according to the regular process." In this way, the third reconciliation prompt word provides clear guidance for subsequent judgment and processing of the initial reconciliation request.

[0107] In an embodiment of the present invention, the first reconciliation prompt word is obtained through the following process and can be implemented through the following example.

[0108] Obtain the first contract clause specifications and the first clause performance instances of each first information item;

[0109] Construct a second logical determination component according to the first contract clause specification and the first clause performance example, where the second logical determination component is used to determine whether the initial reconciliation request contains the first information item according to the first contract clause specification and the first clause performance example;

[0110] Construct a clause matching prompt component and a clause missing prompt component. The clause matching prompt component configures the contract clause association prompt when the initial reconciliation request contains the first information item, and the clause missing prompt component configures the warning of contract element missing when the initial reconciliation request does not contain the first information item;

[0111] Construct the first reconciliation prompt word according to the second logical determination component, the clause matching prompt component and the clause missing prompt component.

[0112] In an embodiment of the present invention, by way of example, continuing with the server of a large supermarket chain as an example, when the server receives an initial reconciliation request "Query the reconciliation information regarding recent food purchases with [Supplier Name]", it begins to obtain the first reconciliation prompt words. First, the server obtains the first contract clause specifications and the first clause performance instances of each first information item. For example, for the first information item "Purchase contract number" that may be missing, the first contract clause specification clearly stipulates that the purchase contract number should consist of 10 digits, with the first 4 digits representing the year and the last 6 digits being the sequential number, and it should be marked prominently in the upper left corner of the first page of the purchase contract. The first clause performance instances are the actual purchase contract numbers extracted from a large number of past purchase contracts, such as specific examples like "2023000001" and "2024000005". Next, based on these first contract clause specifications and the first clause performance instances, the server constructs a second logical determination component. This component is like a precise detector that can carefully analyze the initial reconciliation request according to the first contract clause specifications and the first clause performance instances to determine whether the request contains the first information item "Purchase contract number". For example, it will check whether there is a string that meets the 10 - digit number and specific format in the request text to determine whether the initial reconciliation request contains this information item. After that, the server constructs a clause matching prompt component and a clause missing prompt component. The clause matching prompt component is configured to provide contract clause association prompts when it detects that the initial reconciliation request contains the first information item "Purchase contract number". For example, it prompts "In the contract corresponding to the purchase contract number [specific number], Article 3 stipulates the commodity quality standards, and Article 5 clarifies the payment method", helping the staff quickly understand the key clauses of the contract with this number. The clause missing prompt component is configured to give a warning of missing contract elements when it detects that the initial reconciliation request does not contain the "Purchase contract number", such as "Your reconciliation request is missing the purchase contract number, which is a key element for accurate reconciliation. Please supplement it", reminding the staff to pay attention and supplement the missing information. Finally, the server constructs the first reconciliation prompt words based on the second logical determination component, the clause matching prompt component, and the clause missing prompt component. This first reconciliation prompt word integrates the functions of the above - mentioned components and may be expressed as: "If the request contains a purchase contract number consisting of 10 digits that meet the format (the first 4 digits represent the year and the last 6 digits are the sequential number), you can view the key clauses of the corresponding contract; if not, the key reconciliation element, the purchase contract number, is missing. Please supplement it." In this way, the first reconciliation prompt words provide clear and definite guidance for the staff to process the initial reconciliation request subsequently.

[0113] In an embodiment of the present invention, the construction of the clause matching prompt component and the clause missing prompt component can be implemented through the following examples.

[0114] Obtain the contract clause parsing process, clause matching status code, and clause performance description. The contract clause parsing process is the parsing process in which the initial reconciliation request includes the first information item. The clause matching status code indicates that the initial reconciliation request includes the first information item. The clause performance description is the text information in which the initial reconciliation request includes the first information item;

[0115] Construct the clause matching prompt component according to the contract clause parsing process, the clause matching status code, and the clause performance description;

[0116] Obtain the contract write-off exception analysis process, element missing identifier, and element missing reason description. The contract write-off exception analysis process is the analysis process in which the initial reconciliation request does not include the first information item. The element missing identifier indicates that the initial reconciliation request does not include the first information item. The element missing reason description indicates the missing first information item in the initial reconciliation request;

[0117] Construct the clause missing prompt component according to the contract write-off exception analysis process, the element missing identifier, and the element missing reason description.

[0118] In an embodiment of the present invention, by way of example, take the server of a large chain supermarket group processing a reconciliation request as an example. Suppose the initial reconciliation request is "Query the reconciliation information regarding the purchase of grain and oil this month from [supplier name]". At this time, the server constructs relevant components for the first information item "purchase contract number" that may be missing. First, obtain the content required to construct the clause matching prompt component. The contract clause parsing process stipulates that if the initial reconciliation request contains the "purchase contract number", the number needs to be extracted first, then the corresponding contract is located in the contract database based on the number, and then each clause of the contract is parsed. The clause matching status code is set to "01", indicating that the initial reconciliation request contains the "purchase contract number". The description of the clause performance situation is text information such as "The purchase contract number is used to trace the detailed content of the purchase contract, including performance information such as product specifications, prices, delivery times, etc.". Then, construct the clause matching prompt component based on this information. This component integrates the contract clause parsing process, the clause matching status code, and the description of the clause performance situation. When it detects that the initial reconciliation request contains the "purchase contract number", it will give a prompt such as "Status code 01: The purchase contract number has been identified. Please follow the parsing process, first extract the number to locate the contract, and then understand the performance information such as product specifications, prices, delivery times, etc.", guiding the subsequent operations of the staff. Then, obtain the information for constructing the clause missing prompt component. The contract cancellation exception analysis process shows that if the initial reconciliation request does not contain the "purchase contract number", the current reconciliation process needs to be suspended, the staff is notified to supplement the information, and the abnormal situation is recorded. The element missing identifier is set to "M001", representing that the initial reconciliation request does not contain the "purchase contract number". The reason for the missing element is described as "The current reconciliation request lacks the purchase contract number for accurately locating the contract and reconciling accounts". Finally, construct the clause missing prompt component based on this information. This component combines the contract cancellation exception analysis process, the element missing identifier, and the reason for the missing element. When it detects that the initial reconciliation request does not contain the "purchase contract number", it will prompt "Identifier M001: The purchase contract number is missing. Since this number is used to accurately locate the contract and reconcile accounts, the reconciliation is suspended. Please supplement the number and resend the request. The exception has been recorded", clearly informing the staff of the problem and the handling method. Through the above steps, the server completes the construction of the clause matching prompt component and the clause missing prompt component, providing a complete and clear guidance for processing the initial reconciliation request.

[0119] In an embodiment of the present invention, the second reconciliation prompt word is obtained through the following process and can be implemented through the following example.

[0120] Obtain the second contract clause specifications and second clause performance examples of each second information item;

[0121] Construct a third logical determination component according to the second contract clause specification and the second clause performance example, where the third logical determination component is used to extract the second information item from the initial reconciliation request according to the second contract clause specification and the second clause performance example;

[0122] Construct a fourth logical determination component according to the write-off element priority rule for extracting the second information item; the write-off element priority rule includes the mandatory parsing principle of the amount element, the conflict handling principle of the time limit element, and the signature validity verification principle;

[0123] Construct the second reconciliation prompt word according to the third logical determination component and the fourth logical determination component.

[0124] In an embodiment of the present invention, by way of example, still taking the server of a large chain supermarket group as an example, when the server receives an initial reconciliation request "Query the reconciliation information regarding the recent large - volume beverage procurement with [supplier name]", it begins to obtain the second reconciliation prompt word. First, the server obtains the second contract - term specifications and second - term performance instances of each second information item. For example, for the second information item "recent" which may be ambiguous, the second contract - term specification stipulates that "recent" generally refers to within 30 days from the contract - signing date, but for seasonal goods, the scope can be appropriately adjusted according to seasonal characteristics. The second - term performance instance is an actual application example of "recent" collected from past contracts. For example, in a certain beverage procurement contract, "recent" is clearly defined as within 20 days before the summer promotion event. Then, based on these second contract - term specifications and second - term performance instances, the server constructs a third logical determination component. This component can screen and analyze the initial reconciliation request according to the second contract - term specifications and second - term performance instances, and extract second information items such as "recent" from the request. For example, it will analyze the expression of "recent" in the request, and combine the contract - term specification and performance instance to judge its potential meaning in the current context. Then, the server constructs a fourth logical determination component according to the write - off - element priority rules for extracting second information items. The write - off - element priority rules include the mandatory parsing principle for amount elements, the conflict - handling principle for time - limit elements, and the signature - validity verification principle. Taking the second information item "recent" as an example, the mandatory parsing principle for amount elements requires that if the calculation of the beverage procurement amount is related to the time range of "recent", the accurate time range corresponding to the amount calculation must be clearly defined first; the conflict - handling principle for time - limit elements stipulates that if the time definition of "recent" conflicts with the time limit stipulated in other contracts, it shall be processed according to specific rules, such as taking the special time limit of seasonal goods as the standard; the signature - validity verification principle emphasizes that if there is a dispute over the interpretation of "recent", the final interpretation right shall be confirmed according to the signature on the contract. Finally, the server constructs the second reconciliation prompt word according to the third logical determination component and the fourth logical determination component. This second reconciliation prompt word combines the functions of the two components and may be expressed as: "Regarding the expression 'recent', extract and analyze it according to the contract - term specification and past performance instances. If it involves amount calculation, clearly define the corresponding time range first; if it conflicts with other time limits, handle it according to the special time limit of seasonal goods; if there is a dispute over the interpretation, confirm the interpretation right with the signature." In this way, the second reconciliation prompt word provides detailed and orderly processing guidelines for the staff when dealing with ambiguous information items in the initial reconciliation request.

[0125] In an embodiment of the present invention, the target procurement - entity archive contains a procurement - contract - term knowledge graph, and the procurement - contract - term knowledge graph contains multiple contract - term nodes;

[0126] Determining multiple pending confirmation information items corresponding to the second information item according to the second information item and the target procurement entity filing library can be implemented through the following examples.

[0127] For each of the second information items, query for the first associated clause instance corresponding to the second information item among the multiple contract clause nodes;

[0128] If multiple first associated clause instances are queried, add the multiple first associated clause instances to the procurement term feature library;

[0129] With reference to the procurement term feature library, perform contract element structured parsing on the second information item to obtain performance element atomic items;

[0130] Determine multiple pending confirmation information items corresponding to the performance element atomic items.

[0131] In an embodiment of the present invention, by way of example, taking the server of a large supermarket chain group processing the reconciliation business as an example, the target procurement entity archive library managed by the server contains a knowledge graph of procurement contract terms, which is filled with multiple contract term nodes. Suppose the initial reconciliation request received by the server is "Query the reconciliation information regarding a large amount of recent snack purchases from [supplier name]", and it is analyzed that "recent" is the second information item. First, for this second information item of "recent", the server conducts a query among the multiple contract term nodes in the knowledge graph of procurement contract terms to find the corresponding first associated clause instance. Since the knowledge graph covers numerous contract term information, during the query process, the server discovers multiple first associated clause instances related to "recent". For example, in some snack procurement contract terms, "recent" is defined as within 15 days after the contract is signed, for the emergency replenishment scenario; in some other contracts, "recent" refers to within one month, applicable to the regular promotion preparation stage. Because there are multiple first associated clause instances found through the query, the server adds these instances to the procurement term feature library. This procurement term feature library collects the definitions and application instances of various procurement terms in different contract scenarios. Then, the server refers to the procurement term feature library to conduct a contract element structured analysis on the second information item of "recent". Based on the different definitions and instances in the feature library, it disassembles the relatively broad expression of "recent" into more specific performance element atomic items. For example, it analyzes and obtains performance element atomic items such as "within 15 days (emergency replenishment scenario)" and "within one month (regular promotion preparation stage)". Finally, the server determines multiple pending confirmation information corresponding to these performance element atomic items. Based on the performance element atomic items analyzed previously, combined with factors such as the past transaction habits between the supermarket and the supplier, the current market situation, and the overall contract background, multiple pending confirmation information is determined. For example, it may determine pending confirmation information such as "within 15 days (considering the large fluctuations in the recent market demand for snacks, it may be for emergency replenishment)" and "within one month (based on the past regular promotion preparation cycle)", providing multiple possibilities for further clarifying the exact meaning of "recent" subsequently, so that the staff can make a final confirmation in combination with the actual situation.

[0132] In an embodiment of the present invention, the knowledge graph of procurement contract terms further includes multiple clause element constraint conditions corresponding to the contract term nodes;

[0133] The determination of the multiple pending confirmation information corresponding to the performance element atomic items can be implemented through the following example.

[0134] Query the knowledge graph of procurement contract terms according to the performance element atomic items;

[0135] If the purchase contract terms knowledge graph is queried to obtain a second associated terms instance corresponding to the performance element atomic item, and in the purchase contract terms knowledge graph, the second associated terms instance corresponds to a plurality of second terms element constraints, then the plurality of pending confirmation information is constructed based on the plurality of second terms element constraints.

[0136] In the embodiment of the present invention, illustratively, taking the server of a large supermarket chain group as an example, the knowledge graph of the purchase contract terms in the archive library of its target purchase subject not only has multiple contract terms nodes, but each node also corresponds to multiple terms element constraints. Still taking the above initial reconciliation request of "querying the reconciliation information with [supplier name] about the recent large-scale purchase of snacks" as an example, the server has parsed the "recent" performance element atomic items, such as "within 15 days (emergency replenishment scenario)" and "within one month (regular promotion preparation stage)". For the performance element atomic item "within 15 days (emergency replenishment scenario)", the server queries the purchase contract terms knowledge graph based on it. In the graph, the server obtains the second associated clause instance corresponding to the performance element atomic item. For example, a contract clause is found that stipulates that in the emergency replenishment scenario, if the purchase amount exceeds 50,000 yuan, the replenishment must be completed within 15 days, and the supplier must provide expedited delivery service; if the purchase amount is less than 50,000 yuan, although the replenishment must also be completed within 15 days, the delivery service is carried out according to the conventional process. Here, "whether the purchase amount exceeds 50,000 yuan" is a plurality of second clause element constraints. Due to the existence of these second-clause element constraints, the server constructs multiple pending confirmation information based on them. That is, if the purchase amount of the emergency replenishment of snacks with [supplier name] exceeds 50,000 yuan, the pending confirmation information is "replenishment will be completed within 15 days, and the supplier will provide expedited delivery service"; if the purchase amount is less than 50,000 yuan, the pending confirmation information is "replenishment will be completed within 15 days, according to the regular delivery process". Similarly, for the atomic item of the performance element "within one month (regular promotion preparation stage)", the server queries the corresponding second-related clause instance and related second-clause element constraints in the purchase contract clause knowledge graph. For example, in the regular promotion preparation stage, if the promotion covers all stores in the supermarket, the purchase must be completed in two batches within one month, and the first batch will arrive within 15 days; if it only covers some regional stores, the purchase can be completed at one time within one month. Based on these conditions, the server constructs the corresponding pending confirmation information. If the regular promotion activities corresponding to this snack purchase cover all stores in the supermarket, the pending confirmation information is "the purchase will be completed in two batches within one month, and the first batch will arrive within 15 days"; if it only covers some regional stores, the pending confirmation information is "the purchase will be completed in one go within one month." In this way, the server uses the constraints of the terms and conditions in the knowledge graph of the purchase contract terms to construct multiple pending confirmation information for each atomic item of the performance element, providing a more comprehensive and detailed reference for subsequent accurate reconciliation.

[0137] In an embodiment of the present invention, if a second associated clause instance corresponding to the performance element atomic item is obtained by querying the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions, then according to the multiple second clause element constraint conditions, the multiple to-be-confirmed information can be constructed, and the implementation can be executed through the following examples.

[0138] It is determined that a second associated clause instance corresponding to the performance element atomic item is obtained by querying the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions;

[0139] The performance element atomic item and the fifth reconciliation prompt word corresponding to the performance element atomic item are loaded into the clause element constraint condition inference model to obtain at least one supplementary clause element constraint condition inferred corresponding to the performance element atomic item;

[0140] According to the multiple second clause element constraint conditions and the at least one supplementary clause element constraint condition, the multiple to-be-confirmed information is constructed.

[0141] In an embodiment of the present invention, by way of example, taking the server of a large supermarket chain as an example, the server is processing a reconciliation request for snack purchases from [Supplier Name] and has obtained the atomic items of performance elements for the "recent period", such as "within 15 days (emergency replenishment scenario)". First, the server determines that querying the procurement contract clause knowledge graph has obtained a second associated clause instance corresponding to the atomic item of the performance element "within 15 days (emergency replenishment scenario)". For example, a contract clause regarding emergency replenishment is found, which stipulates that if the snacks purchased are perishable categories, the delivery must be completed within 15 days and cold chain transportation must be ensured; if they are ordinary categories, the regular transportation method shall be used. Here, "the category of snacks purchased (perishable or ordinary)" is a constraint condition of multiple second clause elements. Next, the server loads the atomic item of the performance element "within 15 days (emergency replenishment scenario)" and the corresponding fifth reconciliation prompt word into the clause element constraint condition inference model. The fifth reconciliation prompt word may contain guiding information such as "According to past emergency replenishment contracts, consider the impact of product characteristics on replenishment conditions". Through this model, the server obtains at least one inferred supplementary clause element constraint condition corresponding to the atomic item of the performance element. For example, based on past contract data and business logic, the model infers that if the weight of the snacks purchased exceeds a certain standard, additional loading and unloading equipment may be required, which is a supplementary clause element constraint condition. Finally, the server constructs multiple pending confirmation messages based on the existing multiple second clause element constraint conditions (such as the category of snacks purchased) and the at least one newly inferred supplementary clause element constraint condition (such as additional loading and unloading equipment is required if the weight of the snacks purchased exceeds a certain standard). If perishable snacks are purchased and the weight exceeds the standard, the pending confirmation message is "Replenishment to be completed within 15 days, ensure cold chain transportation and be equipped with additional loading and unloading equipment"; if ordinary snacks are purchased and the weight does not exceed the standard, the pending confirmation message is "Replenishment to be completed within 15 days, use the regular transportation method, no additional loading and unloading equipment is required". Similarly, for other atomic items of performance elements in the "recent period", such as "within one month (regular promotion preparation stage)", the server also operates according to this process. First, determine the corresponding second associated clause instance and multiple second clause element constraint conditions, such as the scale of the promotion activity will affect the procurement arrangement. Then input the atomic item of the performance element and the corresponding fifth reconciliation prompt word into the clause element constraint condition inference model to obtain supplementary clause element constraint conditions, such as inferring that if the procurement quantity reaches a certain value, the supplier needs to provide additional gifts. Finally, construct pending confirmation messages based on these conditions. If the scale of the promotion activity is large and the procurement quantity reaches the value, the pending confirmation message is "Procurement to be completed within one month, the supplier provides additional gifts"; if the scale of the promotion activity is small and the procurement quantity does not reach the value, the pending confirmation message is "Procurement to be completed within one month, no additional gifts". Through such detailed steps, the server provides comprehensive and accurate pending confirmation messages for accurate reconciliation.

[0142] In an embodiment of the present invention, after querying the procurement contract clause knowledge graph according to the atomic item of the performance element, the method further includes:

[0143] If no second associated clause instance corresponding to the atomic item of the performance element is obtained by querying the procurement contract clause knowledge graph, load the atomic item of the performance element and the fourth reconciliation prompt word corresponding to the atomic item of the performance element into the contract element compliance verification model;

[0144] If the contract element compliance verification model identifies that the atomic item of the performance element is not a standard contract atomic item, construct a contract element update instruction for indicating to update the atomic item of the performance element.

[0145] In an embodiment of the present invention, by way of example, still taking the server of a large supermarket chain group processing the snack procurement reconciliation request with a supplier as an example. When the server processes the second information item of "recently", it has parsed out the atomic item of the performance element such as "(in the special event preparation scenario) within 15 days". The server queries the procurement contract clause knowledge graph according to this atomic item of the performance element in accordance with the process. However, this time no second associated clause instance corresponding to "(in the special event preparation scenario) within 15 days" is obtained in the knowledge graph. Therefore, the server loads the atomic item of the performance element "(in the special event preparation scenario) within 15 days" and the corresponding fourth reconciliation prompt word into the contract element compliance verification model. The fourth reconciliation prompt word may include content similar to "Check whether the atomic item of the performance element complies with the standard specification of the supermarket procurement contract, paying attention to aspects such as time definition and activity scenario description" to guide the model to perform targeted verification. The contract element compliance verification model starts to work and conducts a comprehensive analysis of "(in the special event preparation scenario) within 15 days". During the analysis process, based on the preset standard contract atomic item specification, the model identifies that the description of "special event preparation scenario" is not clear, does not meet the requirements of the standard contract atomic item, and does not belong to the standard contract atomic item. Based on this identification result, the server constructs a contract element update instruction. For example, the contract element update instruction may be expressed as "The description of'special event preparation scenario' in '(in the special event preparation scenario) within 15 days' is vague and does not meet the requirements of the standard contract atomic item. Please clarify key information such as the activity type and purpose to update the atomic item of the performance element". This instruction clearly informs the relevant staff that they need to modify the atomic item of the performance element to make it comply with the standard contract atomic item specification, so as to perform subsequent reconciliation and related business operations more accurately, ensure the standardization and consistency of the procurement contract, and avoid potential risks and disputes caused by unclear contract elements. Through such a rigorous process, the server can timely discover and handle the non-standard problems of contract elements that occur during the reconciliation process, ensuring the smooth progress of the procurement business and the accuracy of financial data.

[0146] In the embodiment of the present invention, the fourth reconciliation prompt word is obtained through the following process and can be implemented through the following examples.

[0147] According to the contract clause specification of the standard contract atomic item, a fifth logic determination component is constructed, and the fifth logic determination component is used to configure to determine whether the performance element atomic item is the standard contract atomic item according to the contract clause specification of the standard contract atomic item;

[0148] According to the clause compliance verification rule of the standard contract atomic item, a compliance verification rule determination component is constructed;

[0149] According to the fifth logic determination component and the compliance verification rule determination component, the fourth reconciliation prompt word is constructed.

[0150] In an embodiment of the present invention, by way of example, take the server of a large chain supermarket group handling affairs related to snack purchase reconciliation as an example. During the processing, for a performance element atomic item such as "(in the scenario of special event preparation) within 15 days", the server starts to obtain the fourth reconciliation prompt word. First, the server constructs a fifth logical determination component according to the contract clause specifications of the standard contract atomic item. For example, the standard contract atomic item stipulates that for time-related descriptions, it is necessary to be precise to the specific activity type, such as "(in the scenario of Spring Festival promotion event preparation) within 15 days". The fifth logical determination component judges whether "(in the scenario of special event preparation) within 15 days" conforms to the standard contract atomic item according to this specification. It will analyze whether the expression of "special event preparation scenario" is clear and whether it follows the requirements of the standard contract atomic item for the description of the activity scenario. Then, the server constructs a compliance verification rule determination component according to the clause compliance verification rules of the standard contract atomic item. Suppose the clause compliance verification rules stipulate that the description of the activity scenario needs to be closely related to common supermarket promotion, procurement and other business activities, and there should be no ambiguous expressions. The compliance verification rule determination component will further examine whether "(in the scenario of special event preparation) within 15 days" is compliant according to these rules. It will check whether the "special event preparation scenario" is related to supermarket business and whether there is any ambiguity. Finally, the server constructs the fourth reconciliation prompt word according to the fifth logical determination component and the compliance verification rule determination component. If the fifth logical determination component finds that the expression of "special event preparation scenario" is not clear and does not conform to the standard contract atomic item, and the compliance verification rule determination component also determines that it does not conform to the clause compliance verification rules, the fourth reconciliation prompt word may be "The '(in the scenario of special event preparation) within 15 days' you entered does not conform to the specifications of the standard contract atomic item. The standard requires that the description of the activity scenario be precise and closely related to supermarket business. The expression of'special event preparation scenario' is ambiguous, which may lead to deviations in contract understanding and execution. Please re-clarify key information such as the activity type." Through such a construction process, the fourth reconciliation prompt word can clearly inform relevant personnel of the problems existing in the performance element atomic item, guide them to make corrections according to the requirements of the standard contract atomic item, ensure the compliance of contract elements, and lay a foundation for subsequent accurate reconciliation and business development.

[0151] In an embodiment of the present invention, the optimizing the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the multiple pending confirmation information to obtain the target reconciliation request can be implemented through the following examples.

[0152] Receiving the required supplementary content for the first information item loaded by the current administrator and the confirmation instruction for selecting a target confirmation information from the multiple pending confirmation information;

[0153] Supplementing the first information item in the initial reconciliation request according to the required supplementary content;

[0154] Optimize the second information item in the initial reconciliation request according to the target confirmation information, and use the initial reconciliation request after supplementing the first information item and optimizing the second information item as the target reconciliation request.

[0155] In an embodiment of the present invention, by way of example, still taking a large chain supermarket group as an example, the server receives an initial reconciliation request "Query the reconciliation information regarding a large amount of snack purchases recently with [supplier name]". After processing, the server determines that the first information item such as "purchase contract number" is missing, and at the same time, multiple pending confirmation information is obtained for "recently", such as "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)", "within one month (regular promotion preparation stage, promotion covering the entire supermarket, purchased in two batches)", etc. At this time, the current administrator (the staff responsible for procurement reconciliation in the supermarket) performs relevant operations on the system operation interface. The administrator enters the required supplementary content, such as "SC202411001", at the position where the system prompts the missing "purchase contract number". At the same time, according to the actual situation of this procurement business, the administrator selects "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)" as the target confirmation information and issues a confirmation instruction. The server receives the required supplementary content "SC202411001" for the "purchase contract number" loaded by the administrator, and the confirmation instruction selecting "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)". Then, the server supplements the first information item "purchase contract number" in the initial reconciliation request according to the required supplementary content "SC202411001". Then, the server optimizes the second information item "recently" in the initial reconciliation request according to the target confirmation information "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)". Originally, the expression "recently" was ambiguous, and now it is clarified as "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)". Finally, the server uses the initial reconciliation request after supplementing the "purchase contract number" as "SC202411001" and optimizing "recently" as "within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required)" as the target reconciliation request. At this time, the target reconciliation request becomes "Query the reconciliation information regarding a large amount of snack purchases within 15 days (emergency replenishment scenario, purchase amount exceeding 50,000, urgent delivery required) with [supplier name] for the contract number SC202411001". The optimized target reconciliation request in this way is more accurate and detailed, and the server can perform accurate queries in the target procurement entity archive library based on this, providing a reconciliation result that better meets the needs of the administrator.

[0156] In an embodiment of the present invention, the initial reconciliation request is optimized according to the required supplementary content for the first information item and the confirmation instruction for the multiple pending confirmation information to obtain a target reconciliation request, which can be implemented through the following examples.

[0157] Obtain the first contract classification identifier and the first transaction scenario feature of the initial reconciliation request;

[0158] Obtain the set of past reconciliation requests of the current administrator, where the current administrator is the administrator who initiated the initial reconciliation request;

[0159] For each past reconciliation request in the set of past reconciliation requests, obtain the second contract classification identifier and the second transaction scenario feature of the past reconciliation request;

[0160] If the second contract classification identifier of a past reconciliation request in the set of past reconciliation requests matches the first contract classification identifier, and the second transaction scenario feature of the past reconciliation request matches the first transaction scenario feature, then obtain the first past information item and the second past information item in the past reconciliation request;

[0161] Construct the required supplementary content according to the first past information item, and construct the confirmation instruction according to the second past information item;

[0162] Optimize the initial reconciliation request according to the required supplementary content and the confirmation instruction to obtain the target reconciliation request.

[0163] In an embodiment of the present invention, by way of example, taking a large chain supermarket group as an example, the server receives an initial reconciliation request initiated by an administrator: "Query the reconciliation information regarding the recent purchase of a batch of daily necessities from [Supplier A]". First, the server obtains the first contract classification identifier and the first transaction scenario characteristics of the initial reconciliation request. After analysis, the first contract classification identifier is determined to be "Daily Necessities Purchase Contract", and the first transaction scenario characteristics are identified as "Regular bulk purchase, recent time range". Next, the server obtains the set of past reconciliation requests of the administrator who initiated this request currently. This administrator is responsible for the reconciliation work of the supermarket's daily necessities purchases, and there are many relevant reconciliation request records in the past. Then, for each past reconciliation request in the set of past reconciliation requests, the server obtains its second contract classification identifier and second transaction scenario characteristics. For example, among the past reconciliation requests, there is one that is "Query the reconciliation information regarding the purchase of a batch of shampoo from [Supplier A] within one month", its second contract classification identifier is "Daily Necessities Purchase Contract", and the second transaction scenario characteristics are "Regular bulk purchase, within one month time range". At this time, the server discovers that the second contract classification identifier of this past reconciliation request matches the first contract classification identifier "Daily Necessities Purchase Contract" of the current initial reconciliation request, and the second transaction scenario characteristics "Regular bulk purchase, within one month time range" have a relatively high degree of match with the first transaction scenario characteristics "Regular bulk purchase, recent time range" (here "recent" and "within one month" are considered to match in this business scenario). Therefore, the server obtains the first past information item and the second past information item in this past reconciliation request. Assume that the first past information item is "Purchase contract number: RJ20241005", and the second past information item is "Within one month (regular purchase scenario)". Next, the server constructs the required supplementary content based on the first past information item "Purchase contract number: RJ20241005", and constructs a confirmation instruction based on the second past information item "Within one month (regular purchase scenario)". Finally, the server optimizes the initial reconciliation request based on the constructed required supplementary content and confirmation instruction. The "Purchase contract number: RJ20241005" is supplemented to the position of the missing first information item (purchase contract number) in the initial reconciliation request, and the second information item "recent" is optimized to "Within one month (regular purchase scenario)". In this way, the optimized initial reconciliation request becomes the target reconciliation request: "Query the reconciliation information regarding the purchase of a batch of daily necessities from [Supplier A] within one month (regular purchase scenario) with the contract number RJ20241005". In this way, the server optimizes the initial reconciliation request by leveraging the administrator's past reconciliation experience, making it more accurate, so as to conduct a more effective query in the target procurement entity archive in the future and provide a more accurate reconciliation result for the administrator.

[0164] An embodiment of the present invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 executes the aforementioned intelligent reconciliation method for the procurement entity. As Figure 2 shown, Figure 2 is a structural block diagram of the computer device 100 provided by an embodiment of the present invention. The computer device 100 includes a memory 111, a processor 112, and a communication unit 113. To achieve data transmission or interaction, the elements of the memory 111, the processor 112, and the communication unit 113 are electrically connected to each other directly or indirectly. For example, these elements can be electrically connected to each other through one or more communication buses or signal lines.

[0165] For illustrative purposes, the foregoing description has been made with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Numerous modifications and variations are possible in light of the above teachings. The embodiments were chosen and described in order to best illustrate the principles of the disclosure and its practical application, thereby enabling those skilled in the art to best utilize the disclosure and utilize various embodiments with different modifications to suit the particular applications contemplated.

Claims

1. An intelligent reconciliation method for procurement entities, characterized in that, Including: Obtain a target procurement entity archive library and an initial reconciliation request for querying the target procurement entity archive library; Load the initial reconciliation request and a first reconciliation prompt word corresponding to the initial reconciliation request into a first information item requirement parsing model to obtain a first information item, where the first information item is a required information item missing in the initial reconciliation request; Load the initial reconciliation request and a second reconciliation prompt word corresponding to the initial reconciliation request into a second information item requirement parsing model to obtain a second information item, where the second information item is an ambiguous information item in the initial reconciliation request; Determine a plurality of pending confirmation information corresponding to the second information item according to the second information item and the target procurement entity archive library; Optimize the initial reconciliation request according to the required supplementary content for the first information item and the confirmation instruction for the plurality of pending confirmation information to obtain a target reconciliation request, and perform a query in the target procurement entity archive library according to the target reconciliation request.

2. The method according to claim 1, wherein Before loading the initial reconciliation request and a first reconciliation prompt word corresponding to the initial reconciliation request into a first information item requirement parsing model to obtain a first information item, the method further includes: Query the target procurement entity archive library according to the initial reconciliation request to obtain an initial reconciliation result; Load the initial reconciliation request, the initial reconciliation result, and a third reconciliation prompt word corresponding to the initial reconciliation request and the initial reconciliation result into a reconciliation complexity assessment model to obtain an inferred reconciliation complexity assessment result; If the reconciliation complexity assessment result indicates that the initial reconciliation request does not require complex reconciliation processing, terminate the query according to the initial reconciliation request.

3. The method according to claim 2, characterized in that, The querying the target procurement entity archive library according to the initial reconciliation request to obtain an initial reconciliation result includes: Obtain a contract element extraction rule; Extract contract elements from the initial reconciliation request according to the contract element extraction rule; Determine the clause coverage of the associated transaction instance with respect to the contract element according to the number of times the clause in the associated transaction instance in the target procurement entity archive library matches the contract element; Determine the clause constraint strength of the associated transaction instance with respect to the contract element according to the number of associated transaction instances containing the contract element in the target procurement entity archive library; Determine a write-off candidate set in each associated transaction instance in the target procurement entity archive library according to the clause coverage and the clause constraint strength as the initial reconciliation result.

4. The method according to claim 2, wherein The third reconciliation prompt word is obtained by the following process, including: Obtain a first business rule preset condition for a complex reconciliation processing determination condition; Construct a first logical determination component according to the first business rule preset condition, where the first logical determination component is used to configure whether the initial reconciliation request requires complex reconciliation processing according to the first business rule preset condition; Construct an execution policy component and an archiving policy component. The execution policy component configures the processing when the initial reconciliation request requires complex reconciliation processing, and the archiving policy component configures the processing when the initial reconciliation request does not require complex reconciliation processing; Construct the third reconciliation prompt word according to the first logic determination component, the execution policy component, and the archiving policy component; The first reconciliation prompt word is obtained through the following process, including: Obtain the first contract clause specification and the first clause performance instance of each first information item; Construct a second logic determination component according to the first contract clause specification and the first clause performance instance. The second logic determination component is used to configure to determine whether the initial reconciliation request includes the first information item according to the first contract clause specification and the first clause performance instance; Construct a clause matching prompt component and a clause missing prompt component. The clause matching prompt component configures the contract clause association prompt when the initial reconciliation request includes the first information item, and the clause missing prompt component configures the warning of contract element missing when the initial reconciliation request does not include the first information item; Construct the first reconciliation prompt word according to the second logic determination component, the clause matching prompt component, and the clause missing prompt component; The construction of the clause matching prompt component and the clause missing prompt component includes: Obtain the contract clause parsing process, the clause matching status code, and the clause performance situation description. The contract clause parsing process is the parsing process when the initial reconciliation request includes the first information item, the clause matching status code represents that the initial reconciliation request includes the first information item, and the clause performance situation description is the text information when the initial reconciliation request includes the first information item; Construct the clause matching prompt component according to the contract clause parsing process, the clause matching status code, and the clause performance situation description; Obtain the contract cancellation exception analysis process, the element missing identifier, and the element missing reason description. The contract cancellation exception analysis process is the analysis process when the initial reconciliation request does not include the first information item, the element missing identifier represents that the initial reconciliation request does not include the first information item, and the element missing reason description represents the missing first information item in the initial reconciliation request; Construct the clause missing prompt component according to the contract cancellation exception analysis process, the element missing identifier, and the element missing reason description; The second reconciliation prompt word is obtained through the following process, including: Obtain the second contract clause specification and the second clause performance instance of each second information item; Construct a third logic determination component according to the second contract clause specification and the second clause performance instance. The third logic determination component is used to configure to extract the second information item from the initial reconciliation request according to the second contract clause specification and the second clause performance instance; Construct a fourth logical determination component according to the write-off element priority rule for extracting the second information item; the write-off element priority rule includes a mandatory parsing principle for amount elements, a conflict handling principle for time limit elements, and a signature validity verification principle; Construct the second reconciliation prompt word according to the third logical determination component and the fourth logical determination component.

5. The method according to claim 1, wherein The target procurement entity archive contains a procurement contract clause knowledge graph, and the procurement contract clause knowledge graph contains multiple contract clause nodes; Determine multiple pending confirmation information items corresponding to the second information item according to the second information item and the target procurement entity archive, including: For each of the second information items, query for a first associated clause instance corresponding to the second information item among the multiple contract clause nodes; If multiple first associated clause instances are retrieved, add the multiple first associated clause instances to the procurement term feature library; With reference to the procurement term feature library, perform contract element structured parsing on the second information item to obtain performance element atomic items; Determine multiple pending confirmation information items corresponding to the performance element atomic items.

6. The method according to claim 5, characterized in that, The procurement contract clause knowledge graph further includes multiple clause element constraint conditions corresponding to the contract clause nodes; The determining of multiple pending confirmation information items corresponding to the performance element atomic items includes: Query the procurement contract clause knowledge graph according to the performance element atomic items; If a second associated clause instance corresponding to the performance element atomic item is retrieved from the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions, then construct the multiple pending confirmation information items according to the multiple second clause element constraint conditions.

7. The method according to claim 6, characterized in that, The if a second associated clause instance corresponding to the performance element atomic item is retrieved from the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions, then construct the multiple pending confirmation information items according to the multiple second clause element constraint conditions includes: Determine that a second associated clause instance corresponding to the performance element atomic item is retrieved from the procurement contract clause knowledge graph, and in the procurement contract clause knowledge graph, the second associated clause instance corresponds to multiple second clause element constraint conditions; Load the performance element atomic item and a fifth reconciliation prompt word corresponding to the performance element atomic item into a clause element constraint condition inference model to obtain at least one inferred supplementary clause element constraint condition corresponding to the performance element atomic item; Construct the multiple pending confirmation information items according to the multiple second clause element constraint conditions and the at least one supplementary clause element constraint condition; After querying the procurement contract clause knowledge graph according to the performance element atomic items, the method further includes: If the second associated clause instance corresponding to the performance element atomic item is not obtained by querying the purchase contract clause knowledge graph, the performance element atomic item and the fourth reconciliation prompt word corresponding to the performance element atomic item are loaded into the contract element compliance verification model; If the contract element compliance verification model identifies that the performance element atomic item is not a standard contract atomic item, construct a contract element update instruction, where the contract element update instruction is used to instruct to update the performance element atomic item; The fourth reconciliation prompt word is obtained by the following process, including: According to the contract clause specification of the standard contract atomic item, a fifth logic determination component is constructed, wherein the fifth logic determination component is configured to determine whether the performance element atomic item is the standard contract atomic item according to the contract clause specification of the standard contract atomic item; According to the compliance verification rules of the atomic items of the standard contract, a compliance verification rule determination component is constructed; The fourth reconciliation prompt word is constructed according to the fifth logic determination component and the compliance verification rule determination component.

8. The method according to claim 1, wherein The step of optimizing the initial reconciliation request according to the required supplementary content of the first information item and the confirmation instructions of the plurality of pending confirmation information to obtain the target reconciliation request includes: receiving the mandatory supplementary content for the first information item loaded by the current administrator and the confirmation instruction for selecting target confirmation information from the plurality of pending confirmation information; Supplement the first information item in the initial reconciliation request according to the required supplementary content; According to the target confirmation information, the second information item in the initial reconciliation request is optimized, and the initial reconciliation request after the first information item is supplemented and the second information item is optimized is used as the target reconciliation request.

9. The method according to claim 1, wherein The step of optimizing the initial reconciliation request according to the required supplementary content of the first information item and the confirmation instructions of the plurality of pending confirmation information to obtain the target reconciliation request includes: Acquire a first contract classification identifier and a first transaction scenario feature of the initial reconciliation request; Obtaining a set of past reconciliation requests of a current administrator, where the current administrator is the administrator who initiated the initial reconciliation request; For each past reconciliation request in the past reconciliation request set, obtaining a second contract classification identifier and a second transaction scenario feature of the past reconciliation request; If the second contract classification identifier of a past reconciliation request in the past reconciliation request set matches the first contract classification identifier, and the second transaction scenario feature of the past reconciliation request matches the first transaction scenario feature, obtaining the first past information item and the second past information item in the past reconciliation request; constructing the required supplementary content according to the first past information item, and constructing the confirmation instruction according to the second past information item; According to the required supplementary content and the confirmation instruction, the initial reconciliation request is optimized to obtain the target reconciliation request.

10. A server system, characterized in that, The method comprises a server, wherein the server is used to execute the method described in any one of claims 1 to 9.

Images