Model-based reconciliation method and apparatus

Abstract

This specification provides one or more embodiments of a model-based reconciliation method and apparatus. The method includes acquiring reconciliation request information input by a user; determining reconciliation parameters based on the reconciliation request information; after determining the reconciliation parameters, acquiring target reconciliation data corresponding to the reconciliation parameters; extracting target attribute tags related to the reconciliation parameters from the target reconciliation data and determining whether the reconciliation parameters match the target attribute tags; if the reconciliation parameters match the target attribute tags, acquiring target data features of the target reconciliation data and determining whether the reconciliation parameters match the target data features; when the reconciliation parameters match the target data features, generating a first prompt word based on the reconciliation parameters and the target reconciliation data and inputting it into a trained reconciliation model to obtain a reconciliation result matching the reconciliation request information.

Description

Technical Field

[0001] This specification relates to one or more embodiments in the field of model-based reconciliation technology, and more particularly to a model-based reconciliation method and apparatus. Background Technology

[0002] As industry specialization becomes increasingly refined, many business scenarios require multi-party collaboration. For example, credit reporting involves multiple participants, including financial institutions, credit reporting platforms, data processors, and data source providers. Each participant needs to conduct regular reconciliations to ensure the accuracy and consistency of their data.

[0003] Currently, in related technologies, reconciliation is mainly done manually. However, the reconciliation process itself is quite tedious, and the complexity of reconciliation increases with the volume of business, resulting in low overall efficiency of manual reconciliation. Summary of the Invention

[0004] In view of this, one or more embodiments of this specification provide a model-based reconciliation method and apparatus.

[0005] To achieve the above objectives, one or more embodiments of this specification provide the following technical solutions: According to a first aspect of one or more embodiments of this specification, a model-based reconciliation method is proposed, comprising: Obtain the reconciliation request information input by the user, and determine the reconciliation parameters based on the reconciliation request information; Obtain the target reconciliation data corresponding to the reconciliation parameters, extract the target attribute tags related to the reconciliation parameters from the target reconciliation data, and determine whether the reconciliation parameters match the target attribute tags; If it is determined that the reconciliation parameters match the target attribute tags, the target data features of the target reconciliation data are obtained, and it is determined whether the reconciliation parameters match the target data features. If the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

[0006] According to a second aspect of one or more embodiments of this specification, a model-based reconciliation apparatus is provided, comprising: The acquisition module acquires the reconciliation request information input by the user and determines the reconciliation parameters based on the reconciliation request information; The first determining module acquires target reconciliation data corresponding to the reconciliation parameters, extracts target attribute tags related to the reconciliation parameters from the target reconciliation data, and determines whether the reconciliation parameters match the target attribute tags. The second determining module, when it is determined that the reconciliation parameters match the target attribute tags, obtains the target data features of the target reconciliation data and determines whether the reconciliation parameters match the target data features; The reconciliation module, upon determining that the reconciliation parameters match the target data features, generates a first prompt word based on the reconciliation parameters and the target reconciliation data, and inputs it into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

[0007] According to a third aspect of one or more embodiments of this specification, an electronic device is provided, comprising: processor; Memory used to store processor-executable instructions; The processor implements the method as described in the first aspect by running the executable instructions.

[0008] According to a fourth aspect of one or more embodiments of this specification, a computer-readable storage medium is provided that stores computer instructions thereon, which, when executed by a processor, implement the steps of the method as described in the first aspect.

[0009] According to a fifth aspect of one or more embodiments of this specification, a computer program product is provided, comprising: a computer program / instructions that, when executed by a processor, implement the method as described in the first aspect.

[0010] As can be seen from the above embodiments, the model-based reconciliation method and apparatus provided in one or more embodiments of this specification first obtains the reconciliation request information input by the user, then determines the reconciliation parameters based on the reconciliation request information, and after determining the reconciliation parameters, obtains the target reconciliation data corresponding to the reconciliation parameters, extracts the target attribute tags related to the reconciliation parameters from the target reconciliation data, and determines whether the reconciliation parameters match the target attribute tags. If it is determined that the reconciliation parameters match the target attribute tags, obtains the target data features of the target reconciliation data, and determines whether the reconciliation parameters match the target data features. Through the dual verification of the target attribute tags and target data features of the target reconciliation data, the accuracy of the target reconciliation data to be reconciled can be further guaranteed. When it is determined that the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information. By processing the reconciliation parameters and target reconciliation data through the reconciliation model, the overall efficiency of reconciliation can be significantly improved. Moreover, for users, they only need to input the reconciliation requirement information to complete the entire reconciliation process. This not only enables users to actively configure the reconciliation parameters but also further improves the automation level of the reconciliation work. Attached Figure Description

[0011] Figure 1 This is an exemplary embodiment of the architecture diagram of an application scenario for a model-based reconciliation method.

[0012] Figure 2 This is a schematic diagram of a model-based reconciliation method provided in an exemplary embodiment.

[0013] Figure 3 This is a schematic diagram of the structure of an interface for users to input reconciliation request information, provided as an exemplary embodiment.

[0014] Figure 4 This is a flowchart illustrating another model-based reconciliation method provided in an exemplary embodiment.

[0015] Figure 5 This is a schematic diagram of the structure of a device provided in an exemplary embodiment.

[0016] Figure 6 This is a block diagram of a model-based reconciliation device provided in an exemplary embodiment. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of related data shall comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals shall be provided for users to choose to authorize or refuse.

[0019] As described in the background section, reconciliation is a crucial step in ensuring data consistency and verifying the authenticity of business transactions across various business scenarios. In related technologies, reconciliation primarily relies on manual operation, requiring manual verification of transaction vouchers, account records, and business data item by item, identification of discrepancies, and confirmation of results. However, reconciliation is inherently cumbersome and highly repetitive, with its complexity directly proportional to the volume of business. Therefore, manual reconciliation is generally inefficient and fails to meet the demands for high-efficiency reconciliation in large-scale business scenarios.

[0020] Furthermore, in related technologies, users often need to perform different reconciliation operations to meet various reconciliation requirements based on different business needs. For example, even for the same user, the reconciliation parameters, such as the reconciliation subject, reconciliation period, and reconciliation scope, will vary each time a reconciliation is performed. Therefore, each time a reconciliation is conducted, it is necessary to determine the reconciliation parameters in advance and accurately obtain the data to be reconciled based on these parameters. This process itself requires a significant amount of manpower and time.

[0021] In summary, to address the technical problem of low reconciliation efficiency in related technologies, this specification provides a model-based reconciliation method. First, it acquires the reconciliation request information input by the user. Then, it determines reconciliation parameters based on this request information, allowing the user to accurately control the current reconciliation parameters according to their input, thus facilitating automatic reconciliation in various business scenarios. After determining the reconciliation parameters, to further improve efficiency, target reconciliation data corresponding to the parameters can be automatically acquired, eliminating the need for manual preparation of the data to be reconciled. Simultaneously, to avoid deviations in the automatically acquired target reconciliation data, after acquiring the target data, target attribute tags related to the reconciliation parameters are extracted from the target reconciliation data, and it is determined whether the reconciliation parameters match the target attribute tags. If the reconciliation parameters match the target attribute tags, target data features of the target reconciliation data are further acquired, and it is determined whether the reconciliation parameters match the target data features. By employing dual verification of the target reconciliation data's target attribute tags and target data features, the accuracy of the target reconciliation data to be reconciled can be further guaranteed, preventing deviations in the target data itself from affecting the final reconciliation result. Finally, when it is determined that the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result matching the reconciliation requirement information. By processing the reconciliation parameters and the verified target reconciliation data through the reconciliation model, the overall efficiency of reconciliation can be significantly improved while further ensuring the accuracy of the reconciliation results. Furthermore, for users, only the reconciliation requirement information needs to be entered to complete the entire reconciliation process, enabling users to actively configure reconciliation parameters and further enhancing the automation of the reconciliation work.

[0022] Figure 1 This is a schematic diagram illustrating the architecture of an application scenario for a model-based reconciliation method, provided by an exemplary embodiment. For example... Figure 1 As shown, the method may include a server 11, a network 12, and several electronic devices, such as a PC (Personal Computer) 13, a mobile phone 14, etc.

[0023] Server 11 can be a physical server containing an independent host, or it can be a virtual server hosted in a host cluster. During operation, server 11 can run server-side programs for a certain application to implement the relevant functions of that application. For example, when server 11 runs a model-based reconciliation program, it can act as the execution entity of the model-based reconciliation program.

[0024] PC13 and mobile phone14 are just some of the types of electronic devices that users can use. In reality, users can obviously also use electronic devices such as tablets, laptops, PDAs (Personal Digital Assistants), wearable devices (such as smart glasses, smartwatches, etc.), etc., and one or more embodiments in this specification do not limit this. During operation, the electronic device can run a client-side program of an application to implement the relevant functions of that application. For example, when the electronic device runs the aforementioned model-based reconciliation method program service, it can act as a client for that program. That is, the user can input reconciliation request information through the client, and then the client sends the reconciliation request information to the server so that the server runs the model-based reconciliation method and returns the reconciliation result matching the reconciliation request information to the client. In some embodiments, the aforementioned model-based reconciliation method can also run on the electronic device, that is, the client can also act as the execution subject running the model-based reconciliation program. The client application of the aforementioned program service can be launched and run on the electronic device. The client-side program can be a native application installed on the electronic device, or the client-side program can be a mini-program, quick app, or other similar form. Of course, when using web technologies such as HTML5 or similar, the relevant functions can be achieved through the page displayed by the browser. The browser here can be a standalone browser application or a browser module embedded in some applications.

[0025] As for the network 12 that enables interaction between electronic devices such as PC13 and mobile phone 14 and server 11, wired or wireless networks can be selected for communication based on the communication methods supported by the respective electronic devices. This specification does not impose any restrictions on this. For example, PC13 can support both wired and wireless communication, so it can use either wired or wireless networks as needed. Mobile phone 14 typically only supports wireless communication, so it can use a wireless network for communication.

[0026] refer to Figure 2 Here is a flowchart of a model-based reconciliation method provided in this specification, which includes the following steps: S202, obtain the reconciliation request information input by the user, and determine the reconciliation parameters based on the reconciliation request information.

[0027] The reconciliation request information entered by the user can be text information that the user directly enters on the reconciliation request interface, for example, reference Figure 3The user inputs the reconciliation request information as "Please reconcile the bills for Institution A and Institution B this month." After obtaining the user's input reconciliation request information, in order to accurately determine the current reconciliation parameters, it is necessary to further determine the reconciliation parameters based on the user's input reconciliation request information. In some embodiments, the reconciliation parameters can be obtained from the reconciliation request information through keyword extraction. For example, in the above example, Institution A, Institution B, and "this month" can be used as keywords for extraction, and the reconciliation parameters can be obtained accordingly. In some implementations, considering that the user's input reconciliation request information is sometimes not standard text information, for example, in the above example, it is necessary to clarify which month "this month" specifically refers to, or that the user inputs Institution A as an abbreviation, and the full name of the institution needs to be further restored, so as to provide a basis for accurate reconciliation later. In order to accurately obtain the reconciliation parameters, a trained semantic recognition model can be used to process the user's input reconciliation request information to obtain more standardized reconciliation parameters. It should be noted that the trained semantic recognition model can be a large model or a traditional neural network model, and there is no limitation on this.

[0028] It should be noted that the reconciliation parameters in the embodiments of this specification generally refer to various parameters used to instruct and regulate the reconciliation process. In some embodiments, the reconciliation parameters may include one or more of multiple reconciliation parameters such as the subject to be reconciled, the reconciliation period, the reconciliation scope, the reconciliation rules, the reconciliation basis, and the reconciliation product. In some embodiments, when the reconciliation request information entered by the user only contains some reconciliation parameters, the reconciliation parameters not included in the reconciliation request information can be set as default values. For example, in one instance, some general reconciliation rules can be set first, such as total revenue = total cost + total profit, call volume = cost-side call volume, reduction amount < 1% of total call volume, etc. When the reconciliation request information entered by the user does not include additional reconciliation rules, the above-mentioned general reconciliation rules can be directly used as part of the reconciliation parameters.

[0029] In some embodiments, before obtaining the user's input reconciliation request information, the user's reconciliation intent can be determined first. For example, the user can indicate their reconciliation intent by opening a program specifically for reconciliation, or the user can actively input text information indicating their reconciliation intent in a program with multiple functions, or the user can express their reconciliation intent in other ways, without limitation.

[0030] To further ensure the accuracy of the reconciliation parameters, in some embodiments of this specification, the model-based reconciliation method further includes: Display the reconciliation parameters; In response to receiving the parameter correction information input by the user, the reconciliation request information and the parameter correction information are processed by the trained semantic recognition model to obtain the reconciliation parameters again.

[0031] To ensure accurate reconciliation parameters without deviation, the reconciliation parameters can be proactively displayed to the user after they are obtained. This allows the user to determine if the displayed parameters match their input reconciliation requirements. If not, the user can input parameter correction information. Upon receiving this correction information, a trained semantic recognition model processes the reconciliation requirements and correction information to re-obtain the reconciliation parameters. After obtaining the re-obtained parameters, the model-based reconciliation method described in this embodiment can be executed to accurately obtain a reconciliation result that matches the reconciliation requirements.

[0032] To further improve the accuracy of the speech recognition model, upon receiving the parameter correction information input by the user, the model can be compared with the reconciliation request information. If the information contained in the reconciliation request information is less than that in the parameter correction information, it indicates that the previous error in reconciliation parameter recognition may have been due to user input rather than the semantic recognition model itself. In this case, no adjustments can be made to the semantic recognition model. If the information contained in the reconciliation request information is not less than that in the parameter correction information, it indicates that the previous error in reconciliation parameter recognition may have been due to insufficient accuracy of the semantic recognition model. In this case, the parameters of the semantic recognition model can be adjusted using the parameter correction information.

[0033] S204, obtain the target reconciliation data corresponding to the reconciliation parameters, extract the target attribute tags related to the reconciliation parameters from the target reconciliation data, and determine whether the reconciliation parameters match the target attribute tags.

[0034] After determining the reconciliation parameters, the target reconciliation data corresponding to those parameters can be further obtained. In some embodiments, all data requiring reconciliation can be saved to a database beforehand, and then the target reconciliation data corresponding to the current reconciliation parameters can be found from the database. For example, if a reconciliation parameter specifies that the reconciliation entities are institution A and institution B, and the reconciliation period is December, then the December reconciliation data for institutions A and B can be filtered from the database to serve as the target reconciliation data. In some embodiments, the target reconciliation data can also be obtained from user-provided reconciliation data using the current reconciliation parameters; this is not limited.

[0035] It should be noted that in this embodiment, actively obtaining the target reconciliation data through reconciliation parameters can reduce the process of manually compiling the target reconciliation data, improving efficiency and avoiding errors in reconciliation data compilation caused by human error. Furthermore, considering that data matching errors may occur when finding the corresponding target reconciliation data from a large amount of data using reconciliation parameters, this embodiment does not directly perform subsequent reconciliation based on the target reconciliation data after obtaining it. Instead, it first extracts target attribute tags related to the reconciliation parameters from the target reconciliation data. These target attribute tags can be tags that characterize certain attribute information of the target reconciliation data, and this attribute information is related to the reconciliation parameters. For example, in some embodiments, the target attribute tags can be tags that indicate the time period and / or the reconciliation subject to which the target reconciliation data belongs. These target attribute tags can generally be obtained directly from the header, file name, or title of the target reconciliation data. In some embodiments, the target attribute tags can also be obtained through fields in the target reconciliation data specifically used to record information related to the reconciliation parameters. For example, in one instance, the target reconciliation data may include fields such as month and customer name. These fields can be used to obtain attribute labels related to the reconciliation period and the reconciliation subject in the target reconciliation data.

[0036] After determining the target attribute tags for the target reconciliation data, it's possible to further determine whether the reconciliation parameters match the target attribute tags to verify the accurate acquisition of the target reconciliation data. If it's determined that the target attribute tags and reconciliation parameters do not match, the target reconciliation data needs to be reacquired based on the reconciliation parameters. For example, if the reconciliation period in the current reconciliation parameters is September, while the target reconciliation data in the target attribute tags belongs to June, this indicates that the acquired target reconciliation data does not match the reconciliation parameters. Directly reconciling based on the June reconciliation data would undoubtedly lead to errors in the final reconciliation result. Therefore, it's necessary to reacquire the target reconciliation data based on the reconciliation parameters.

[0037] In some embodiments, when determining whether the reconciliation parameters match the target attribute labels, semantic recognition can be used to determine whether the meanings of the reconciliation parameters and the target attribute labels are consistent. If they are consistent, then the two can be determined to match. In some embodiments, the distance between the corresponding vectors of the reconciliation parameters and the target attribute labels in space can also be used to determine whether the two match. If the distance between the corresponding vectors of the two in space is less than a preset distance, then the reconciliation parameters match the target data features.

[0038] S206, if it is determined that the reconciliation parameters match the target attribute tag, obtain the target data features of the target reconciliation data, and determine whether the reconciliation parameters match the target data features.

[0039] Considering that in some scenarios, attribute tags such as titles or file names of reconciliation data may inherently contain errors—for example, the title of a reconciliation entity's December reconciliation data might be mistakenly labeled as February, and the corresponding reconciliation period in the reconciliation parameters is February—in such cases, although the extracted target attribute tags and reconciliation parameters match in form, the actual target reconciliation data may not match the reconciliation parameters. Therefore, simply verifying the target reconciliation data based on the consistency between target attribute tags and reconciliation parameters cannot guarantee the accuracy of the reconciliation data. Therefore, in order to accurately determine whether the currently acquired target reconciliation data matches the reconciliation parameters when the target attribute tags of the target invoice data deviate, in this embodiment, if it is determined that the reconciliation parameters match the target attribute tags, the target data features of the target reconciliation data are further obtained, and it is determined whether the reconciliation parameters match the target data features.

[0040] It's important to note that reconciliation data is generally generated from real business transactions. Therefore, the data characteristics in reconciliation data typically follow certain patterns. For example, in the months leading up to the end of the year, reconciliation data from various entities usually increases significantly in order to meet the targets set for the previous year. Conversely, in the first few months of the year, reconciliation data from each entity tends to be relatively less. Furthermore, for the same entity, its reconciliation data generally shows continuous fluctuations. Alternatively, for certain stable entities in certain industries, the amount of reconciliation data may fluctuate around a specific quantity. The data characteristics of reconciliation data can effectively reflect these patterns; that is, the data characteristics of reconciliation data represent the inherent developmental patterns of reconciliation data. Therefore, after determining whether the reconciliation parameters match the target attribute tags, further verifying whether the target data characteristics match the reconciliation parameters can further improve the accuracy of the target reconciliation data.

[0041] To accurately determine whether reconciliation parameters match target data characteristics, in some embodiments of this specification, the reconciliation parameters include a reconciliation time period and various reconciliation entities, and the target data characteristics include statistical values ​​of data from each reconciliation entity within the reconciliation time period; determining whether the reconciliation parameters match the target data characteristics includes: For each reconciliation entity, determine the predicted data statistics for that entity during the reconciliation period, and match the predicted data statistics with the data statistics for that entity during the reconciliation period. Based on the matching results of each reconciliation entity, it is determined whether the reconciliation parameters match the target data features.

[0042] In this embodiment, the reconciliation parameters include the reconciliation time period and each reconciliation entity. In some embodiments, the reconciliation parameters may include two or more reconciliation entities. The target data features may include the statistical values ​​of each reconciliation entity during the reconciliation time period. These statistical values ​​generally refer to the quantitative results obtained after processing the original reconciliation data using a preset statistical method, which can quantify the data characteristics, inherent patterns, or relationships between data. In some embodiments, these statistical values ​​may include the total cost, average cost, variance, and other statistical values ​​of the reconciliation data. When the target data features include the statistical values ​​of each reconciliation entity during the reconciliation time period, for each reconciliation entity in the target reconciliation data, the predicted statistical value of that reconciliation entity during the reconciliation time period can be determined first. Then, the predicted statistical value is matched with the statistical value of that reconciliation entity during the reconciliation time period. Finally, based on the matching results of each reconciliation entity, it is determined whether the reconciliation parameters match the target data features. That is, when the matching results of all reconciliation entities are all matched, it can be determined that the reconciliation parameters match the target data features. In some embodiments, a preset predicted data statistical value can be configured in advance for each reconciliation entity in each reconciliation time period based on industry experience or needs. Then, based on the preset predicted data statistical value for each reconciliation entity in each time period, the predicted data statistical value for each reconciliation entity in the target reconciliation data in the current time period can be determined.

[0043] To more accurately determine the statistical values ​​of the predicted data, in some embodiments of this specification, determining the statistical values ​​of the predicted data for the reconciliation entity during the reconciliation period includes: Determine the historical time period preceding the reconciliation time period, and obtain the data statistics of the reconciliation entity during the historical time period; The predicted data statistics are determined based on the data statistics of the reconciliation entity during the historical period.

[0044] Considering the differences between different reconciliation entities, we can first determine the historical time periods preceding the reconciliation period. For example, if the current reconciliation period is December, the corresponding historical time periods could be January to November of the same year, or other time periods preceding December. After determining the historical time periods, we first obtain the statistical data values ​​for each reconciliation entity within those historical time periods. Then, we determine the predicted statistical data values ​​based on these historical statistical values. It should be noted that for each reconciliation entity, reconciliation data generally changes continuously over time. Therefore, the predicted statistical data values ​​for the current time period can be well estimated using the statistical data values ​​for each reconciliation entity within its historical time periods. If the statistical data value for a particular reconciliation entity in the obtained target reconciliation data does not match its corresponding predicted statistical data value, it indicates a deviation in the acquisition of the reconciliation data for that entity.

[0045] In some embodiments of this specification, determining the predicted data statistics based on the data statistics of the reconciliation entity during the historical time period includes: Based on the data statistics of the reconciliation entity during the historical period, determine the target trend of the data statistics of the reconciliation entity in the time dimension; Based on the reconciliation period and the target change trend, the predicted data statistics are determined.

[0046] To accurately determine the predicted statistical values ​​of reconciliation entities during the reconciliation period based on their historical statistical values, a target trend in the statistical values ​​over time can be determined for each entity. This target trend can be a change in the statistical values ​​over time, such as an increase or decrease, or a fitting function that reflects the target trend based on the historical statistical values. There is no limitation on this. After determining the target trend for each reconciliation entity, the predicted statistical values ​​can be determined based on this trend. For example, when the target trend is represented by a fitting function, the reconciliation period can be substituted into the fitting function to obtain the predicted statistical values ​​corresponding to that period. When the target trend is represented as an increase or decrease in the statistical values ​​over time, adjacent historical periods can be found based on the reconciliation period. Then, the statistical values ​​of the reconciliation entity in those adjacent historical periods can be adjusted according to the target trend to obtain the predicted statistical values.

[0047] In some embodiments of this specification, determining the predicted data statistics based on the data statistics of the reconciliation entity during the historical time period includes: Obtain the industry knowledge base corresponding to the reconciliation entity; Based on the data statistics of the reconciliation entity during the historical period, the industry knowledge base, and the reconciliation period, a second prompt word is generated and input into the trained prediction model to determine the predicted data statistics.

[0048] To further improve the accuracy of predicted statistical values, considering that in some industries, the statistical values ​​of reconciliation data fluctuate significantly over time and exhibit a degree of randomness, and that the relationship between statistical values ​​and time is not a simple linear one, it is difficult to accurately predict future trends using historical statistical values. Therefore, this embodiment proposes using a trained large-scale prediction model to obtain the predicted statistical values ​​of the reconciliation entity. Furthermore, when determining the prompt words for the input model, the industry knowledge base corresponding to each reconciliation entity is first obtained. For example, for reconciliation entities primarily dealing in electronic products, the statistical values ​​of reconciliation data often fluctuate significantly during specific industry promotional holidays. When generating the second prompt word, considering both the industry knowledge base corresponding to the reconciliation entity and the statistical values ​​of historical time periods provides a more comprehensive reference basis for the large-scale prediction model, further improving the accuracy of the predicted statistical values.

[0049] S208, if it is determined that the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

[0050] After verifying the target reconciliation data using both target attribute tags and target data features, it can be ensured that the target reconciliation data is free of deviation. Then, a first prompt word can be generated based on the reconciliation parameters and the target reconciliation data, and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirements. It should be noted that, considering that comparing target attribute tags with reconciliation parameters is relatively simpler, when verifying the target reconciliation data, it can be verified first based on the target attribute tags. This way, if it is determined that the reconciliation parameters and the target attribute tags do not match, the target reconciliation data can be directly retrieved again without needing a second verification using target data features, further improving the overall efficiency of the verification process.

[0051] In some embodiments, the training process of the reconciliation model can refer to the training process of large models in related technologies. For example, in one example, a baseline large model can be selected as needed, and then the collected historical reconciliation data can be divided into a training set, a validation set, and a test set for training, validation, and evaluation of the large model, respectively. It should be noted that compared with directly using the general large model in related technologies, the reconciliation model obtained through training can further improve the reconciliation capability of the large model because it has been reinforced by using historical reconciliation data.

[0052] To avoid erroneous reconciliation results caused by generalization of large models, in some embodiments of this specification, the reconciliation result includes the reconciliation inconsistency portion; the model-based reconciliation method further includes: Calculate the first bill statistics for each reconciliation entity in the reconciliation inconsistency portion, and determine whether the first bill statistics for each reconciliation entity match. In response to the determination that there is a mismatch between the first bill statistics data corresponding to each reconciliation entity, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data.

[0053] Since the generalization of the large reconciliation model may lead to some deviation in the output results, this embodiment further verifies the reconciliation results output by the large reconciliation model by using bill statistics data after obtaining reconciliation results that match the reconciliation requirements. In specific implementation, considering that the reconciliation results generally include a reconciliation inconsistency portion and a reconciliation discrepancy portion, and that the reconciliation data of each reconciliation entity needs to match in the reconciliation inconsistency portion, this embodiment first calculates the first bill statistics data of each reconciliation entity in the reconciliation inconsistency portion, and then determines whether the first bill statistics data corresponding to each reconciliation entity match. If they do not match, it indicates that the reconciliation results output by the large reconciliation model are incorrect. Therefore, it is necessary to regenerate the first prompt word based on the reconciliation parameters and the target reconciliation data so that the large reconciliation model can re-output the reconciliation results based on the regenerated first prompt word. In some embodiments, to ensure that the reconciliation model avoids the same error when regenerating the reconciliation results, the information that there is an error in the reconciliation inconsistency section can be added to the first prompt text of the regeneration, so as to prevent the reconciliation model from generating the same reconciliation results a second time.

[0054] To more accurately determine the range of data where reconciliation errors occurred, in some embodiments of this specification, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data, including: The reconciliation inconsistency portion is divided into multiple reconciliation inconsistency sub-regions; A target reconciliation inconsistency sub-region is determined from multiple reconciliation inconsistency sub-regions; wherein, the billing statistics of each reconciliation entity do not match among the billing statistics in the target reconciliation inconsistency sub-region. The first prompt word is regenerated based on the target reconciliation inconsistency area, the reconciliation parameters, and the target reconciliation data.

[0055] It should be noted that, generally, the deviation in reconciliation results caused by the generalization of a large model is not too extensive; that is, most of the data output by the large model is reliable. Therefore, to further improve the efficiency of the model in regenerating reconciliation results, in this embodiment, the reconciliation inconsistency portion is first divided into multiple reconciliation inconsistency sub-regions. The specific number of these sub-regions can be set as needed and is not limited. Then, a target reconciliation inconsistency sub-region is determined from these sub-regions. This target sub-region is the sub-interval where the reconciliation error occurred, i.e., the bill statistics data of each reconciliation entity do not match in the target reconciliation inconsistency sub-region. After determining the target reconciliation inconsistency sub-region, the first prompt word can be regenerated based on this target reconciliation inconsistency sub-region, the reconciliation parameters, and the target reconciliation data.

[0056] In some embodiments of this specification, the reconciliation result includes a reconciliation discrepancy portion; the model-based reconciliation method further includes: Calculate the second billing statistics for each reconciliation entity in the reconciliation discrepancy section, and determine whether the second billing statistics for each reconciliation entity match. In response to determining the matching between the second bill statistics data corresponding to each reconciliation entity, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data.

[0057] Since the generalization of the large model may not only classify erroneous reconciliation results into the reconciliation inconsistency part, but also classify correct reconciliation results into the reconciliation discrepancy part, this embodiment will further calculate the second bill statistics data of each reconciliation entity in the reconciliation discrepancy part, and determine whether the second bill statistics data corresponding to each reconciliation entity match. When it is determined that the second bill statistics data corresponding to each reconciliation entity match, the first prompt word can be regenerated based on the reconciliation parameters and the target reconciliation data.

[0058] Considering that when the correct reconciliation result only exists in a certain sub-region of the reconciliation discrepancy, directly judging whether the second bill statistics of the entire reconciliation discrepancy portion match still cannot yield a definite result, in some embodiments, regenerating the first prompt word based on the reconciliation parameters and the target reconciliation data may include: dividing the reconciliation discrepancy portion into multiple reconciliation discrepancy sub-regions, and then determining the target reconciliation discrepancy sub-region from the multiple reconciliation discrepancy sub-regions; wherein, each reconciliation entity matches the bill statistics data in the target reconciliation discrepancy sub-region; finally, regenerating the first prompt word based on the target reconciliation discrepancy region, the reconciliation parameters, and the target reconciliation data.

[0059] To further improve the accuracy of the reconciliation model, in some embodiments of this specification, the model-based reconciliation method further includes: Obtain the reconciliation results after annotation; The reconciliation model is adjusted based on the annotated reconciliation results.

[0060] To continuously improve the accuracy of the reconciliation model, after obtaining reconciliation results that match the reconciliation requirements, these results can be labeled. The labeling can indicate whether the reconciliation result is correct or incorrect. Then, the reconciliation model is adjusted based on the labeled reconciliation results. In some embodiments, the reconciliation results can be labeled manually, or the labeled results can be labeled by a trained labeling model; there are no limitations on this.

[0061] refer to Figure 4 Here is a flowchart of another model-based reconciliation method provided in this specification, which includes the following steps: S402, obtain the reconciliation request information input by the user, and determine the reconciliation parameters based on the reconciliation request information.

[0062] S404, obtain the target reconciliation data corresponding to the reconciliation parameters, and extract the target attribute tags related to the reconciliation parameters from the target reconciliation data.

[0063] S406, determine whether the reconciliation parameters match the target attribute tags. If they match, proceed to step S408; otherwise, re-execute step S404.

[0064] S408, Obtain the target data characteristics of the target reconciliation data.

[0065] S410, determine whether the reconciliation parameters match the target data characteristics. If they match, proceed to step S412. If they do not match, re-execute step S404.

[0066] S412, based on the reconciliation parameters and target reconciliation data, generates the first prompt word and inputs it into the reconciliation model to obtain the reconciliation result that matches the reconciliation requirement information.

[0067] It should be noted that the detailed descriptions of steps S402 to S412 above can be found in the above descriptions. Figure 2 The relevant parts of the corresponding embodiments will not be described in detail here.

[0068] The model-based reconciliation method provided in this specification first acquires the reconciliation request information input by the user, and then determines the reconciliation parameters based on this information. This allows the user to accurately control the current reconciliation parameters according to their input, facilitating automatic reconciliation in various business scenarios. After determining the reconciliation parameters, to further improve reconciliation efficiency, target reconciliation data corresponding to the parameters can be automatically acquired, eliminating the need for manual preparation of the data to be reconciled. Simultaneously, to avoid deviations in the automatically acquired target reconciliation data, after acquiring the target data, target attribute tags related to the reconciliation parameters are extracted from the target reconciliation data, and it is determined whether the reconciliation parameters match the target attribute tags. If the reconciliation parameters match the target attribute tags, target data features of the target reconciliation data are further acquired, and it is determined whether the reconciliation parameters match the target data features. Through dual verification of the target reconciliation data's target attribute tags and target data features, the accuracy of the target reconciliation data to be reconciled can be further guaranteed, preventing deviations in the target data itself from affecting the final reconciliation result. Finally, when it is determined that the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information. By processing the reconciliation parameters and the validated target reconciliation data through the reconciliation model, the overall efficiency of reconciliation can be significantly improved while further ensuring the accuracy of the reconciliation results. Furthermore, for users, the entire reconciliation process can be completed simply by inputting the reconciliation requirement information. This not only allows for user-initiated configuration of reconciliation parameters but also further enhances the automation of the reconciliation process.

[0069] Figure 5 This is a schematic structural diagram of a device provided in an exemplary embodiment. For example... Figure 5As shown, device 500 mainly consists of a communication interface 502, a user interface 504, a processor 506, and a data storage 508. These components are interconnected and communicate with each other via a system bus, network, or other connection mechanism 510. The communication interface 502 enables device 500 to communicate with other devices, access networks, and transmission networks via analog or digital modulation. For example, the communication interface 502 may include a chipset and antenna for wireless communication with a radio access network or access point. Furthermore, the communication interface 502 can be a wired interface such as Ethernet, Token Ring, or a USB port, or a wireless interface such as Wi-Fi, Bluetooth, Global Positioning System (GPS), or a wide-area wireless interface (e.g., WiMAX or LTE). Of course, the communication interface 502 can also support other forms of physical layer interfaces and standard or proprietary communication protocols. The communication interface 502 may also include multiple physical communication interfaces, such as Wi-Fi interfaces, Bluetooth interfaces, and wide-area wireless interfaces.

[0070] User interface 504 includes receiving user input and providing output to the user. Therefore, user interface 504 may include input components such as a keypad, keyboard, touch-sensitive or presence-sensitive panel, computer mouse, trackball, joystick, microphone, still camera, and video camera, and output components such as a display screen (which may be combined with a touch-sensitive panel), CRT, LCD, LED, display using DLP technology, printer, and other similar devices known or developed in the future. User interface 504 may also generate auditory output via speakers, speaker jacks, audio output ports, audio output devices, headphones, and other similar devices known or developed in the future. In some embodiments, user interface 504 may include software, circuitry, or other forms of logic capable of transmitting and receiving data from external user input / output devices. Additionally or alternatively, device 500 may support remote access from other devices via communication interface 502 or another physical interface (not shown). User interface 504 may be configured to receive user input, the position and movement of which may be indicated by indicators or cursors described herein. User interface 504 may also be configured as a display device for rendering or displaying text fragments.

[0071] Processor 506 may contain one or more general-purpose processors and / or special-purpose processors.

[0072] Data storage 508 may include one or more volatile and / or non-volatile storage components and may be integrated wholly or partially with processor 506. Data storage 508 may include removable and non-removable components.

[0073] Processor 506 is capable of executing program instructions 518 (e.g., compiled or uncompiled program logic and / or machine code) stored in data storage 508 to perform the various functions described herein. Data storage 508 may contain a non-transitory computer-readable medium on which program instructions are stored, which, when executed by device 500, enable device 500 to perform any methods, processes, or functions disclosed in this specification and / or the accompanying drawings. Execution of program instructions 518 by processor 506 may result in processor 506 using data 512.

[0074] For example, program instructions 518 may include an operating system 522 (e.g., an operating system kernel, device drivers, and / or other modules) installed on device 500 and one or more applications 520 (e.g., a browser, social application, or game application). Similarly, data 512 may include operating system data 516 and application data 514. Operating system data 516 is primarily accessible to the operating system 522, while application data 514 is primarily accessible to one or more applications 520. Application data 514 may reside in a file system visible or hidden from the user of device 500.

[0075] Application 520 can communicate with operating system 522 through one or more application programming interfaces (APIs). These APIs help application 520 read and / or write application data 514, transmit or receive information via communication interface 502, receive or display information on user interface 504, etc.

[0076] In some terminology, application 520 may be simply referred to as "app". Furthermore, application 520 can be downloaded to device 500 through one or more online app stores or app markets. However, applications can also be installed on device 500 in other ways, such as through a web browser or a physical interface on device 500 (e.g., a USB port).

[0077] Please refer to Figure 6 Model-based reconciliation devices can be applied to, for example... Figure 5 The device shown implements the technical solution described in this specification. This model-based reconciliation device may include: The acquisition module 602 acquires the reconciliation request information input by the user and determines the reconciliation parameters based on the reconciliation request information; The first determining module 604 acquires target reconciliation data corresponding to the reconciliation parameters, extracts target attribute tags related to the reconciliation parameters from the target reconciliation data, and determines whether the reconciliation parameters match the target attribute tags. The second determining module 606, when determining that the reconciliation parameters match the target attribute tags, obtains the target data features of the target reconciliation data and determines whether the reconciliation parameters match the target data features; The reconciliation module 608, upon determining that the reconciliation parameters match the target data features, generates a first prompt word based on the reconciliation parameters and the target reconciliation data and inputs it into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

[0078] In some embodiments of this specification, the reconciliation parameters include a reconciliation time period and each reconciliation entity, and the target data features include statistical values ​​of data for each reconciliation entity during the reconciliation time period; the second determining module includes: The prediction data determination unit determines the predicted data statistics for each reconciliation entity during the reconciliation period and matches the predicted data statistics with the data statistics for that reconciliation entity during the reconciliation period. The matching unit determines whether the reconciliation parameters match the target data features based on the matching results of each reconciliation entity.

[0079] In some embodiments of this specification, the prediction data determination unit includes: The historical data acquisition component determines the historical time period preceding the reconciliation time period and acquires the data statistics of the reconciliation entity during the historical time period. The predictive data determination component determines the predictive data statistics based on the data statistics of the reconciliation entity during the historical time period.

[0080] In some embodiments of this specification, the prediction data determination component is specifically used for: Based on the data statistics of the reconciliation entity during the historical period, determine the target trend of the data statistics of the reconciliation entity in the time dimension; Based on the reconciliation period and the target change trend, the predicted data statistics are determined.

[0081] In some embodiments of this specification, the prediction data determination component is specifically used for: Obtain the industry knowledge base corresponding to the reconciliation entity; Based on the data statistics of the reconciliation entity during the historical period, the industry knowledge base, and the reconciliation period, a second prompt word is generated and input into the trained prediction model to determine the predicted data statistics.

[0082] In some embodiments of this specification, the reconciliation result includes the portion where there is no discrepancy; the model-based reconciliation device further includes a first verification module, which includes: The billing statistics unit calculates the first billing statistics for each reconciliation entity in the reconciliation discrepancy portion and determines whether the first billing statistics for each reconciliation entity match. The prompt word unit, in response to determining a mismatch between the first bill statistics data corresponding to each reconciliation entity, regenerates the first prompt word based on the reconciliation parameters and the target reconciliation data.

[0083] In some embodiments of this specification, the prompt word unit is specifically used for: The reconciliation inconsistency portion is divided into multiple reconciliation inconsistency sub-regions; A target reconciliation inconsistency sub-region is determined from multiple reconciliation inconsistency sub-regions; wherein, the billing statistics of each reconciliation entity do not match among the billing statistics in the target reconciliation inconsistency sub-region. The first prompt word is regenerated based on the target reconciliation inconsistency area, the reconciliation parameters, and the target reconciliation data.

[0084] In some embodiments of this specification, the reconciliation result includes a reconciliation discrepancy portion; the model-based reconciliation device further includes a second verification module, used for: Calculate the second billing statistics for each reconciliation entity in the reconciliation discrepancy section, and determine whether the second billing statistics for each reconciliation entity match. In response to determining the matching between the second bill statistics data corresponding to each reconciliation entity, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data.

[0085] In some embodiments of this specification, the model-based reconciliation device further includes a display module for: Display the reconciliation parameters; In response to receiving the parameter correction information input by the user, the reconciliation request information and the parameter correction information are processed by the trained semantic recognition model to obtain the reconciliation parameters again.

[0086] In some embodiments of this specification, the model-based reconciliation device further includes an adjustment module for: Obtain the reconciliation results after annotation; The reconciliation model is adjusted based on the annotated reconciliation results.

[0087] For ease of description, the above devices are described by dividing them into various modules or units based on their functions. Of course, when implementing one or more of these specifications, the functions of each module or unit can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another method, or some features may be ignored or not executed.

[0088] Based on the same concept as the methods described above, this specification also provides an electronic device, including: a processor; a memory for storing processor-executable instructions; wherein the processor executes the executable instructions to implement the steps of the model-based reconciliation method as described in any of the above embodiments.

[0089] Based on the same concept as the methods described above, this specification also provides a computer-readable storage medium having computer instructions stored thereon that, when executed by a processor, implement the steps of the model-based reconciliation method as described in any of the above embodiments.

[0090] Based on the same concept as the methods described above, this specification also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the model-based reconciliation method as described in any of the above embodiments.

[0091] What those skilled in the art will understand is: In this specification, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, product, or apparatus that includes said elements is not excluded.

[0092] In this specification, “a,” “an,” and “the” do not specifically refer to the singular, but may also include the plural.

[0093] In this specification, ordinal numbers such as "first," "second," etc., do not necessarily indicate order; they are often used to distinguish between objects. For example, "first server" and "second server" usually refer to two servers. To differentiate between these two servers, they are described as "first server" and "second server." Of course, sometimes these two servers may be the same server.

[0094] In this specification, unless explicitly stated otherwise, "receiving and sending data" does not necessarily mean direct receiving and sending; it can also mean indirect receiving and sending. For example, A receiving data sent by B can be understood as A directly receiving the data sent by B, or it can be understood as A indirectly receiving the data sent by B through other entities such as C. Similarly, B sending data to A can be understood as B sending the data directly to A, or it can be understood as B indirectly sending the data to A through other entities such as C. Here, C can be one entity, or it can be two or more entities.

[0095] In this specification, unless explicitly stated otherwise, the relationships between structures can be direct or indirect. For example, when describing "A is connected to B," unless it is explicitly stated that A and B are directly connected, it should be understood that A can be directly connected to B or indirectly connected to B. Similarly, when describing "A is on top of B," unless it is explicitly stated that A is directly above B (AB is adjacent and A is above B), it should be understood that A can be directly above B or indirectly above B (AB is separated by other elements, and A is above B). And so on.

[0096] This specification uses specific terms to describe embodiments thereof. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of those different embodiments or examples, without contradiction.

[0097] Although one or more embodiments of this specification provide method steps as described in the embodiments or flowcharts, it is understood that the order of steps listed in the embodiments or flowcharts is only one of many possible execution orders and does not represent the only execution order. Therefore, when the claims involve method steps, any changes or adjustments to the order of such steps, or the parallelism between steps, are also within the scope of protection of the claims.

Claims

1. A model-based reconciliation method, comprising: Obtain the reconciliation request information input by the user, and determine the reconciliation parameters based on the reconciliation request information; Obtain the target reconciliation data corresponding to the reconciliation parameters, extract the target attribute tags related to the reconciliation parameters from the target reconciliation data, and determine whether the reconciliation parameters match the target attribute tags; If it is determined that the reconciliation parameters match the target attribute tags, the target data features of the target reconciliation data are obtained, and it is determined whether the reconciliation parameters match the target data features. If the reconciliation parameters match the target data features, a first prompt word is generated based on the reconciliation parameters and the target reconciliation data and input into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

2. The method according to claim 1, wherein the reconciliation parameters include a reconciliation time period and each reconciliation entity, and the target data features include the data statistics of each reconciliation entity during the reconciliation time period; Determining whether the reconciliation parameters match the target data characteristics includes: For each reconciliation entity, determine the predicted data statistics for that entity during the reconciliation period, and match the predicted data statistics with the data statistics for that entity during the reconciliation period. Based on the matching results of each reconciliation entity, it is determined whether the reconciliation parameters match the target data features.

3. The method according to claim 2, determining the predicted data statistics of the reconciliation entity during the reconciliation period, includes: Determine the historical time period preceding the reconciliation time period, and obtain the data statistics of the reconciliation entity during the historical time period; The predicted data statistics are determined based on the data statistics of the reconciliation entity during the historical period.

4. The method according to claim 3, wherein determining the predicted data statistics based on the data statistics of the reconciliation entity during the historical time period includes: Based on the data statistics of the reconciliation entity during the historical period, determine the target trend of the data statistics of the reconciliation entity in the time dimension; Based on the reconciliation period and the target change trend, the predicted data statistics are determined.

5. The method according to claim 3, wherein determining the predicted data statistics based on the data statistics of the reconciliation entity during the historical time period comprises: Obtain the industry knowledge base corresponding to the reconciliation entity; Based on the data statistics of the reconciliation entity during the historical period, the industry knowledge base, and the reconciliation period, a second prompt word is generated and input into the trained prediction model to determine the predicted data statistics.

6. The method according to claim 1, wherein the reconciliation result includes the portion of the reconciliation with no discrepancies; the method further includes: Calculate the first bill statistics for each reconciliation entity in the reconciliation inconsistency portion, and determine whether the first bill statistics for each reconciliation entity match. In response to the determination that there is a mismatch between the first bill statistics data corresponding to each reconciliation entity, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data.

7. The method according to claim 6, wherein the first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data, comprising: The reconciliation inconsistency portion is divided into multiple reconciliation inconsistency sub-regions; A target reconciliation inconsistency sub-region is determined from multiple reconciliation inconsistency sub-regions; wherein, the billing statistics of each reconciliation entity do not match among the billing statistics in the target reconciliation inconsistency sub-region. The first prompt word is regenerated based on the target reconciliation inconsistency area, the reconciliation parameters, and the target reconciliation data.

8. The method according to claim 1, wherein the reconciliation result includes the reconciliation discrepancy portion; the method further includes: Calculate the second billing statistics for each reconciliation entity in the reconciliation discrepancy section, and determine whether the second billing statistics for each reconciliation entity match. In response to determining the matching between the second bill statistics data corresponding to each reconciliation entity, a first prompt word is regenerated based on the reconciliation parameters and the target reconciliation data.

9. The method according to claim 1, further comprising: Display the reconciliation parameters; In response to receiving the parameter correction information input by the user, the reconciliation request information and the parameter correction information are processed by the trained semantic recognition model to obtain the reconciliation parameters again.

10. The method according to claim 1, further comprising: Obtain the reconciliation results after annotation; The reconciliation model is adjusted based on the annotated reconciliation results.

11. A model-based reconciliation device, comprising: The acquisition module acquires the reconciliation request information input by the user and determines the reconciliation parameters based on the reconciliation request information; The first determining module acquires target reconciliation data corresponding to the reconciliation parameters, extracts target attribute tags related to the reconciliation parameters from the target reconciliation data, and determines whether the reconciliation parameters match the target attribute tags. The second determining module, when it is determined that the reconciliation parameters match the target attribute tags, obtains the target data features of the target reconciliation data and determines whether the reconciliation parameters match the target data features; The reconciliation module, upon determining that the reconciliation parameters match the target data features, generates a first prompt word based on the reconciliation parameters and the target reconciliation data, and inputs it into the trained reconciliation model to obtain a reconciliation result that matches the reconciliation requirement information.

12. An electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor executes the executable instructions to implement the method as described in any one of claims 1-10.

13. A computer-readable storage medium having stored thereon computer instructions that, when executed by a processor, implement the steps of the method as claimed in any one of claims 1-10.

14. A computer program product comprising: A computer program / instruction that, when executed by a processor, implements the method as described in any one of claims 1-10.

Images