Customer service knowledge base data reconciliation methods, storage media, and electronic devices

By automating the acquisition, reasonable segmentation, and parallel execution of data reconciliation tasks, the problem of low data reconciliation efficiency in the customer service knowledge base was solved, achieving efficient and accurate data alignment and improving the data processing efficiency of the customer service knowledge base.

CN123070264APending Publication Date: 2026-11-10TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510592187.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-05-08
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

The data reconciliation efficiency of the customer service knowledge base is low, and the existing technology relies on manual operation, which takes a long time when dealing with large-scale data.

Method used

By automating data acquisition, reasonably dividing the data set into multiple subsets, ensuring that the difference in data volume between each subset is within a preset threshold, executing data reconciliation tasks in parallel, utilizing multi-threading or distributed computing technologies for data reconciliation, and automatically aligning data based on the reconciliation results.

Benefits of technology

It has automated and parallelized the reconciliation of customer service knowledge base data, improved the efficiency of data reconciliation, ensured the balanced use of computing resources, fully leveraged the advantages of parallel processing, shortened the reconciliation time, and improved the accuracy and consistency of data processing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123070264A_ABST
    Figure CN123070264A_ABST
Patent Text Reader

Abstract

This application discloses a data reconciliation method, apparatus, storage medium, and electronic device for a customer service knowledge base. The method includes: in response to a data reconciliation request, acquiring a knowledge base dataset for the customer service knowledge base; dividing the knowledge base dataset into at least two knowledge base data subsets; executing in parallel data reconciliation tasks corresponding to each knowledge base data subset to obtain data reconciliation results, wherein the data reconciliation tasks are used to compare data differences between the knowledge base data subsets and the source knowledge base, and the data reconciliation results are used to represent data differences between the customer service knowledge base and the source knowledge base; and aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results, using the data in the source knowledge base as a benchmark. This application solves the technical problem of low data reconciliation efficiency for customer service knowledge bases.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of computers, and more specifically, to a data reconciliation method, apparatus, storage medium, and electronic device for a customer service knowledge base. Background Technology

[0002] In customer service knowledge base data reconciliation scenarios, the reconciliation process is typically initiated manually, such as on fixed workdays or weekly periods, with personnel manually operating the relevant systems to verify data. However, when dealing with large-scale data, the entire reconciliation process is time-consuming, leading to low efficiency in customer service knowledge base data reconciliation. Therefore, the issue of low efficiency in customer service knowledge base data reconciliation exists.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This application provides a data reconciliation method, apparatus, storage medium, and electronic device for a customer service knowledge base, to at least solve the technical problem of low data reconciliation efficiency in customer service knowledge bases.

[0005] According to one aspect of the embodiments of this application, a data reconciliation method for a customer service knowledge base is provided, comprising: responding to a data reconciliation request, obtaining a knowledge base data set of the customer service knowledge base, wherein the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the alignment of data between the customer service knowledge base and a source knowledge base; dividing the knowledge base data set to obtain at least two knowledge base data subsets, wherein the data volume difference between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold; executing each data reconciliation task corresponding to each knowledge base data subset in parallel to obtain a data reconciliation result, wherein the data reconciliation task is used to compare the data differences between the knowledge base data subsets and the source knowledge base, and the data reconciliation result is used to represent the data differences between the customer service knowledge base and the source knowledge base; and aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation result and using the data in the source knowledge base as a benchmark.

[0006] According to another aspect of the embodiments of this application, a data reconciliation device for a customer service knowledge base is also provided, comprising: a first acquisition unit, configured to acquire a knowledge base data set of a customer service knowledge base in response to a data reconciliation request, wherein the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the alignment of data between the customer service knowledge base and a source knowledge base; a segmentation unit, configured to segment the knowledge base data set to obtain at least two knowledge base data subsets, wherein the data volume difference between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold; a parallel unit, configured to execute each data reconciliation task corresponding to each knowledge base data subset in parallel and acquire data reconciliation results, wherein the data reconciliation task is used to compare the data differences between the knowledge base data subsets and the source knowledge base, and the data reconciliation results are used to represent the data differences between the customer service knowledge base and the source knowledge base; and an alignment unit, configured to align the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark.

[0007] As an optional solution, the above-mentioned parallel unit includes: a parallel module, used to execute the above-mentioned data reconciliation tasks in parallel and obtain the execution results of each of the above-mentioned data reconciliation tasks; and an integration module, used to integrate the execution results of the above-mentioned tasks to obtain the above-mentioned data reconciliation results.

[0008] As an optional solution, the above-mentioned apparatus further includes: a first acquisition module, configured to acquire a first question corresponding to a target reconciliation task during the process of parallel execution of the various data reconciliation tasks and acquisition of the execution results of the various tasks corresponding to the various data reconciliation tasks, wherein the target reconciliation task is used to compare the data differences between a target data subset and the source knowledge base, the various data reconciliation tasks include the target reconciliation task, and the various knowledge base data subsets include the target data subset; and a second acquisition module, configured to acquire the first question corresponding to a target reconciliation task during the process of parallel execution of the various data reconciliation tasks and acquisition of the execution results of the various tasks corresponding to the various data reconciliation tasks. In the process of parallel execution of the above-mentioned data reconciliation tasks and obtaining the execution results of the corresponding tasks, the third acquisition module is used to acquire the source knowledge base for the above-mentioned first question, which is a second answer to the above-mentioned first question; the fourth acquisition module is used to compare the above-mentioned first answer and the above-mentioned second answer during the process of parallel execution of the above-mentioned data reconciliation tasks and obtaining the execution results of the corresponding tasks to obtain the first answer information difference, wherein the execution results of the above-mentioned target reconciliation tasks include the above-mentioned first answer information difference.

[0009] As an optional solution, the above apparatus further includes: a fifth acquisition module, configured to acquire a second question corresponding to the target reconciliation task during the process of parallel execution of the various data reconciliation tasks and acquisition of the execution results of the various data reconciliation tasks, wherein the second question is related to the first question; a sixth acquisition module, configured to acquire a third answer to the second question for the target data subset during the process of parallel execution of the various data reconciliation tasks and acquisition of the execution results of the various data reconciliation tasks; and a seventh acquisition module, configured to acquire a third answer to the second question for the target data subset during the process of parallel execution of the various data reconciliation tasks and acquisition of the execution results of the various data reconciliation tasks. During the execution of each task corresponding to the reconciliation task, the aforementioned source knowledge base is obtained, along with the fourth answer to the aforementioned second question. The eighth acquisition module is used to obtain the associated coherence information corresponding to the aforementioned target reconciliation task based on the aforementioned first answer, second answer, third answer, and fourth answer during the parallel execution of the aforementioned data reconciliation tasks and the acquisition of the execution results of each task corresponding to the aforementioned data reconciliation tasks. The associated coherence information is used to represent the coherence of the aforementioned target data subset and the aforementioned source knowledge base in terms of related questions and answers in business logic. The task execution results corresponding to the aforementioned target reconciliation task include the aforementioned associated coherence information.

[0010] As an optional solution, the above-mentioned apparatus further includes: a second acquisition unit, configured to acquire the data reconciliation request before acquiring the knowledge base data set of the customer service knowledge base in response to the data reconciliation request, provided that at least one reconciliation triggering condition is met, wherein the at least one reconciliation triggering condition includes at least one of the following: reconciliation frequency condition, data change condition of the customer service knowledge base, and customer service interaction condition of the customer service knowledge base.

[0011] As an optional solution, the above-mentioned device further includes: a first determining unit, configured to determine that the first triggering condition is met before obtaining the data reconciliation request, when the amount of data added or modified in the customer service knowledge base within a preset time period is greater than or equal to a first threshold, wherein the first triggering condition includes the data change condition of the customer service knowledge base, and the at least one reconciliation triggering condition includes the first triggering condition; or, a second determining unit, configured to determine that the second triggering condition is met before obtaining the data reconciliation request, when the reconciliation time interval of the customer service knowledge base is greater than or equal to a second threshold, and the number of customer service interactions in the customer service knowledge base is greater than or equal to a third threshold, wherein the second triggering condition includes the reconciliation frequency condition and the customer service interaction condition of the customer service knowledge base, and the at least one reconciliation triggering condition includes the second triggering condition.

[0012] As an optional solution, the above-mentioned apparatus further includes: a third determining unit, configured to determine a data segmentation method based on the data type of the knowledge base data set before segmenting the knowledge base data set to obtain at least two knowledge base data subsets; the segmentation unit includes: a segmentation module, configured to segment the knowledge base data set according to the data segmentation method to obtain the at least two knowledge base data subsets.

[0013] As an optional solution, the above-mentioned segmentation module includes: a first segmentation submodule, used to segment the above-mentioned knowledge base data set based on a hash value modulo segmentation algorithm when the amount of data of different data types in the above-mentioned knowledge base data set is greater than or equal to a fourth threshold, to obtain the above-mentioned at least two knowledge base data subsets; and a second segmentation submodule, used to segment the above-mentioned knowledge base data set based on a time range segmentation algorithm when the amount of data of different data types in the above-mentioned knowledge base data set is less than the above-mentioned fourth threshold, to obtain the above-mentioned at least two knowledge base data subsets.

[0014] As an optional solution, the above-mentioned apparatus further includes: a fourth determining unit, used to determine a data synchronization method based on the data reconciliation results before aligning the data between the customer service knowledge base and the source knowledge base based on the data in the source knowledge base according to the data reconciliation results; the alignment unit includes: an alignment module, used to align the data between the customer service knowledge base and the source knowledge base according to the data synchronization method based on the data in the source knowledge base.

[0015] As an optional solution, the alignment module includes: a first alignment submodule, used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base using an incremental data synchronization method when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to the fifth threshold; and a second alignment submodule, used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base using a full data synchronization method when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold.

[0016] As an optional solution, the above-mentioned apparatus further includes: a third acquisition unit, configured to acquire various types of data in the customer service knowledge base in response to a data preprocessing request before acquiring the knowledge base data set of the customer service knowledge base in response to the data reconciliation request; and a preprocessing unit, configured to preprocess the various types of data to obtain multiple target data before acquiring the knowledge base data set of the customer service knowledge base in response to the data reconciliation request, wherein the knowledge base data set includes multiple target data.

[0017] As an optional solution, the above-mentioned device further includes: a verification unit, used to perform format verification on the above-mentioned types of data during the process of preprocessing the above-mentioned types of data to obtain multiple target data; and a repair unit, used to mark the first data in the above-mentioned types of data if the first data fails the format verification during the process of preprocessing the above-mentioned types of data to obtain multiple target data, and record it in an error log, and repair the format of the first data.

[0018] As an optional solution, the above-mentioned apparatus further includes: a cleaning unit, used to perform a cleaning operation on the above-mentioned data during the process of preprocessing the above-mentioned data to obtain multiple target data, and to remove the second data in the above-mentioned data, wherein the second data is duplicate or invalid data.

[0019] As an optional solution, the above-mentioned device further includes: a fourth acquisition unit, used to acquire the amount of data that has changed during the data alignment process between the customer service knowledge base and the source knowledge base, based on the data reconciliation results and using the data in the source knowledge base as a benchmark; and a display unit, used to display an abnormality prompt message when the amount of data that has changed is greater than or equal to a sixth threshold, after the data reconciliation process between the customer service knowledge base and the source knowledge base is aligned based on the data reconciliation results and using the data in the source knowledge base as a benchmark. The abnormality prompt message is used to indicate that there is an abnormality in the data reconciliation process of the customer service knowledge base.

[0020] As an optional solution, the above-mentioned display unit includes at least one of the following: a first display module for displaying a first prompt message, wherein the first prompt message is used to indicate the degree of impact of the above-mentioned abnormal situation on the above-mentioned customer service knowledge base related business; a second display module for displaying a second prompt message, wherein the second prompt message is used to indicate the scope of impact of the above-mentioned abnormal situation on the above-mentioned customer service knowledge base related business; a third display module for displaying a third prompt message, wherein the third prompt message is used to indicate the description corresponding to the abnormal data covered by the above-mentioned abnormal situation; a fourth display module for displaying a fourth prompt message, wherein the fourth prompt message is used to indicate the cause of the above-mentioned abnormal situation; and a fifth display module for displaying a fifth prompt message, wherein the fifth prompt message is used to indicate the suggested handling measures for the above-mentioned abnormal situation.

[0021] As an optional solution, the above-mentioned device further includes: a correction module, used to correct the erroneous data according to a preset correction method after the above-mentioned display of the abnormality prompt information, when the above-mentioned abnormality prompt information indicates that the abnormality is caused by the input of incorrect data; and a sixth display module, used to display auxiliary prompt information after the above-mentioned display of the abnormality prompt information, when the above-mentioned abnormality prompt information indicates that the abnormality is caused by the change of business rules, wherein the auxiliary prompt information is used to prompt the relevant business rules.

[0022] As an optional solution, the above-mentioned apparatus further includes: a fifth acquisition unit, used to acquire all data in the customer service knowledge base that needs to be changed in the process of aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark; and a change unit, used to perform orderly changes on each piece of data in the above-mentioned data based on the data in the source knowledge base until all the data are aligned with the data in the source knowledge base, wherein the orderly change process allows for tracing the data version of any piece of data in the above-mentioned data.

[0023] According to another aspect of the embodiments of this application, a computer program product or computer program is provided, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the data reconciliation method for the customer service knowledge base described above.

[0024] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the aforementioned data reconciliation method for the customer service knowledge base through the computer program.

[0025] In this embodiment, in response to a data reconciliation request, a knowledge base data set of a customer service knowledge base is obtained. The customer service knowledge base is a knowledge base that provides data support for customer service accounts. The data reconciliation request is used to request the alignment of data between the customer service knowledge base and the source knowledge base. The knowledge base data set is then divided to obtain at least two knowledge base data subsets. The data volume difference between each of the at least two knowledge base data subsets is less than or equal to a preset threshold. Data reconciliation tasks corresponding to each of the knowledge base data subsets are executed in parallel to obtain data reconciliation results. The data reconciliation tasks are used to compare the data differences between the knowledge base data subsets and the source knowledge base. The data reconciliation results represent the data differences between the customer service knowledge base and the source knowledge base. Based on the data reconciliation results, and using the data in the source knowledge base as a benchmark, the data between the customer service knowledge base and the source knowledge base is aligned.

[0026] This embodiment utilizes a series of technical means, such as automated data acquisition, reasonable data segmentation, parallel execution of reconciliation tasks, and automatic data alignment based on results, to enable multiple data subsets to utilize multiple data reconciliation tasks and perform reconciliation operations simultaneously. This achieves automated and parallel processing of customer service knowledge base data reconciliation, thereby improving the data reconciliation efficiency of the customer service knowledge base.

[0027] Furthermore, considering that if computational resources are unevenly distributed during the execution of each data reconciliation task, leading to uneven load, some data reconciliation tasks may be assigned excessively large subsets of data, resulting in heavy processing loads and requiring significant time to complete computation; while other data reconciliation tasks may be assigned smaller subsets of data, completing their tasks quickly. In this situation, the overall processing progress depends on the data reconciliation tasks with large datasets, causing resources for other data reconciliation tasks to be idle and wasted, failing to fully utilize the advantages of parallel processing, significantly reducing overall data processing efficiency, and potentially leading to a decrease in the data reconciliation efficiency of the customer service knowledge base.

[0028] In response, this embodiment also maintains the difference in data volume between subsets of the knowledge base within a reasonable range, ensuring a balanced distribution of computing resources when each data reconciliation task is executed. This achieves the goal of fully leveraging the advantages of parallel processing and improving overall data processing efficiency, thereby improving the technical effect of data reconciliation efficiency of the customer service knowledge base and solving the technical problem of low data reconciliation efficiency of the customer service knowledge base. Attached Figure Description

[0029] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0030] Figure 1 This is a schematic diagram illustrating the application environment of an optional customer service knowledge base data reconciliation method according to an embodiment of this application;

[0031] Figure 2 This is a schematic diagram of the flow of an optional customer service knowledge base data reconciliation method according to an embodiment of this application;

[0032] Figure 3 This is a schematic diagram of an optional customer service knowledge base data reconciliation method according to an embodiment of this application;

[0033] Figure 4 This is a schematic diagram of another optional data reconciliation method for a customer service knowledge base according to an embodiment of this application;

[0034] Figure 5 This is a schematic diagram of another optional data reconciliation method for a customer service knowledge base according to an embodiment of this application;

[0035] Figure 6 This is a schematic diagram of another optional data reconciliation method for a customer service knowledge base according to an embodiment of this application;

[0036] Figure 7 This is a schematic diagram of another optional data reconciliation method for a customer service knowledge base according to an embodiment of this application;

[0037] Figure 8 This is a schematic diagram of an optional customer service knowledge base data reconciliation device according to an embodiment of this application;

[0038] Figure 9 This is a schematic diagram of the structure of an optional electronic device according to an embodiment of this application. Detailed Implementation

[0039] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0040] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0041] According to one aspect of the embodiments of this application, a data reconciliation method for a customer service knowledge base is provided. Optionally, as an optional implementation, the above-mentioned data reconciliation method for a customer service knowledge base may be applied, but is not limited to, to applications such as... Figure 1 The environment shown may include, but is not limited to, user equipment 102 and server 112. User equipment 102 may include, but is not limited to, a display 104, a processor 106 and a memory 108. Server 112 includes a database 114 and a processing engine 116.

[0042] The specific process can be summarized in the following steps:

[0043] In step S102, user equipment 102 obtains a data reconciliation request, wherein the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base;

[0044] Step S104: Send the data reconciliation request to the server 112 via network 110;

[0045] In steps S106-S110, server 112 responds to the data reconciliation request, obtains the knowledge base data set of the customer service knowledge base, and then divides the knowledge base data set by the processing engine 116 to obtain at least two knowledge base data subsets. Furthermore, it executes each data reconciliation task corresponding to each knowledge base data subset in parallel to obtain the data reconciliation results.

[0046] In step S112, the data reconciliation result is sent to the user equipment 102 via the network 110. The user equipment 102, based on the data reconciliation result and using the data in the source knowledge base as a reference, aligns the data between the customer service knowledge base and the source knowledge base via the processor 106, displays the alignment result on the display 104, and stores the above data reconciliation result in the memory 108.

[0047] remove Figure 1Beyond the examples shown, the user equipment described above can be a terminal device configured with a target client, including but not limited to at least one of the following: mobile phones (such as Android phones, iOS phones, etc.), laptops, tablets, PDAs, MIDs (Mobile Internet Devices), PADs, desktop computers, smart TVs, etc. The target client can be a video client, instant messaging client, browser client, educational client, etc. The network described above can include, but is not limited to, wired networks and wireless networks. The wired network includes local area networks (LANs), metropolitan area networks (MANs), and wide area networks (WANs). The wireless network includes Bluetooth, Wi-Fi, and other networks that enable wireless communication. The server described above can be a single server, a server cluster consisting of multiple servers, or a cloud server. The above is merely an example, and no limitations are imposed in this embodiment.

[0048] As an example, the terminal device may include, but is not limited to, mobile phones, computers, smart voice interaction devices, smart home appliances, vehicle terminals, and aircraft. This embodiment can be applied to various scenarios, including but not limited to digital humans, virtual humans, games, virtual reality, and extended reality (XR).

[0049] For example, a server can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0050] Alternatively, as an alternative implementation method, such as Figure 2 As shown, the data reconciliation method for the customer service knowledge base can be performed by an electronic device, such as... Figure 1 The user equipment or server shown includes the following specific steps:

[0051] S202, in response to the data reconciliation request, obtain the knowledge base data set of the customer service knowledge base, where the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base;

[0052] In an optional embodiment, the data reconciliation request can be the starting signal that triggers the entire data reconciliation process, such as being initiated by the system or manually, expressing the need to align the customer service knowledge base with the source knowledge base data.

[0053] In an optional embodiment, the customer service knowledge base can be a knowledge base that provides data support for customer service accounts, containing various data related to customer service work, such as frequently asked questions, product information, service processes, etc.

[0054] In an optional embodiment, the knowledge base dataset may be all data obtained from the customer service knowledge base, which serves as the basis for subsequent data processing.

[0055] Specifically, upon receiving a data reconciliation request, this embodiment initiates the corresponding processing flow. Further, based on the request, relevant data is extracted from the customer service knowledge base to form a knowledge base data set, providing raw data for subsequent data processing.

[0056] By automating the response to requests and the retrieval of data, the tediousness and errors of manual operations are avoided, the efficiency and accuracy of data acquisition are improved, and a foundation is laid for efficient reconciliation in the future.

[0057] S204, the knowledge base data set is divided to obtain at least two knowledge base data subsets, wherein the difference in data volume between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold.

[0058] In an optional embodiment, a subset of knowledge base data can be a smaller data unit obtained by dividing the knowledge base data set, with each subset containing a portion of the original data.

[0059] In an optional embodiment, the preset threshold can be a pre-set upper limit for the difference in data volume, used to control the difference in data volume between various data subsets.

[0060] Specifically, this embodiment employs a reasonable segmentation algorithm, such as hash value modulo or range-based segmentation, to divide the knowledge base dataset into at least two data subsets. During the segmentation process, it is ensured that the difference in data volume between each data subset is less than or equal to a preset threshold.

[0061] The purpose of data partitioning is to achieve a balanced distribution of data, facilitating subsequent parallel processing. By reasonably controlling the difference in data volume, the problem of excessively large data volumes in a certain subset leading to excessively long processing times is avoided. This allows multiple subsets to be processed simultaneously, making full use of computing resources and improving overall processing efficiency.

[0062] S206, execute each data reconciliation task corresponding to each knowledge base data subset in parallel, and obtain the data reconciliation results. The data reconciliation task is used to compare the data differences between the knowledge base data subset and the source knowledge base, and the data reconciliation results are used to represent the data differences between the customer service knowledge base and the source knowledge base.

[0063] In an optional embodiment, the data reconciliation task may be a comparison task performed for each subset of knowledge base data to identify data differences between the subset and the source knowledge base.

[0064] In an optional embodiment, the data reconciliation result may be information recording the data differences between the customer service knowledge base and the source knowledge base, such as which data are inconsistent and the specific content of the differences.

[0065] Specifically, this embodiment utilizes technologies such as multithreading, multiprocessing, or distributed computing to execute multiple data reconciliation tasks simultaneously, with each task processing a subset of knowledge base data. After each data reconciliation task is completed, the corresponding data reconciliation result is obtained.

[0066] Parallel execution of data reconciliation tasks significantly improves processing speed, substantially reducing reconciliation time compared to serial reconciliation methods. By processing multiple data subsets simultaneously, it fully leverages computing resources, improving the efficiency of data reconciliation. The data reconciliation results provide a basis for subsequent data alignment.

[0067] S208, based on the data reconciliation results, aligns the data between the customer service knowledge base and the source knowledge base using the data in the source knowledge base as a benchmark.

[0068] In an alternative embodiment, the source knowledge base may serve as a benchmark for data alignment, containing accurate and authoritative data.

[0069] Specifically, this embodiment analyzes the data reconciliation results to determine the data differences between the customer service knowledge base and the source knowledge base. Based on the analysis results, the data in the customer service knowledge base is updated, corrected, or supplemented using the data in the source knowledge base as a benchmark to ensure consistency with the source knowledge base.

[0070] Data alignment based on reconciliation results ensures the accuracy and consistency of the customer service knowledge base data. Using the source knowledge base as a benchmark ensures the data's authority. The automated data alignment process avoids the tedious work of manual verification and modification, improving the efficiency and accuracy of data alignment, achieving effective alignment between the customer service knowledge base and the source knowledge base, and enhancing the data quality of the customer service knowledge base.

[0071] Optionally, this embodiment can be applied to e-commerce scenarios. Further, it is assumed that the customer service knowledge base and source knowledge base of an e-commerce platform need to reconcile data. After the e-commerce platform completes its daily business, it initiates a data reconciliation request. The system responds to the request, retrieving a set of data from the customer service knowledge base, including product information, promotional activity rules, and frequently asked customer questions. For example, there are 1000 records of product information, 200 records of promotional activity rules, and 500 records of frequently asked customer questions.

[0072] The above dataset is divided into subsets, with a preset threshold of 10 records. Assume the division results in 5 subsets: subset 1 contains 200 product information records, 40 promotional activity rule records, and 100 frequently asked customer questions (FAQ) records; subset 2 contains 200 product information records, 40 promotional activity rule records, and 100 FFAQ records; subset 3 contains 200 product information records, 40 promotional activity rule records, and 100 FFAQ records; subset 4 contains 200 product information records, 40 promotional activity rule records, and 100 FFAQ records; and subset 5 contains 200 product information records, 40 promotional activity rule records, and 100 FFAQ records. The difference in data volume between each subset is less than or equal to the preset threshold, satisfying the requirements for parallel processing.

[0073] For each data subset, a data reconciliation task is performed in parallel. For example, for subset 1, the comparison reveals that the prices of 5 product information entries are inconsistent with the source knowledge base, and the validity periods of 2 promotional activity rules are incorrect. For subset 2, the content of 3 frequently asked customer questions differs from the source knowledge base. Similar comparisons are performed on other subsets, ultimately yielding the data reconciliation results for each subset.

[0074] Based on the data reconciliation results, the customer service knowledge base is updated using the data in the source knowledge base as a benchmark. For example, the five product listings with inconsistent prices in subset 1 are updated to the correct prices in the source knowledge base, and the validity periods of two promotional activity rules are corrected. The content of three frequently asked customer questions in subset 2 is updated. Differences in other subsets are also processed accordingly to ensure consistency between the customer service knowledge base and the source knowledge base.

[0075] Through the above steps, the e-commerce platform's customer service knowledge base and the source knowledge base have completed data reconciliation and alignment, ensuring that the data obtained by customer service personnel (customer service accounts) is accurate and improving the quality of customer service.

[0076] It should be noted that this embodiment automates and parallelizes the data reconciliation process for the customer service knowledge base through a series of steps, including responding to data reconciliation requests to acquire data, appropriately partitioning the data, executing reconciliation tasks in parallel, and automatically aligning the data based on the results. From data acquisition to final alignment, the entire process is efficient and accurate, significantly improving the data reconciliation efficiency of the customer service knowledge base and solving the problem of long processing times associated with traditional manual reconciliation methods when dealing with large-scale data. Simultaneously, by controlling the data volume differences between data subsets, a balanced distribution of computing resources is ensured, fully leveraging the advantages of parallel processing and further enhancing overall data processing efficiency.

[0077] To further illustrate, optional examples include... Figure 3As shown, in response to the data reconciliation request, the knowledge base data set 308 of the customer service knowledge base 304 is obtained. The customer service knowledge base 304 is the knowledge base that provides data support for the customer service account 302. The data reconciliation request is used to request the reconciliation of data between the customer service knowledge base 304 and the source knowledge base 306. The knowledge base data set 308 is then divided to obtain at least two knowledge base data subsets 310 (such as knowledge base data subset 1, knowledge base data subset 2, knowledge base data subset n, etc.). The data volume difference between each knowledge base data subset 310 in at least two knowledge base data subsets 310 is less than or equal to... A preset threshold is set; each data reconciliation task 312 (such as data reconciliation task 1, data reconciliation task 2, data reconciliation task n, etc.) corresponding to each knowledge base data subset 310 is executed in parallel to obtain data reconciliation results 314. Among them, data reconciliation task 312 is used to compare the data differences between the knowledge base data subset 310 and the source knowledge base 306, and data reconciliation result 314 is used to represent the data differences between the customer service knowledge base 304 and the source knowledge base 306. Based on data reconciliation result 314, the data between the customer service knowledge base 304 and the source knowledge base 306 is aligned with the data in the source knowledge base 306 as the benchmark.

[0078] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0079] Through the embodiments provided in this application, a series of technical means such as automated data acquisition, reasonable data segmentation, parallel execution of reconciliation tasks, and automatic data alignment based on results are used to enable multiple data subsets to perform reconciliation operations simultaneously using multiple data reconciliation tasks. This achieves automated and parallel processing of customer service knowledge base data reconciliation, thereby improving the data reconciliation efficiency of the customer service knowledge base.

[0080] Furthermore, considering that if computational resources are unevenly distributed during the execution of each data reconciliation task, leading to uneven load, some data reconciliation tasks may be assigned excessively large subsets of data, resulting in heavy processing loads and requiring significant time to complete computation; while other data reconciliation tasks may be assigned smaller subsets of data, completing their tasks quickly. In this situation, the overall processing progress depends on the data reconciliation tasks with large datasets, causing resources for other data reconciliation tasks to be idle and wasted, failing to fully utilize the advantages of parallel processing, significantly reducing overall data processing efficiency, and potentially leading to a decrease in the data reconciliation efficiency of the customer service knowledge base.

[0081] In this regard, this embodiment also keeps the difference in data volume between subsets of knowledge base data within a reasonable range, ensuring a balanced distribution of computing resources when each data reconciliation task is executed. This achieves the goal of fully leveraging the advantages of parallel processing and improving overall data processing efficiency, thereby realizing the technical effect of improving the data reconciliation efficiency of the customer service knowledge base.

[0082] As an optional approach, data reconciliation tasks corresponding to various subsets of knowledge base data are executed in parallel to obtain the data reconciliation results, including:

[0083] S1-1, execute each data reconciliation task in parallel and obtain the execution results of each data reconciliation task;

[0084] In an optional embodiment, each data reconciliation task can be a comparison task carried out for each subset of knowledge base data, with each task responsible for comparing the data differences between a subset and the source knowledge base.

[0085] In an optional embodiment, the execution result of each task can be the result obtained after each data reconciliation task is completed, reflecting the data differences between the subset of knowledge base data processed by the task and the source knowledge base.

[0086] Specifically, this embodiment utilizes technologies such as multi-threading, multi-processing, or distributed computing to simultaneously launch and run various data reconciliation tasks, making full use of computing resources and shortening the overall processing time. After each data reconciliation task is completed, its execution results are collected and recorded. These results contain specific information about the differences between each subset and the source knowledge base data.

[0087] By executing each data reconciliation task in parallel, data processing efficiency is greatly improved. Traditional serial processing requires completing each task sequentially, which is time-consuming, while parallel execution can process multiple tasks simultaneously, significantly shortening the data reconciliation time. Obtaining the results of each task provides a foundation for subsequent integration to obtain comprehensive data reconciliation results.

[0088] S1-2 integrates the results of each task execution to obtain the data reconciliation results.

[0089] In an optional embodiment, the results of each task execution can be the results obtained by executing each data reconciliation task in parallel, which is the integrated raw material.

[0090] In an optional embodiment, the data reconciliation results can integrate the final results obtained after the execution of various tasks, comprehensively demonstrating the data differences between the customer service knowledge base and the source knowledge base.

[0091] Specifically, this embodiment summarizes, analyzes, and merges the results of each task execution, removes duplicate information, standardizes the format, and forms a complete data reconciliation result. For example, it merges and statistically analyzes identical data differences found in different subsets, and converts differences in different formats into a standardized format. After integration processing, the final data reconciliation result is generated, which clearly shows the inconsistencies between the customer service knowledge base and the source knowledge base in various aspects.

[0092] It's important to note that integrating the results of various tasks is a crucial step in ensuring the comprehensiveness and accuracy of data reconciliation. Since each data reconciliation task targets different subsets of data, their results may exhibit some dispersion and discrepancies. Integration allows these scattered results to be aggregated into a cohesive whole, making the data reconciliation results more readable and usable. The resulting data reconciliation results provide a clear direction for subsequent data alignment, guiding the system to accurately update and correct the customer service knowledge base based on the data in the source knowledge base.

[0093] To illustrate further, it is optional to assume that the customer service knowledge base dataset is divided into three subsets A, B, and C, each corresponding to a data reconciliation task.

[0094] Task 1 processes subset A and finds that the prices of 10 product information entries are inconsistent with the source knowledge base. This is the result of Task 1.

[0095] Task 2 processed subset B and found that the validity periods of 5 promotional activities were incorrect. This is the result of Task 2.

[0096] Task 3 processes subset C and discovers that the processing status of 3 customer feedback items differs from the source knowledge base. This is the result of Task 3.

[0097] By integrating the execution results of tasks 1, 2, and 3, the data reconciliation results are as follows: Compared with the source knowledge base, the customer service knowledge base has 10 product information price discrepancies, 5 promotional activity validity periods are incorrect, and 3 customer feedback processing statuses are different.

[0098] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0099] The embodiments provided in this application execute various data reconciliation tasks in parallel, obtaining the execution results of each task. These results are then integrated to obtain the final data reconciliation result. Parallel execution of the data reconciliation tasks fully utilizes computing resources, significantly shortening the data reconciliation time, especially when processing large-scale data. Furthermore, by integrating the execution results of each task, omissions or duplications that might occur due to decentralized processing are avoided, ensuring the comprehensiveness and accuracy of the data reconciliation result. This facilitates more precise data alignment in subsequent processes, ensuring data consistency between the customer service knowledge base and the source knowledge base, thereby achieving the technical effect of improving the accuracy of data reconciliation in the customer service knowledge base.

[0100] As an optional approach, the method also includes the following steps in the process of executing various data reconciliation tasks in parallel and obtaining the execution results of each corresponding task:

[0101] S2-1, Obtain the first question corresponding to the target reconciliation task, wherein the target reconciliation task is used to compare the data differences between the target data subset and the source knowledge base, each data reconciliation task includes the target reconciliation task, and each knowledge base data subset includes the target data subset;

[0102] In an optional embodiment, the target reconciliation task may be a specific task selected from various data reconciliation tasks for further in-depth analysis, responsible for comparing the data differences between the target data subset and the source knowledge base.

[0103] In an optional embodiment, the first problem may be a specific problem related to the target reconciliation task, used to further explore the data differences between the target data subset and the source knowledge base.

[0104] Specifically, this embodiment determines the target reconciliation task and extracts the corresponding first question by means of preset rules, manual specification, or automatic filtering based on data characteristics. For example, the system automatically selects a task as the target reconciliation task based on the priority of the data reconciliation task or the probability of data discrepancies, and retrieves the matching first question from the question database.

[0105] Identifying the first question corresponding to the target reconciliation task provides direction for subsequent in-depth analysis of data discrepancies. By focusing on specific questions, we can more effectively uncover potential differences between the target data subset and the source knowledge base, improving the accuracy and effectiveness of data reconciliation.

[0106] S2-2, Obtain the target data subset, which is the first answer to the first question;

[0107] In an optional embodiment, the target data subset may be the set of data processed by the target reconciliation task within each knowledge base data subset.

[0108] In an optional embodiment, the first answer may be the answer given by the target data subset to the first question, reflecting the data situation in the target data subset related to the first question.

[0109] Specifically, in this embodiment, based on the requirements of the first question, relevant data is extracted from the target data subset and processed to obtain the first answer. For example, if the first question is "whether the price of goods in the target data subset exceeds 100 yuan", the system will filter out the price information of goods from the target data subset and determine which goods have a price exceeding 100 yuan, thus forming the first answer.

[0110] Obtaining the first answer from the target data subset to the first question provides a basis for comparison with the answer from the source knowledge base. Understanding the target data subset clarifies its performance on a specific question, providing a foundation for discovering data discrepancies.

[0111] S2-3, Obtain the source knowledge base for the second answer to the first question;

[0112] In an optional embodiment, the source knowledge base can be a knowledge base that serves as a data alignment benchmark and contains accurate and authoritative data.

[0113] In an optional embodiment, the second answer may be the answer given by the source knowledge base to the first question, reflecting the data related to the first question in the source knowledge base.

[0114] Specifically, in this embodiment, according to the requirements of the first question, relevant data is retrieved from the source knowledge base, and after processing, a second answer is obtained. For example, for the question "Does the price of the commodity exceed 100 yuan?", the system obtains commodity price information from the source knowledge base, determines which commodity prices exceed 100 yuan, and forms the second answer.

[0115] Obtaining the second answer from the source knowledge base to the first question provides a benchmark for comparison with the first answer from the target data subset. The data in the source knowledge base is considered accurate and reliable; by comparing it with the answer, data discrepancies within the target data subset can be accurately identified.

[0116] S2-4, compare the first answer and the second answer to obtain the information difference of the first answer, wherein the task execution result corresponding to the target reconciliation task includes the information difference of the first answer.

[0117] In an optional embodiment, the first answer may be the answer of the target data subset to the first question.

[0118] In an optional embodiment, the second answer may be the answer from the source knowledge base to the first question.

[0119] In an optional embodiment, the first answer information difference can be the difference information between the first answer and the second answer, reflecting the inconsistency between the target data subset and the source knowledge base on the first question.

[0120] Specifically, this embodiment compares the first and second answers item by item to identify differences. For example, by comparing the lists of goods priced over 100 yuan in the two answers, it identifies goods present in the target data subset but not in the source knowledge base, or goods with inconsistent prices. Based on the comparison results, it generates the first answer information difference, clarifying the data differences between the target data subset and the source knowledge base on the first question.

[0121] It should be noted that by comparing the first and second answers to obtain the information difference of the first answer, the inconsistencies between the target data subset and the source knowledge base on specific issues can be presented concretely and clearly. This information difference enriches the task execution results of the target reconciliation task and provides a more detailed basis for subsequent data alignment.

[0122] To illustrate further, an optional assumption is that the target reconciliation task is responsible for comparing the data differences between the target data subset (containing 100 product information entries) and the source knowledge base. The first question is "whether the product price exceeds 200 yuan".

[0123] First question: The system automatically selects the target reconciliation task and retrieves "Does the price of the goods exceed 200 yuan" from the question database as the first question.

[0124] Obtaining the first answer: By filtering product price information from the target data subset, 20 products with prices exceeding 200 yuan were found, forming the first answer.

[0125] Obtaining a second answer: Product price information was retrieved from the source knowledge base, and 8 products with prices exceeding 200 yuan were found, thus forming a second answer.

[0126] The information difference obtained from the first answer: By comparing the first answer and the second answer, it was found that there are two products in the target data subset with prices exceeding 200 yuan that do not exist in the source knowledge base, or whose prices are inconsistent. This is the information difference from the first answer, which is part of the task execution result corresponding to the target reconciliation task.

[0127] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0128] The embodiments provided in this application obtain a first question corresponding to a target reconciliation task. The target reconciliation task is used to compare data differences between a target data subset and a source knowledge base. Each data reconciliation task includes a target reconciliation task, and each knowledge base data subset includes a target data subset. The process involves obtaining a first answer to the first question from the target data subset; obtaining a second answer to the first question from the source knowledge base; and comparing the first and second answers to obtain a first answer information difference. The task execution result corresponding to the target reconciliation task includes this first answer information difference. By obtaining and comparing answers to specific questions, data differences can be discovered more deeply, avoiding subtle issues that might be missed by only performing an overall comparison. This achieves the goal of improving the accuracy of data reconciliation and thus realizes the technical effect of improving the accuracy of data reconciliation in the customer service knowledge base.

[0129] As an optional approach, the method also includes the following steps in the process of executing various data reconciliation tasks in parallel and obtaining the execution results of each corresponding task:

[0130] S3-1, Obtain the second question corresponding to the target reconciliation task, wherein the second question is related to the first question;

[0131] In an optional embodiment, the target reconciliation task may be a specific task selected from various data reconciliation tasks for further in-depth analysis, responsible for comparing the data differences between the target data subset and the source knowledge base.

[0132] In an optional embodiment, the second question may be an issue associated with the first question corresponding to the target reconciliation task, used to explore the data relationship between the target data subset and the source knowledge base from another perspective.

[0133] Specifically, this embodiment determines a related second question based on the characteristics of the first question, business logic rules, or data analysis needs. For example, if the first question is "whether the price of the product exceeds 200 yuan", then the second question could be "whether the inventory quantity of products with a price exceeding 200 yuan is greater than 10 units". Through the association in business logic, the relationships between the data are further explored.

[0134] Obtaining the second question corresponding to the target reconciliation task provides direction for subsequent analysis of data differences between the target data subset and the source knowledge base from multiple dimensions. By setting associated questions, a more comprehensive understanding of data performance under different business scenarios can be achieved, improving the depth and breadth of data reconciliation.

[0135] S3-2, Obtain the target data subset, for the third answer to the second question;

[0136] In an optional embodiment, the target data subset may be the set of data processed by the target reconciliation task within each knowledge base data subset.

[0137] In an optional embodiment, the third answer may be the answer given by the target data subset to the second question, reflecting the data situation in the target data subset related to the second question.

[0138] Specifically, in this embodiment, according to the requirements of the second question, relevant data is extracted from the target data subset and processed to obtain the third answer. For example, for the question "whether the inventory quantity of goods with a price exceeding 200 yuan is greater than 10 units", goods with a price exceeding 200 yuan are selected from the target data subset, their inventory quantity is counted, and it is determined whether it is greater than 10 units, thus forming the third answer.

[0139] Obtaining the third answer from the target data subset to the second question provides a foundation for comparison with the answers from the source knowledge base and for subsequent analysis. By understanding the performance of the target data subset on related questions, a more comprehensive assessment of its data differences from the source knowledge base can be achieved.

[0140] S3-3, Obtain the source knowledge base for the fourth answer to the second question;

[0141] In an optional embodiment, the source knowledge base can be a knowledge base that serves as a data alignment benchmark and contains accurate and authoritative data.

[0142] In an optional embodiment, the fourth answer may be the answer given by the source knowledge base to the second question, reflecting the data related to the second question in the source knowledge base.

[0143] Specifically, in this embodiment, based on the requirements of the second question, relevant data is retrieved from the source knowledge base, and after processing, a fourth answer is obtained. For example, for the question "whether the inventory quantity of goods with a price exceeding 200 yuan is greater than 10 units", the inventory quantity information of goods with a price exceeding 200 yuan is obtained from the source knowledge base, and it is determined whether it is greater than 10 units, thus forming the fourth answer.

[0144] Obtaining the fourth answer from the source knowledge base to the second question provides a benchmark for comparison with the third answer from the target data subset and for analyzing the consistency of business logic. The data in the source knowledge base is considered accurate and reliable; by comparing it with the answers, data discrepancies and business logic issues within the target data subset can be accurately identified.

[0145] S3-4: Based on the first answer, second answer, third answer, and fourth answer, obtain the associated coherence information corresponding to the target reconciliation task. The associated coherence information is used to represent the coherence of the target data subset and the source knowledge base in terms of related questions and answers in business logic. The task execution result corresponding to the target reconciliation task includes the associated coherence information.

[0146] In an optional embodiment, the association coherence information may be used to represent the coherence of the target data subset and the source knowledge base in terms of the business logic of the associated questions and answers.

[0147] Specifically, this embodiment performs a comprehensive analysis of the first, second, third, and fourth answers to identify the business logic relationships between them. For example, it analyzes the logical relationship between whether the price of a product exceeds 200 yuan (the first question and answer) and whether the inventory quantity of products with a price exceeding 200 yuan is greater than 10 units (the second question and answer), and determines whether the answers of the target data subset and the source knowledge base on these related questions are consistent and conform to the business logic.

[0148] Based on the analysis results, correlation and coherence information is generated to clarify the consistency between the target data subset and the source knowledge base in terms of business logic. If the answers to the related questions in the target data subset and the source knowledge base are consistent and conform to business logic, the correlation and coherence information indicates that the business logic is consistent; if there are inconsistencies or non-compliance with business logic, the correlation and coherence information will point out the specific problems.

[0149] It should be noted that by obtaining relevant and coherent information, the consistency of the target data subset with the source knowledge base in terms of business logic can be comprehensively assessed. This information enriches the task execution results of the target reconciliation task, not only focusing on data differences but also considering the business logic relationships between data, providing a more comprehensive and accurate basis for subsequent data alignment, and helping to improve data quality and the accuracy of business processing.

[0150] To further illustrate, an optional assumption is that the target reconciliation task is responsible for comparing the data differences between the target data subset (containing 100 product information entries) and the source knowledge base.

[0151] The first question is "whether the price of the goods exceeds 200 yuan". Based on the business logic, the second question is "whether the inventory quantity of goods with a price exceeding 200 yuan is greater than 10 pieces".

[0152] From the target data subset, 20 products with prices exceeding 200 yuan were selected, of which 15 had an inventory quantity of more than 10 pieces, forming the third answer.

[0153] Information on the inventory quantity of goods priced over 200 yuan was obtained from the source knowledge base. Eight goods were found to be priced over 200 yuan, of which 16 had an inventory quantity of more than 10 pieces, forming the fourth answer.

[0154] An analysis was conducted combining the first response (the target data subset contains 20 items with prices exceeding 200 yuan, and the source knowledge base contains 8), the second response (compared to the first response), the third response, and the fourth response. It was found that there are discrepancies between the target data subset and the source knowledge base in the number of items with prices exceeding 200 yuan, and also in the number of items with an inventory quantity greater than 10. Further analysis of the business logic suggests potential issues such as data entry errors or untimely updates. This is the relevant and coherent information, which serves as part of the task execution result corresponding to the target reconciliation task.

[0155] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0156] The embodiments provided in this application obtain a second question corresponding to a target reconciliation task, wherein the second question is associated with a first question; a target data subset is obtained, along with a third answer to the second question; a source knowledge base is obtained, along with a fourth answer to the second question; based on the first, second, third, and fourth answers, correlation and coherence information corresponding to the target reconciliation task is obtained, wherein the correlation and coherence information represents the business logic coherence between the target data subset and the source knowledge base in related questions and answers, and the task execution result corresponding to the target reconciliation task includes the correlation and coherence information. By processing related questions and analyzing data differences from multiple dimensions, more potential data problems can be discovered, improving the comprehensiveness of data reconciliation. Furthermore, the correlation and coherence information considers the business logic relationships between data, helping to more accurately locate and correct data problems, thereby achieving the goal of improving the accuracy of data alignment, and thus realizing the technical effect of improving the accuracy of data reconciliation in the customer service knowledge base.

[0157] As an optional approach, before retrieving the knowledge base dataset of the customer service knowledge base in response to a data reconciliation request, the method also includes:

[0158] A data reconciliation request is obtained when at least one reconciliation trigger condition is met, wherein at least one of the following is included: reconciliation frequency condition, data change condition of customer service knowledge base, and customer service interaction condition of customer service knowledge base.

[0159] In an optional embodiment, at least one of the following reconciliation trigger conditions includes: reconciliation frequency condition, data change condition of customer service knowledge base, and customer service interaction condition of customer service knowledge base. It can be understood that when any one or more of the following conditions are met, it is considered that at least one reconciliation trigger condition is met.

[0160] To further illustrate, you can set a condition that meets the reconciliation frequency requirement to be considered as meeting at least one reconciliation trigger condition. Alternatively, you can set a condition that meets the data change requirement in the customer service knowledge base to be considered as meeting at least one reconciliation trigger condition. Or, you can set a condition that meets the customer service interaction requirement in the customer service knowledge base to be considered as meeting at least one reconciliation trigger condition.

[0161] You can set a condition that considers both reconciliation frequency requirements and data change requirements in the customer service knowledge base as fulfilling at least one reconciliation trigger condition. Alternatively, you can set a condition that considers both reconciliation frequency requirements and customer service interaction requirements in the customer service knowledge base as fulfilling at least one reconciliation trigger condition. Or, you can set a condition that considers both reconciliation frequency requirements, data change requirements in the customer service knowledge base, and customer service interaction requirements in the customer service knowledge base as fulfilling at least one reconciliation trigger condition.

[0162] In an optional embodiment, the reconciliation trigger condition can be a set of conditions used to determine whether to initiate a data reconciliation request, which may include various different types of conditions.

[0163] In an optional embodiment, the reconciliation frequency condition can be a trigger condition for data reconciliation set based on a time period, such as reconciliation once a day, once a week, or once a month.

[0164] In an optional embodiment, the data change condition of the customer service knowledge base can be a condition that triggers data reconciliation when specific changes occur in the data of the customer service knowledge base, such as the addition, modification or deletion of a certain number of knowledge entries.

[0165] In an optional embodiment, the customer service interaction conditions of the customer service knowledge base can be conditions that trigger data reconciliation based on specific situations generated during customer service interactions with customers, such as the number of times customer feedback on inaccurate knowledge base information reaches a certain threshold.

[0166] Specifically, in this embodiment, timing is performed according to a preset time period. When the set reconciliation time is reached, the reconciliation frequency condition is met, and a data reconciliation request is triggered.

[0167] Alternatively, monitor data changes in the customer service knowledge base in real time. When data changes are detected that meet preset conditions, such as more than 10 new knowledge entries, the data change conditions are met, and a data reconciliation request is triggered.

[0168] Alternatively, collect feedback information from customer service interactions. When the number of times customer feedback on inaccurate knowledge base information is counted reaches 5, the customer service interaction conditions are met, triggering a data reconciliation request.

[0169] It should be noted that these three reconciliation trigger conditions cover the data reconciliation needs from different perspectives. The reconciliation frequency condition ensures the regularity of reconciliation, the data change condition ensures timely reconciliation when significant data changes occur, and the customer service interaction condition triggers reconciliation from the perspective of user feedback. Together, they constitute a comprehensive and flexible reconciliation trigger mechanism that can adapt to different business scenarios and data changes.

[0170] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0171] Through the embodiments provided in this application, a data reconciliation request is obtained when at least one reconciliation trigger condition is met. The at least one reconciliation trigger condition includes at least one of the following: reconciliation frequency condition, data change condition of customer service knowledge base, and customer service interaction condition of customer service knowledge base. By using different reconciliation trigger conditions, data reconciliation can be accurately initiated according to actual business needs and data changes, thereby avoiding blind reconciliation and achieving the technical effect of improving the data reconciliation efficiency of customer service knowledge base.

[0172] As an optional approach, the method also includes the following steps before obtaining the data reconciliation request:

[0173] S4-1, if the amount of data added or modified in the customer service knowledge base within a preset time period is greater than or equal to the first threshold, it is determined that the first triggering condition is met, wherein the first triggering condition includes the data change condition of the customer service knowledge base, and at least one reconciliation triggering condition includes the first triggering condition.

[0174] In an optional embodiment, the first triggering condition includes the data change condition of the customer service knowledge base. It can be understood that when the data change condition of the customer service knowledge base is met, it is considered that the first triggering condition is met. The data change condition of the customer service knowledge base may include the amount of data added or modified in the customer service knowledge base within a preset time period being greater than or equal to a first threshold.

[0175] In an optional embodiment, the preset time period can be a pre-defined time range for statistical analysis of changes in customer service knowledge base data, such as one day or one week.

[0176] In an optional embodiment, the amount of data added or modified can be the sum of the number of knowledge entries added and the number of knowledge entries modified in the customer service knowledge base within a preset time period.

[0177] In an optional embodiment, the first threshold may be a set threshold used to determine whether the change in data volume has reached a critical value that needs to be triggered for reconciliation, such as 50 records.

[0178] Specifically, in this embodiment, the amount of new or modified data in the customer service knowledge base is continuously counted within a preset time period, and the counted amount of new or modified data is compared with a first threshold. If the amount of new or modified data is greater than or equal to the first threshold, the system determines that at least one reconciliation trigger condition is met.

[0179] S4-2, if the reconciliation time interval of the customer service knowledge base is greater than or equal to the second threshold, and the number of customer service interactions in the customer service knowledge base is greater than or equal to the third threshold, it is determined that the second triggering condition is met. The second triggering condition may include the reconciliation frequency condition and the customer service interaction condition of the customer service knowledge base. At least one reconciliation triggering condition may include the second triggering condition.

[0180] In an optional embodiment, the second triggering condition includes a reconciliation frequency condition and a customer service interaction condition of the customer service knowledge base. It can be understood that the second triggering condition is considered to be met when both the reconciliation frequency condition and the customer service interaction condition of the customer service knowledge base are met. Meeting the reconciliation frequency condition includes the reconciliation time interval of the customer service knowledge base being greater than or equal to a second threshold, and meeting the customer service interaction condition of the customer service knowledge base includes the number of customer service interactions of the customer service knowledge base being greater than or equal to a third threshold.

[0181] In an optional embodiment, the reconciliation time interval can be the time difference between the last data reconciliation completion time and the current time.

[0182] In an optional embodiment, the second threshold may be a set critical value used to determine whether the reconciliation time interval is too long, such as 15 days.

[0183] In an optional embodiment, the number of customer service interactions can be the number of times a customer service representative interacts with a customer within a certain period of time.

[0184] In an optional embodiment, the third threshold may be a set threshold for determining whether the number of customer service interactions has reached a critical value that requires triggering reconciliation, such as 100 times.

[0185] Specifically, this embodiment calculates the reconciliation time interval between the last data reconciliation completion time and the current time, and counts the number of interactions between customer service and customers within the corresponding time interval. The reconciliation time interval is compared with a second threshold, and the number of customer service interactions is compared with a third threshold. When the reconciliation time interval is greater than or equal to the second threshold, and the number of customer service interactions is greater than or equal to the third threshold, the system determines that at least one reconciliation trigger condition is met.

[0186] It should be noted that this embodiment comprehensively considers both the reconciliation time interval and customer service interaction. A longer reconciliation time interval may lead to the accumulation of data discrepancies, while a higher number of customer service interactions may indicate that customers have encountered numerous problems while using the knowledge base, possibly due to inaccurate or incomplete data. By combining these two conditions, a more comprehensive judgment can be made as to whether data reconciliation needs to be initiated, thereby improving the rationality and effectiveness of data reconciliation.

[0187] To illustrate further, consider a hypothetical timeframe of one week and a first threshold of 30 entries. In a given week, due to course updates, the customer service knowledge base adds 20 knowledge entries and modifies 15 others, totaling 35 entries. This exceeds the first threshold of 30, and the system determines that the reconciliation trigger condition is met, thus initiating a data reconciliation request.

[0188] Alternatively, suppose the second threshold is set to 20 days and the third threshold is set to 150 times. The last data reconciliation was completed 25 days ago, and customer service interacted with students 80 times in the past week. At this point, the reconciliation interval of 25 days is greater than the second threshold of 20 days, and the number of customer service interactions (80) is greater than the third threshold of 150. The system determines that the reconciliation trigger conditions are met and requests a data reconciliation request.

[0189] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0190] Through the embodiments provided in this application, when the amount of data added or modified in the customer service knowledge base within a preset time period is greater than or equal to a first threshold, it is determined that at least one reconciliation trigger condition is met; or, when the reconciliation time interval of the customer service knowledge base is greater than or equal to a second threshold and the number of customer service interactions in the customer service knowledge base is greater than or equal to a third threshold, it is determined that at least one reconciliation trigger condition is met. By setting specific thresholds, it is possible to more accurately determine when data reconciliation is needed, thereby avoiding excessive reconciliation or untimely reconciliation, and thus achieving the technical effect of improving the timeliness of data reconciliation in the customer service knowledge base.

[0191] As an optional approach, before splitting the knowledge base dataset to obtain at least two subsets of knowledge base data, the method also includes: determining the data splitting method based on the data type of the knowledge base dataset;

[0192] The knowledge base dataset is divided into at least two subsets, including: dividing the knowledge base dataset according to the data partitioning method to obtain at least two subsets.

[0193] In an optional embodiment, the knowledge base dataset may be a collection of all data obtained from the customer service knowledge base, containing various types of data.

[0194] In an optional embodiment, the data type can be different categories of data in the knowledge base dataset, such as text data, numerical data, date data, etc.

[0195] In an optional embodiment, the data segmentation method can be a specific method for dividing a data set into multiple subsets based on the data type.

[0196] Specifically, this embodiment performs a detailed analysis of the data in the knowledge base dataset to identify the different data types it contains. Based on the identified data types, combined with the characteristics of the data and processing requirements, an appropriate data segmentation method is selected. For example, for text data, segmentation may be based on text length or keywords; for numerical data, segmentation may be based on the range of values.

[0197] The knowledge base dataset is then further segmented according to the determined data segmentation method. After the segmentation, at least two subsets of the knowledge base data are obtained.

[0198] It should be noted that this embodiment determines the data segmentation method by considering the data type, which can fully utilize the characteristics of the data and make the data segmentation more reasonable and effective. Different data types may have different distribution patterns and processing requirements; selecting a segmentation method accordingly can improve the efficiency and quality of data processing.

[0199] To illustrate further, consider the optional assumption that the customer service knowledge base dataset of an e-commerce platform contains the following data types: product information (text data), product price (numerical data), and product listing time (date data).

[0200] For product information (text data), due to the large differences in text length, we choose to split it according to text length, grouping product information with a text length of less than 50 characters into one group and those with a length of 50 characters or more into another group.

[0201] For product prices (numerical data), they are divided according to price range: products priced less than 100 yuan are grouped together, products priced between 100 and 500 yuan are grouped together, and products priced greater than 500 yuan are grouped together.

[0202] For product listing time (date data), it is divided by month, and products listed in different months are divided into different groups.

[0203] Based on the data segmentation method determined above, the customer service knowledge base dataset is segmented. This results in multiple subsets of the knowledge base data, such as subsets containing short text product information, long text product information, low-priced product data, mid-priced product data, high-priced product data, and subsets of products listed in different months.

[0204] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0205] The embodiments provided in this application determine the data segmentation method based on the data type of the knowledge base dataset. According to the segmentation method, the knowledge base dataset is divided into at least two subsets. A reasonable data segmentation method allows for more parallel data processing, reducing processing time and improving overall processing efficiency. Furthermore, using different segmentation methods for different types of data better meets the processing needs of different data types, thereby improving the accuracy and quality of data processing and achieving the technical effect of improving the data reconciliation efficiency of the customer service knowledge base.

[0206] As an alternative approach, the knowledge base dataset is divided according to a data partitioning method to obtain at least two subsets of the knowledge base data, including:

[0207] S5-1, when the amount of data of different data types in the knowledge base data set is greater than or equal to the fourth threshold, the knowledge base data set is divided by a partitioning algorithm based on hash value modulo to obtain at least two knowledge base data subsets;

[0208] In an optional embodiment, the knowledge base dataset can be a collection of all data containing multiple data types.

[0209] In an optional embodiment, the amount of data for different data types can be the amount of data contained in each data type in the knowledge base dataset.

[0210] In an optional embodiment, the fourth threshold may be a critical value used to determine the size of data of different data types.

[0211] In an optional embodiment, the hash value modulo partitioning algorithm can be a partitioning method that determines the subset to which the data belongs by calculating the hash value of the data and taking the modulo of the hash value.

[0212] Specifically, this embodiment counts the amount of data of different data types in the knowledge base dataset and compares it with a fourth threshold. When the amount of data of different data types is greater than or equal to the fourth threshold, a hash value modulo segmentation algorithm is selected. The hash value modulo segmentation algorithm is used to segment the knowledge base dataset, calculate the hash value of each data, and take the modulo of the hash value with a preset modulo. Data with the same modulo result are assigned to the same knowledge base dataset subset.

[0213] When dealing with large volumes of data of different data types, hash value modulo partitioning algorithms can evenly distribute the data across subsets, avoiding data skew. This algorithm is highly efficient and random, enabling rapid data partitioning and making it suitable for large-scale data processing scenarios.

[0214] S5-2, when the amount of data of different data types in the knowledge base dataset is less than the fourth threshold, the time range-based segmentation algorithm is used to segment the knowledge base dataset to obtain at least two knowledge base dataset subsets.

[0215] In an optional embodiment, the time range segmentation algorithm can be a method that segments data according to different time ranges based on the time attributes of the data.

[0216] Specifically, this embodiment counts the amount of data of different data types in the knowledge base dataset and compares it with a fourth threshold. When the amount of data of different data types is less than the fourth threshold, a time range segmentation algorithm is selected. The time range segmentation algorithm is used to segment the knowledge base dataset, dividing the data into different time ranges based on the data's time attributes (such as creation time, update time, etc.), forming multiple knowledge base data subsets.

[0217] When the amount of data of different data types is small, time-range-based segmentation algorithms can better reflect the temporal characteristics of the data, facilitating time-series analysis. This algorithm is simple, intuitive, and easy to implement, making it suitable for data processing scenarios where the data volume is relatively small but the time attribute is important.

[0218] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0219] Through the embodiments provided in this application, when the amount of data of different data types in the knowledge base dataset is greater than or equal to a fourth threshold, a hash value modulo-based segmentation algorithm is used to segment the knowledge base dataset, resulting in at least two knowledge base dataset subsets. When the amount of data of different data types in the knowledge base dataset is less than the fourth threshold, a time range-based segmentation algorithm is used to segment the knowledge base dataset, resulting in at least two knowledge base dataset subsets. Choosing different segmentation algorithms based on the amount of data allows for adaptation to datasets of different sizes and characteristics. For example, using a hash value modulo algorithm for large-scale data enables fast and uniform segmentation; using a time range-based segmentation algorithm for small-scale data fully utilizes the temporal characteristics of the data, improving data processing efficiency, thereby enhancing the flexibility and adaptability of data segmentation and achieving the technical effect of improving the targeting of customer service knowledge base data reconciliation.

[0220] As an optional approach, before aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results, the method also includes: determining the data synchronization method based on the data reconciliation results;

[0221] Based on the data reconciliation results, and using the data in the source knowledge base as a benchmark, the data between the customer service knowledge base and the source knowledge base is aligned. This includes aligning the data between the customer service knowledge base and the source knowledge base according to the data synchronization method, using the data in the source knowledge base as a benchmark.

[0222] In an optional embodiment, the data synchronization method may be a specific method for aligning customer service knowledge base data with source knowledge base data, selected based on the data reconciliation results.

[0223] Specifically, this embodiment analyzes the data reconciliation results to understand the types, extent, and scope of data differences between the customer service knowledge base and the source knowledge base. Based on the reconciliation results obtained from the analysis, and considering the characteristics of the data and business needs, an appropriate data synchronization method is selected. For example, if the differences are small and mainly involve updates to some fields, incremental synchronization can be chosen; if the differences are large, a full synchronization may be necessary.

[0224] Following the established data synchronization method, and using the data in the source knowledge base as a benchmark, operations such as modification, addition, or deletion are performed on the data in the customer service knowledge base. After the synchronization operation, the data between the customer service knowledge base and the source knowledge base reaches a consistent state.

[0225] It's important to note that aligning data according to the data synchronization method ensures that the customer service knowledge base accurately reflects the content of the source knowledge base. A reasonable synchronization method can avoid unnecessary data updates, improve data synchronization efficiency, and guarantee data consistency and integrity, providing accurate knowledge support for customer service.

[0226] To illustrate further, consider the optional assumption that after reconciling the customer service knowledge base of an e-commerce platform with the source knowledge base, the following reconciliation results are obtained: some product information has been updated (such as price and inventory), some product information is missing in the customer service knowledge base, and some product information is outdated in the customer service knowledge base.

[0227] For product information with updated fields, select the incremental synchronization method to update only the fields that have changed.

[0228] For missing product information, a full synchronization method is used to add the complete product information from the source knowledge base to the customer service knowledge base.

[0229] For outdated product information, first delete the outdated information from the customer service knowledge base, and then retrieve the latest information from the source knowledge base to add it.

[0230] Update the price and inventory fields of product information in the customer service knowledge base using incremental synchronization.

[0231] By using a full synchronization method, the missing product information in the source knowledge base is completely added to the customer service knowledge base.

[0232] First, delete outdated product information from the customer service knowledge base, then retrieve the latest product information from the source knowledge base and add it to the customer service knowledge base.

[0233] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0234] The embodiments provided in this application determine the data synchronization method based on the data reconciliation results. Following this method, the data in the source knowledge base is aligned with the data in the customer service knowledge base. Using a suitable data synchronization method ensures a high degree of consistency between the customer service knowledge base and the source knowledge base, reducing problems caused by data discrepancies. Furthermore, selecting an appropriate data synchronization method based on the reconciliation results avoids unnecessary full synchronization, reducing the workload of data transmission and processing, thereby improving resource utilization efficiency and achieving the technical effect of improving the data reconciliation efficiency of the customer service knowledge base.

[0235] As an optional solution, based on the data synchronization method, the data between the customer service knowledge base and the source knowledge base is aligned, including:

[0236] S6-1, If ​​the data reconciliation results indicate that the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to the fifth threshold, the incremental data synchronization method is adopted to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base.

[0237] In an optional embodiment, the fifth threshold may be a critical value used to determine the magnitude of data differences.

[0238] In an optional embodiment, incremental data synchronization can be performed by synchronizing only the changed difference data between the source knowledge base and the customer service knowledge base.

[0239] Specifically, this embodiment analyzes the data reconciliation results to determine whether the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to a fifth threshold. When the data difference is greater than or equal to the fifth threshold, the incremental data synchronization method is selected. Data in the source knowledge base that differs from the customer service knowledge base is identified. The differing data from the source knowledge base is then added to the corresponding knowledge base data subset in the customer service knowledge base.

[0240] When data discrepancies are significant, incremental data synchronization can avoid the large data transfer and processing overhead of full synchronization. Synchronizing only the discrepancies allows for precise updates to the customer service knowledge base, improving data alignment efficiency while reducing system resource consumption.

[0241] S6-2, if the data reconciliation results indicate that the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold, a full data synchronization method is adopted to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base.

[0242] In an optional embodiment, the full data synchronization method can be to synchronize all data (or data within a specific range) in the source knowledge base to the customer service knowledge base.

[0243] Specifically, this embodiment analyzes the data reconciliation results to determine whether the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold. When the data difference is less than the fifth threshold, the full data synchronization method is selected. All data (or the portion containing the differing data) of the corresponding knowledge base data subset in the source knowledge base is obtained. The data in the source knowledge base is then overwritten and updated to the corresponding knowledge base data subset in the customer service knowledge base.

[0244] When data differences are minor, full data synchronization can ensure complete consistency between the customer service knowledge base and the source knowledge base. Although full synchronization may involve some data transfer and processing, the overhead is acceptable due to the small differences, and it avoids minor differences that might be missed by incremental synchronization, thus improving the accuracy of data alignment.

[0245] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0246] Through the embodiments provided in this application, when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to the fifth threshold, an incremental data synchronization method is used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base; when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold, a full data synchronization method is used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base. By selecting an appropriate synchronization method according to the size of the data difference, unnecessary full synchronization is avoided, thereby reducing the time for data transmission and processing, and thus achieving the technical effect of improving the data reconciliation efficiency of the customer service knowledge base.

[0247] As an optional approach, before retrieving the knowledge base dataset of the customer service knowledge base in response to a data reconciliation request, the method also includes:

[0248] S7-1, responding to data preprocessing requests, retrieves various types of data from the customer service knowledge base;

[0249] In an optional embodiment, the various types of data can be different types of raw data from the customer service knowledge base, such as text, numerical values, dates, etc.

[0250] Specifically, in this embodiment, upon receiving a data preprocessing request, various types of raw data are extracted from the customer service knowledge base according to the request.

[0251] S7-2 preprocesses various types of data to obtain multiple target data, including the knowledge base dataset which contains multiple target data.

[0252] In an optional embodiment, preprocessing can be a process of cleaning, transforming, normalizing, or performing other operations on the original data.

[0253] In an optional embodiment, the target data may be preprocessed data suitable for data reconciliation.

[0254] In an optional embodiment, the knowledge base dataset can be a collection of multiple target data for subsequent data reconciliation operations.

[0255] Specifically, this embodiment can remove invalid data such as duplicates, errors, and missing values ​​from various types of data. For example, it can delete garbled characters and empty values ​​from text data, and correct incorrect formats in numerical data. It can also convert data into a unified format or type. For example, it can convert date data in different formats into a standard date format, and expand abbreviations in text data. Furthermore, it can standardize the data to make it comparable. For example, it can normalize numerical data to keep it within a specific range. After the above preprocessing operations, multiple target data are obtained. These multiple target data are then combined into a knowledge base dataset.

[0256] To further illustrate, consider the optional assumption that an e-commerce platform's customer service knowledge base contains the following types of data:

[0257] Product information includes fields such as product name, description, and price. Some product descriptions contain garbled characters, and the price format is inconsistent.

[0258] Customer consultation records: These include fields such as consultation time, consultation content, and customer feedback. Duplicate records and empty values ​​exist.

[0259] Order information includes fields such as order number, order time, and order status. The order time can be formatted in various ways.

[0260] For product information, delete product description records containing garbled characters, correct price formats, and standardize them to two decimal places.

[0261] For customer inquiry records, remove duplicate records and fill in blank values ​​(e.g., fill in blank reviews with "None").

[0262] For order information, delete invalid order number records.

[0263] The price unit in the product information will be standardized to "yuan".

[0264] Convert the consultation time in the customer consultation record to a standard date and time format.

[0265] The order placement time in the order information will be uniformly converted to the format "YYYY-MM-DDHH:MM:SS".

[0266] Normalize the prices in the product information to a range of 0-10000 (assuming that most prices fall within this range).

[0267] After preprocessing, standardized product information, customer inquiry records, and order information are obtained as target data. This target data is then combined into a knowledge base dataset for subsequent data reconciliation.

[0268] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0269] The embodiments provided in this application, in response to a data preprocessing request, acquire various types of data from the customer service knowledge base; preprocess these various types of data to obtain multiple target data, wherein the knowledge base data set includes multiple target data. Through data preprocessing, invalid data is removed, and data formats and types are standardized, thereby improving the accuracy and consistency of the data, and thus achieving the goal of providing high-quality data for subsequent data reconciliation, thereby realizing the technical effect of improving the accuracy of data reconciliation.

[0270] As an optional approach, in the process of preprocessing various types of data to obtain multiple target data, the method also includes:

[0271] S8-1 performs format validation on various types of data;

[0272] In an optional embodiment, format validation may involve checking whether the data conforms to a predefined format standard.

[0273] Specifically, this embodiment pre-defines the format requirements that various types of data should meet based on the data type and business needs. For example, date data should conform to the "YYYY-MM-DD" format, and numerical data should retain two decimal places. It iterates through various types of data, comparing each piece of data with the set format standards to determine whether it passes the format validation.

[0274] S8-2: If the first data in any data category fails the format validation, mark the first data and record it in the error log, and then perform format repair on the first data.

[0275] In an optional embodiment, the first data may be specific data that failed the format validation among various types of data.

[0276] In an optional embodiment, the marker may be used to specially identify data that fails the verification for subsequent processing.

[0277] In an optional embodiment, the error log may be a file used to record data that fails format validation and related information.

[0278] In an alternative embodiment, format repair may involve converting data that does not conform to the format standard into a compliant format.

[0279] Specifically, this embodiment checks the format validation result of each piece of data to determine whether it is the first piece of data that failed validation. For the first piece of data that failed validation, a specific marker is added, such as setting a special field value or adding a comment. Detailed information about the first piece of data, including data content, location, and the format standard that failed, is recorded in the error log.

[0280] Further analysis of the error log reveals the specific reasons why the first data failed format validation, such as incorrect date format or incorrect decimal places. Appropriate repair methods are selected for different error causes. For example, for incorrect date format, date formatting functions can be used for conversion; for incorrect decimal places, rounding or truncation can be performed. Following the selected repair methods, the first data is formatted to conform to predefined format standards.

[0281] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0282] The embodiments provided in this application perform format validation on various types of data. If the first piece of data fails the format validation, it is marked and recorded in an error log, and its format is then repaired. Through format validation, marking, recording, and repair, the data is ensured to conform to predefined format standards, thereby improving data accuracy and consistency, and ultimately enhancing the accuracy of data reconciliation in the customer service knowledge base.

[0283] As an optional approach, in the process of preprocessing various types of data to obtain multiple target data, the method also includes:

[0284] Perform cleaning operations on various types of data, removing the second data in each type of data, where the second data is duplicate or invalid data.

[0285] In an optional embodiment, the cleaning operation can be a series of actions to process duplicates or invalid data.

[0286] In an optional embodiment, the second data may be data that is specifically identified as duplicate or invalid among various types of data.

[0287] In an alternative embodiment, the removal operation may be an action of deleting the second data from the original dataset.

[0288] Specifically, this embodiment reduces data redundancy and errors, improves data accuracy and consistency, and makes subsequent data processing and analysis more efficient and reliable by identifying and removing duplicate or invalid data. Removing duplicate or invalid data is the core objective of data cleaning. This operation optimizes data quality, reduces data storage space usage, improves data processing efficiency, and avoids analytical biases and decision-making errors caused by duplicate or invalid data.

[0289] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0290] The embodiments provided in this application perform data cleaning operations on various types of data, removing second data, which is duplicate or invalid data. After removing duplicate and invalid data, the data becomes more accurate and complete, reducing noise and errors, thereby improving the overall quality of the data and achieving the technical effect of improving the accuracy of data reconciliation in the customer service knowledge base.

[0291] As an optional approach, after aligning the customer service knowledge base with the source knowledge base based on the data reconciliation results, the method also includes:

[0292] S9-1, Obtain the amount of data that has changed during the process of aligning the customer service knowledge base with the source knowledge base;

[0293] In an optional embodiment, the data alignment process may be an operation of matching, adjusting, and synchronizing customer service knowledge base data with source knowledge base data.

[0294] In an optional embodiment, the amount of data changed can be the number of data records in the customer service knowledge base that have been modified, added, or deleted during the alignment process.

[0295] Specifically, in this embodiment, during the alignment operation, changes in the data in the customer service knowledge base are monitored in real time or periodically, and the changed data records are recorded. The number of monitored changed data records is counted to obtain the amount of data that has changed.

[0296] Determining the amount of data changes is fundamental for subsequently judging whether data reconciliation is abnormal. By accurately calculating the amount of changes, the impact of data reconciliation operations on the data can be quantified, providing data support for anomaly detection.

[0297] S9-2, If the amount of data that has been changed is greater than or equal to the sixth threshold, an abnormal prompt message will be displayed. The abnormal prompt message is used to indicate that there is an abnormality in the data reconciliation process of the customer service knowledge base.

[0298] In an optional embodiment, the sixth threshold may be a pre-set threshold used to determine whether the amount of data change is abnormal.

[0299] In an optional embodiment, the error message may be a notification to inform relevant personnel that there is an anomaly in the data reconciliation process.

[0300] Specifically, this embodiment compares the amount of data that has undergone changes with a sixth threshold. If the amount of data that has undergone changes is greater than or equal to the sixth threshold, an anomaly is determined in the data reconciliation process. When an anomaly is determined, an anomaly message is displayed on the system interface or relevant output channels to alert relevant personnel.

[0301] By comparing changes to thresholds and displaying anomaly alerts, anomalies in the data reconciliation process can be detected promptly. This helps relevant personnel take swift action to investigate the root causes of problems and prevent adverse impacts on data quality and business operations.

[0302] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0303] The embodiments provided in this application obtain the amount of data that has changed during the data alignment process between the customer service knowledge base and the source knowledge base. If the amount of data that has changed is greater than or equal to a sixth threshold, an abnormal prompt message is displayed. The abnormal prompt message is used to indicate that there is an abnormality in the data reconciliation process of the customer service knowledge base. By monitoring the amount of data changes and setting a threshold for judgment, it is possible to promptly detect potential data errors, synchronization anomalies, and other problems that may exist during the data reconciliation process, and avoid the accumulation of problems leading to more serious consequences.

[0304] As an optional solution, display an error message, including at least one of the following:

[0305] S10-1, Display the first prompt information, wherein the first prompt information is used to indicate the degree of impact of the abnormal situation on the customer service knowledge base related business;

[0306] In an optional embodiment, the first prompt information is specifically designed to explain the extent to which the abnormal situation affects the customer service knowledge base-related business.

[0307] In an optional embodiment, the customer service knowledge base is associated with various business operations closely related to the customer service knowledge base data, such as customer service response and business consultation answers.

[0308] In an optional embodiment, the degree of impact refers to the specific extent of the impact of the anomaly on related business operations in terms of efficiency, accuracy, and customer satisfaction.

[0309] Specifically, this embodiment conducts a comprehensive analysis of abnormal situations, assessing their impact on customer service knowledge base-related businesses across different dimensions, such as potential increases in customer service response time or higher error rates in answering business inquiries. The degree of impact is quantified or graded, using terms like "slight," "moderate," or "severe," or specific numerical impact indicators, such as "average customer service response time increased by 30%." The analyzed impact information is then displayed to relevant personnel in a clear and easy-to-understand manner.

[0310] Displaying the first alert message allows relevant personnel to intuitively understand the impact of the anomaly on the business, thereby prioritizing the handling of anomalies with a significant impact on the business, allocating resources rationally, and ensuring that critical business processes are not severely affected.

[0311] S10-2, Display the second prompt information, wherein the second prompt information is used to indicate the scope of the impact of the abnormal situation on the customer service knowledge base related business;

[0312] In an optional embodiment, the second prompt information may be used to clarify the specific scope of customer service knowledge base-related services affected by the abnormal situation.

[0313] In an optional embodiment, the scope of impact can be defined as the scope of the business processes, business modules, customer groups, etc. involved.

[0314] Specifically, this embodiment identifies the specific business processes affected by the anomaly, such as which customer service processes are affected, which business modules experience data anomalies, and which customer groups are involved. A detailed description of the scope of impact is provided, for example, "This anomaly affected the online customer service consultation module, involving some customer groups in East China." This defined and described scope of impact information is then presented to relevant personnel.

[0315] Displaying the second notification helps relevant personnel fully understand the scope of the anomaly, enabling them to take targeted measures to prevent further escalation of the impact. It also provides a clear direction for subsequent problem investigation and resolution.

[0316] S10-3, Display the third prompt information, which is used to prompt the description of the abnormal data covered by the abnormal situation;

[0317] In an optional embodiment, the third prompt information may be a detailed description of the abnormal data covered by the abnormal situation.

[0318] In an optional embodiment, the abnormal data may be problematic data records discovered during the data reconciliation process.

[0319] In an optional embodiment, the data description may be a detailed description of the characteristics, content, source, etc. of the abnormal data.

[0320] Specifically, this embodiment collects relevant information about abnormal data, including field values, data sources, and data change history. The collected information is then organized to form a clear and accurate description of the abnormal data. This organized description of the abnormal data is presented to relevant personnel, for example, listing the specific fields and values ​​of the abnormal data.

[0321] Displaying a third prompt message allows relevant personnel to directly understand the specifics of the abnormal data, helping to quickly locate the problematic data and providing a foundation for further data repair and analysis.

[0322] S10-4, Display the fourth prompt message, which is used to indicate the reason for the abnormal situation;

[0323] In an optional embodiment, the fourth prompt message may be used to explain the reason for the abnormal situation.

[0324] In an optional embodiment, the cause could be any of the factors that cause data reconciliation anomalies, such as data entry errors or system synchronization failures.

[0325] Specifically, this embodiment uses technical means and manual analysis to conduct an in-depth investigation of the anomaly and identify the root cause. The identified causes are summarized and categorized into a concise and clear description. This summarized cause information is then presented to relevant personnel, for example, "This anomaly is due to untimely updates to the source knowledge base, resulting in inconsistencies between the customer service knowledge base and the source knowledge base."

[0326] Displaying the fourth prompt helps relevant personnel to fundamentally resolve the problem and prevent similar anomalies from recurring. It also provides a basis for subsequent data management and system optimization.

[0327] S10-5, displays the fifth prompt message, which is used to suggest handling measures for abnormal situations.

[0328] In an optional embodiment, the fifth prompt message may be to provide handling suggestions for abnormal situations.

[0329] In an optional embodiment, the handling measures may be specific methods and steps for resolving abnormal situations.

[0330] Specifically, this embodiment formulates corresponding handling measures based on the type, cause, and degree of impact of the anomaly, such as data repair plans and system adjustment strategies. The handling measures are specified, clearly defining the operational steps, responsible persons, and timeframes. The formulated handling recommendations are presented to relevant personnel, for example, "It is recommended that data maintenance personnel repair the abnormal data within 2 hours and re-perform the data alignment operation."

[0331] The fifth prompt provides clear action guidelines for relevant personnel, helping to quickly and effectively handle abnormal situations, reduce the impact of abnormalities on business, and improve the efficiency of data reconciliation and problem solving.

[0332] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0333] The embodiments provided in this application display a first prompt message, which indicates the degree of impact of the abnormal situation on the customer service knowledge base-related business; a second prompt message, which indicates the scope of impact of the abnormal situation on the customer service knowledge base-related business; a third prompt message, which indicates the description of the abnormal data covered by the abnormal situation; a fourth prompt message, which indicates the cause of the abnormal situation; and a fifth prompt message, which suggests handling measures for the abnormal situation. These multiple prompt messages provide relevant personnel with a comprehensive analysis of the abnormal situation, enabling them to quickly understand the full picture of the problem and take targeted handling measures, greatly shortening the problem handling time.

[0334] As an optional approach, after displaying the error message, the method also includes:

[0335] S11-1, When the error message indicates that the error is caused by incorrect data entry, the erroneous data is corrected according to the preset correction method;

[0336] In an optional embodiment, the error message may be used to indicate abnormal situations during the data reconciliation process, specifically referring to an error message caused by incorrect data entry.

[0337] In an optional embodiment, the erroneous data can be an erroneous data record generated during the data entry process.

[0338] In an optional embodiment, the preset correction method may be a pre-defined method and rule for correcting erroneous data.

[0339] In an optional embodiment, error data correction may involve modifying, deleting, or supplementing erroneously entered data to restore it to its correct state.

[0340] Specifically, this embodiment analyzes the error message to determine that the error is caused by incorrect data entry. Based on preset correction rules, a correction method suitable for the current data type of the error is selected. For example, for numerical data entry errors, the correct value might be re-entered; for text data entry errors, spell checking and correction might be performed. Following the selected correction method, the erroneous data is actually corrected.

[0341] When the error message indicates that the error is caused by incorrect data entry, the data is corrected according to the preset correction method. This can correct data errors in a timely manner, ensure the accuracy of customer service knowledge base data, and avoid the impact of erroneous data on business decisions and customer service quality.

[0342] S11-2 When the exception message indicates that the exception is caused by a change in business rules, auxiliary prompts are displayed. The auxiliary prompts are used to indicate the relevant business rules.

[0343] In an optional embodiment, business rules change: During business operations, relevant rules may change, such as adjustments to customer service processes or changes to product pricing strategies.

[0344] In an optional embodiment, the auxiliary prompt information may be information used to help relevant personnel understand the abnormal situation caused by changes in business rules.

[0345] In an optional embodiment, the business rules involved may be specific content of business rules that have changed and are related to abnormal situations.

[0346] Specifically, this embodiment determines that the anomaly is caused by a change in business rules based on the anomaly message. The correlation between the business rule change and the anomaly is analyzed to identify the specific business rules involved. The relevant business rules are then compiled into supplementary information and displayed to relevant personnel.

[0347] When the error message indicates that the error is caused by changes in business rules, displaying supplementary information allows relevant personnel to understand the changes in business rules in a timely manner, so as to adjust business operation processes and data processing methods, ensure that the customer service knowledge base data is consistent with the new business rules, and guarantee the smooth operation of the business.

[0348] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0349] The embodiments provided in this application demonstrate that when an error message indicates the problem is caused by incorrect data entry, the erroneous data is corrected according to a preset correction method. When an error message indicates the problem is caused by changes in business rules, auxiliary prompts are displayed, indicating the relevant business rules. Correcting incorrect data entry directly improves the accuracy of customer service knowledge base data and reduces the impact of data errors on business operations. Furthermore, displaying auxiliary prompts due to changes in business rules helps relevant personnel understand rule changes promptly, adjust business operations, and ensure that customer service knowledge base data aligns with actual business practices, enhancing the adaptability and flexibility of the business.

[0350] As an optional solution, in the process of aligning the customer service knowledge base with the source knowledge base based on the data reconciliation results, the method also includes:

[0351] S12-1, Based on the data reconciliation results, using the data in the source knowledge base as a benchmark, obtain all the data in the customer service knowledge base that needs to be changed;

[0352] In an optional embodiment, all the data that needs to be changed can be a set of data in the customer service knowledge base that is inconsistent with the data in the source knowledge base and needs to be modified, added or deleted.

[0353] Specifically, this embodiment studies the data reconciliation results to identify data items that differ between the customer service knowledge base and the source knowledge base. Based on the reconciliation results, it determines all data in the customer service knowledge base that needs to be changed, including data records that need to be modified, new data that needs to be added, and redundant data that needs to be deleted. All the identified data requiring change is collected and organized to form a complete dataset.

[0354] Obtaining all the data that needs to be changed is fundamental to subsequent orderly change operations. By accurately identifying and collecting this data, we can ensure the comprehensiveness of data alignment, avoid missing any data that needs to be processed, and guarantee the accuracy and completeness of data alignment.

[0355] S12-2, using the data in the source knowledge base as a benchmark, performs orderly changes on each piece of data in all the data until all the data is aligned with the data in the source knowledge base. During the orderly change process, it is allowed to trace the data version of any piece of data in all the data.

[0356] In an optional embodiment, ordered changes can be performed on various data according to certain rules and order to ensure the controllability and stability of the change process.

[0357] In an optional embodiment, data alignment may involve making the data in the customer service knowledge base completely consistent with the data in the source knowledge base.

[0358] Specifically, this embodiment formulates detailed change rules based on the characteristics and alignment requirements of the data, such as processing important data first and classifying data according to data type. Following the change rules, each piece of data in the entire dataset is modified, added, or deleted sequentially. After each change operation, the data in the customer service knowledge base is checked to ensure alignment with the data in the source knowledge base. If not aligned, the change operation continues. This process of changing data and checking alignment is repeated until all data is aligned with the data in the source knowledge base.

[0359] Orderly changes ensure the stability and controllability of the data alignment process, avoiding data chaos and errors caused by disordered operations. By processing data step by step, problems that arise during the change process can be identified and resolved in a timely manner, improving the success rate of data alignment.

[0360] Furthermore, during the orderly change process, a version record is established for each piece of data, recording information such as the time, content, and person responsible for each change. Corresponding functions or interfaces are developed to allow relevant personnel to query and trace the version of any data. When needed, the data can be restored to a previous version state based on the version record.

[0361] Allowing for data version tracing provides robust assurance for the data alignment process. During changes, if errors occur or the change history needs to be understood, tracing data versions allows for rapid problem location and data recovery, ensuring data accuracy and integrity. Simultaneously, data version records provide a basis for data auditing and compliance checks.

[0362] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0363] The embodiments provided in this application, based on data reconciliation results and using the data in the source knowledge base as a benchmark, acquire all data in the customer service knowledge base that needs to be changed. Using the data in the source knowledge base as a benchmark, each piece of data in all data is changed in an orderly manner until all data is aligned with the data in the source knowledge base. During the orderly change process, it is possible to trace the data version of any piece of data. By acquiring all changed data and performing orderly changes, it ensures that the customer service knowledge base data is completely consistent with the source knowledge base data, improving data accuracy and consistency. Furthermore, allowing data version tracing provides security for the data change process, enabling rapid data recovery in case of problems and ensuring data reliability.

[0364] As an optional solution, for ease of understanding, the above-mentioned data reconciliation method for customer service knowledge base is applied to the reconciliation scenario of a customer service knowledge base in a certain application.

[0365] To further illustrate, optional examples include... Figure 4 The knowledge base interface shown allows users to input knowledge and enable a robot in their customer service account. After configuration, when a customer inquires, the robot will first automatically reply to the customer based on the input knowledge base. If it cannot reply, it will then be transferred to human service, thus improving service efficiency.

[0366] For example Figure 5 The message test window shown allows users to directly consult customer service through an application. The customer service robot will use AI algorithms to find the most matching answer from the knowledge base based on the user's input.

[0367] Specifically, this embodiment utilizes an interface with a customer service system to automatically initiate the reconciliation process based on preset conditions, such as time intervals, data update volume, and the number of customer service interactions. From a technical perspective, this automatic triggering mechanism is based on an advanced event-driven architecture and scheduled task technology. By setting reasonable time intervals, regular reconciliation can be ensured, promptly identifying potential data discrepancies; triggering reconciliation based on data update volume allows for timely responses to significant data changes, preventing the accumulation of inconsistencies; and triggering based on the number of customer service interactions provides focused assurance of data accuracy in high-frequency interaction scenarios. This multi-condition combined automatic triggering method significantly improves the timeliness and efficiency of reconciliation, reducing the cost and error of manual intervention.

[0368] To address the challenges of large-scale data reconciliation, this embodiment employs parallel reconciliation technology. This technology rationally divides the knowledge base data, distributes it across multiple computing units for synchronous processing, and finally merges the results. From a data processing principle perspective, parallel reconciliation technology fully leverages the advantages of multi-core processors and distributed computing. Through a reasonable data partitioning strategy, large-scale data is divided into multiple smaller sub-tasks, enabling multiple computing units to process it simultaneously, thus significantly improving data processing speed. Furthermore, in the result merging stage, an efficient data merging algorithm is used to ensure the accuracy of the merged result. This parallel processing approach effectively shortens the reconciliation time for large-scale data and improves the overall system performance.

[0369] Furthermore, the establishment of a real-time data monitoring and synchronization mechanism is a crucial guarantee for ensuring the accuracy and timeliness of knowledge base data. This mechanism can monitor the updates to knowledge base data in real time and quickly incorporate them into the reconciliation process when data is updated, without affecting ongoing tasks. From a technical implementation perspective, this mechanism is based on real-time message queues and database triggers. The real-time message queue can promptly capture data update messages and transmit them to the reconciliation system; the database triggers can automatically trigger the corresponding processing logic when data changes, ensuring that data updates are reflected in the reconciliation process in a timely manner. Through this real-time monitoring and synchronization mechanism, inaccurate reconciliation caused by data update delays is effectively avoided, ensuring the timeliness and accuracy of the data.

[0370] Furthermore, to provide strong support for the timely detection and handling of abnormal data, an intelligent abnormal data early warning mechanism was designed. This mechanism can automatically monitor outliers and issue timely warnings, while also formulating flexible handling strategies that can automatically select appropriate processing methods based on the type and severity of the abnormal data. From a technical perspective, this mechanism employs advanced data mining and machine learning algorithms. By analyzing and learning from historical data, an abnormal data model is established, enabling accurate identification of outliers.

[0371] Regarding early warning, this embodiment also employs multi-channel warning methods, such as email, SMS, and system messages, to ensure that relevant personnel receive warning information in a timely manner. In terms of processing strategies, corresponding processing rules have been formulated for different types of abnormal data, such as data entry errors and data discrepancies caused by changes in business rules. For example, data entry errors can be automatically corrected; data discrepancies caused by changes in business rules can prompt for manual review. This intelligent abnormal data early warning mechanism effectively improves the processing capacity and efficiency of abnormal data.

[0372] It should be noted that this embodiment applies the aforementioned data reconciliation method for customer service knowledge bases to a reconciliation scenario for a specific application's customer service knowledge base, comprehensively improving the intelligence level and overall efficiency of the reconciliation process. Through the integrated application of an automatic reconciliation trigger mechanism, parallel reconciliation technology, real-time data monitoring and synchronization mechanism, and intelligent abnormal data early warning mechanism, it effectively solves the problems of poor timeliness, low efficiency, and insufficient accuracy inherent in traditional reconciliation methods, demonstrating broad application value and significant innovation. From a technical principle and implementation perspective, the various mechanisms collaborate and complement each other, forming a complete and efficient data reconciliation system, providing strong assurance for the accuracy and consistency of data in the specific application's customer service knowledge base.

[0373] To further illustrate, optional examples include... Figure 6The knowledge base efficient reconciliation system 602 shown consists of a reconciliation trigger management module 604, a data monitoring and synchronization module 606, a parallel reconciliation processing module 608, and an abnormal data processing module 610.

[0374] The reconciliation trigger management module 604 integrates components such as a scheduled task scheduler, a data change monitor, and a condition judgment engine. The scheduled task scheduler initiates reconciliation condition checks according to preset time periods (such as hourly or daily); the data change monitor captures data update events in a certain application's customer service knowledge base in real time; the condition judgment engine makes a comprehensive judgment based on multiple trigger conditions such as set time intervals, data update volume, and number of customer service interactions, and sends a reconciliation start command to the reconciliation processing module when the conditions are met.

[0375] The parallel reconciliation processing module 608 consists of a data segmentation unit, a parallel computing unit, and a result merging unit. The data segmentation unit, based on the structural characteristics and current data volume of the knowledge base data, uses hash algorithms, range partitioning algorithms, and other methods to divide the complete dataset into multiple subsets suitable for parallel processing. The parallel computing unit comprises multiple independent computing threads or processes distributed across different server cores, synchronously performing reconciliation operations (such as data consistency comparison and logical relationship verification) on each subset. The result merging unit is responsible for collecting the processing results from each parallel computing unit, integrating and summarizing them to generate a complete reconciliation result set.

[0376] Data Monitoring and Synchronization Module 606 includes a data change monitoring submodule, a data synchronization coordination submodule, and a data consistency assurance submodule. The data change monitoring submodule monitors data changes in the application's customer service knowledge base in real time through various methods such as database transactions and message queues. The data synchronization coordination submodule coordinates data synchronization operations between the caching module, database, and reconciliation processing module based on factors such as the type and scale of data changes, ensuring that new data is promptly integrated into the reconciliation process. The data consistency assurance submodule uses data version control, transaction rollback, and other technologies to prevent data inconsistencies and data loss during data synchronization, ensuring the accuracy and stability of the entire reconciliation system.

[0377] The anomaly data processing module 610 comprises an anomaly detection model, an early warning notification engine, and an anomaly handling strategy library. The anomaly detection model uses statistical analysis and machine learning methods to perform in-depth analysis of reconciliation results and real-time data, identifying anomaly data (such as outliers, missing data, and logical inconsistencies). The early warning notification engine selects appropriate notification channels (such as in-site messages, SMS, and email) to send early warning notifications to relevant personnel based on the severity and scope of the anomaly. The notification content includes detailed anomaly information, possible cause analysis, and suggested handling measures. The anomaly handling strategy library predefines processing methods for different types of anomaly data, such as automatically correcting simple data entry errors and prompting manual review of data discrepancies caused by complex business rule changes, ensuring that anomaly data is properly handled and guaranteeing the smooth progress of reconciliation work.

[0378] In an optional embodiment, based on Figure 6 The scenario shown continues as follows Figure 7 As shown, the specific steps are as follows:

[0379] S702, Data Acquisition and Preprocessing Stage;

[0380] The reconciliation system (Knowledge Base High-Efficiency Reconciliation System 602) initiates data acquisition requests through an interface with a certain application's customer service system, according to predefined interface specifications and data request frequency. It collects various types of data from the application's customer service knowledge base, including but not limited to question classification data (such as specific question descriptions under categories like business consultation and technical support), answer data (detailed answers to each question, related attachments, etc.), and data update timestamps (recording the last modification time of each piece of data).

[0381] For the acquired data, a preliminary format check is performed to check whether the data meets the format requirements. If data with incorrect format is found, it is marked and recorded in a dedicated error log. At the same time, an attempt is made to repair the format (such as adding missing parentheses, quotation marks, and other basic elements). If the repair fails, the administrator is notified to intervene manually to ensure that the data format is correct for subsequent processes.

[0382] The collected raw data is cleaned to remove duplicate data records (by comparing key identifiers such as issue IDs) and invalid data (data entries that do not meet business rules, such as abnormal characters), thus ensuring the purity of the data.

[0383] S704, Reconciliation Trigger Judgment Phase;

[0384] The scheduled task in the reconciliation trigger management module starts the reconciliation condition monitoring task at set time intervals (such as once every 30 minutes). At the same time, the data change monitor continuously monitors the data update events of a certain application customer service knowledge base (by monitoring database write operations, data change notifications in the message queue, etc.) to ensure that factors that may trigger reconciliation are captured in real time.

[0385] The conditional judgment engine collects time interval information monitored by scheduled tasks, as well as data update volume and customer service interaction counts reported by the data change monitor, and makes a comprehensive judgment based on preset trigger condition rules. For example, if the set data update volume threshold is more than 100 new or modified data entries, when the knowledge base is detected to have 100 new or modified data entries within a certain time period, or when the time interval reaches 1 hour and the number of customer service interactions exceeds 50, the reconciliation trigger condition is determined to be met, and a reconciliation start command is sent to the parallel reconciliation processing module.

[0386] S706, Parallel reconciliation phase;

[0387] After receiving the reconciliation start command, the data segmentation unit in the parallel reconciliation processing module first analyzes the knowledge base data. Based on the business attributes and data volume distribution of the data, it selects an appropriate data segmentation algorithm (such as a hash value modulo-based segmentation algorithm, which distributes the data into different subsets according to certain hash rules; or a range-based segmentation algorithm, which divides the data according to the data update time range, question ID range, etc.) to divide the complete dataset into multiple relatively balanced subsets. This ensures that the data volume difference of each subset is controlled within a reasonable range, which facilitates load balancing of each computing unit during subsequent parallel processing.

[0388] The segmented datasets are distributed to various computing threads or processes within parallel computing units. Each computing unit independently runs the reconciliation logic, performing detailed comparisons of the data within its respective subset. This includes verifying whether questions and answers match, checking the correctness of data relationships (e.g., whether related questions and answers are logically consistent), and verifying data integrity (e.g., whether any required fields are missing). During the comparison process, discrepancies and suspected anomalies are recorded, generating separate sub-reconciliation results.

[0389] In this stage, the segmented datasets are evenly distributed among the computing threads or processes within the parallel computing units. Each computing unit independently performs data comparison work on its assigned dataset according to pre-defined reconciliation logic.

[0390] Specifically, for each question-answer pair contained in the assigned subset of data, the computing unit simultaneously extracts the question-answer content from a specific enterprise application and the corresponding question-answer content stored in that application, and then compares them meticulously one by one. For example, it carefully verifies whether the answer to a question in a specific enterprise application is completely consistent with the answer to the same question in another application.

[0391] Throughout the comparison process, the computing unit constantly monitors for discrepancies between the data. If any inconsistencies are found between the knowledge base data of a specific enterprise application and the knowledge base data stored in another application, these discrepancies and other key information, including any suspected anomalies, are recorded in detail. Furthermore, following established rules, the knowledge base data of the specific enterprise application is explicitly determined as the standard, and the knowledge base data stored in that application is aligned with it, thus ensuring consistency between the two sets of data. Finally, based on the above operations, each computing unit generates sub-reconciliation results corresponding to its respective subset of data, providing accurate foundational data support for subsequent result merging and integration.

[0392] After each parallel computing unit completes the reconciliation operation on its subset, the result merging unit begins to collect the sub-reconciliation results from each unit. According to the pre-set merging rules, all sub-reconciliation results are aggregated into a complete reconciliation result set. This result set contains detailed information on all data discrepancies and anomalies discovered during the entire reconciliation process of the customer service knowledge base data of a certain application, providing basic data support for subsequent abnormal data processing and result display.

[0393] S708, Data monitoring and synchronization phase.

[0394] The data change monitoring submodule in the data monitoring and synchronization module continuously monitors the data status of a customer service knowledge base in a certain application in real time. Through the subscription and publication mode of the message queue, it promptly detects any changes such as updates, deletions, and additions to the data, obtains the details of the changed data (such as the changed data content, change time, change source, etc.), and passes this change information to the data synchronization coordination submodule.

[0395] Upon receiving data change information, the data synchronization coordination submodule formulates corresponding data synchronization strategies based on the type and scale of the change and the current progress of the reconciliation work. For small-scale incremental data updates, incremental synchronization is used to directly integrate the new data into the corresponding subset of the ongoing reconciliation process (through operations such as updating cached data and adjusting data indexes). For large-scale data changes (such as batch importing new knowledge base entries), certain stages of the current reconciliation task are paused (using transaction control, resource locks, and other technical means to ensure data consistency), and a full data synchronization is performed before resuming the reconciliation work. This ensures that the new data can accurately and completely participate in the reconciliation process while avoiding impacting the normal progress of the reconciliation and the accuracy of the data.

[0396] During the data synchronization process, the data consistency assurance submodule provides full protection, using database transaction management mechanisms and data version control technology (ensuring orderly and traceable data version replacement at different stages) to prevent problems such as data inconsistency, data loss, and data conflicts. This ensures that the entire reconciliation system maintains the accuracy and stability of data in a dynamically changing environment, making the reconciliation results true and reliable.

[0397] Once abnormal data is detected, the alert notification engine assesses and classifies the abnormality based on factors such as its severity (e.g., the scope of its impact on business and potential erroneous consequences) and its impact range (e.g., whether it involves a single issue or multiple business segments). It then selects an appropriate notification channel (e.g., SMS notification for high-level abnormalities that severely impact business, and in-app message notification for general abnormalities) to send an alert to relevant personnel. The notification includes a detailed description of the abnormal data, possible cause analysis (based on a pre-established knowledge base of abnormal causes and data analysis results), and suggested handling measures (e.g., checking the data entry source and verifying whether business rules have changed). This ensures that relevant personnel are promptly informed of the abnormality and can take appropriate countermeasures.

[0398] Based on the type of abnormal data and the predefined exception handling strategy library, the system automatically executes corresponding processing operations. For some simple input error types of abnormal data (such as obvious typos, format errors, etc.), the system automatically corrects the data according to preset correction rules (such as replacing incorrect characters, adjusting format, etc.). For abnormal data caused by changes in business rules or complex logical relationships (such as answer changes due to business process adjustments), the system prompts manual handling, provides detailed abnormal data comparison information, information on possible changes in business rules, etc., to assist manual judgment and subsequent processing operations (such as confirming whether to update answers, adjusting business processes, etc.), ensuring that abnormal data is properly handled and guaranteeing the accuracy of the application's customer service knowledge base data and the normal operation of reconciliation work.

[0399] Optionally, the content of this embodiment can be applied to other embodiments described above, or combined with other embodiments described above.

[0400] The embodiments provided in this application introduce an automatically triggered reconciliation mechanism. Through an interface with a customer service system of a certain application, the reconciliation process is automatically initiated based on preset conditions such as time intervals, data update volume, and the number of customer service interactions, achieving intelligent and real-time reconciliation. From a technical perspective, this mechanism is based on an event-driven architecture and scheduled task technology. By setting reasonable time intervals, the system can automatically trigger reconciliation periodically, ensuring the regularity of the reconciliation process. Triggering reconciliation based on data update volume allows for timely responses to significant data changes, preventing the accumulation of data inconsistencies. Triggering based on the number of customer service interactions provides focused assurance of data accuracy in high-frequency interaction scenarios. This multi-condition combined automatic triggering method frees reconciliation from being limited to fixed periods, enabling real-time responses to changes in the knowledge base. Any changes that may affect data accuracy can quickly trigger reconciliation, promptly identifying and resolving data discrepancies, thereby ensuring the timeliness of reconciliation and providing strong data support for the continuous and stable operation of the business.

[0401] When processing large-scale data reconciliation for an application's customer service knowledge base, the inefficiency of traditional sequential processing methods is particularly prominent. From a technical perspective, sequential processing means that data must pass through each processing stage in turn, failing to fully utilize computing resources and resulting in long processing times and low efficiency. This patent utilizes parallel reconciliation technology, rationally dividing the knowledge base data and distributing it to multiple computing units for simultaneous processing, then merging the results. This parallel processing mode, based on distributed computing and parallel algorithm technology, uses a reasonable data partitioning strategy to divide large-scale data into multiple smaller sub-tasks, enabling multiple computing units to process simultaneously. This significantly shortens the reconciliation time for large-scale data and accelerates the entire reconciliation process. In this way, the originally lengthy reconciliation process is completed efficiently, helping to improve overall business operational efficiency, allowing relevant personnel to obtain accurate reconciliation results more quickly, and thus promptly advance subsequent work based on the results.

[0402] To ensure data accuracy, this patent employs several technical measures. First, a real-time data monitoring and synchronization mechanism is established. Technically, this mechanism is based on real-time message queues and database triggers. The real-time message queue promptly captures data update messages and transmits them to the reconciliation system; the database triggers automatically activate corresponding processing logic when data changes, ensuring that data updates are reflected in the reconciliation process in a timely manner. Through this ingenious technical design, when the knowledge base data is updated, the new data can be quickly incorporated into the reconciliation process without affecting ongoing reconciliation tasks. This ensures that reconciliation is always based on the latest and most accurate data, avoiding errors in reconciliation results caused by data asynchrony or untimely updates, effectively guaranteeing data accuracy and improving the reliability of reconciliation results.

[0403] Secondly, the abnormal data early warning mechanism can automatically detect various abnormal values, such as data format mismatches and values ​​exceeding reasonable ranges, and issue timely warnings. Based on data mining and machine learning algorithms, this mechanism builds an abnormal data model by learning from historical and normal data patterns, enabling accurate identification of outliers. Simultaneously, it can automatically select appropriate handling methods based on the type and severity of the abnormal data, such as automatically correcting incorrectly entered data or prompting manual handling of data discrepancies caused by changes in business rules. This effectively prevents abnormal data from being overlooked or mishandled during the reconciliation process, further ensuring the accuracy of the data used to base the final reconciliation results and improving the overall quality of the application's customer service knowledge base data.

[0404] From the perspective of labor costs and resource utilization, the traditional model of manual reconciliation in various stages has many drawbacks. Manual operation not only consumes a significant amount of manpower and time but is also prone to errors due to human negligence. This patent automates and automates the reconciliation process, reducing the workload of manual operations and reliance on human experience. Technically, the automated process automatically executes reconciliation tasks through preset rules and algorithms, requiring no human intervention. This allows relevant staff to devote more energy to more valuable tasks such as business analysis and customer service optimization, thus optimizing the allocation of labor costs. Simultaneously, the efficient reconciliation mechanism avoids the additional hardware resource consumption that may result from inefficient reconciliation, improving overall resource utilization efficiency. By rationally allocating computing and human resources, the system can achieve higher business value at a lower cost while ensuring reconciliation quality.

[0405] It is understood that in the specific embodiments of this application, data such as user information are involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0406] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0407] According to another aspect of the embodiments of this application, a data reconciliation apparatus for a customer service knowledge base is also provided for implementing the above-described data reconciliation method for a customer service knowledge base. For example... Figure 8 As shown, the device includes:

[0408] The first acquisition unit 802 is used to respond to the data reconciliation request and acquire the knowledge base data set of the customer service knowledge base, wherein the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base;

[0409] The segmentation unit 804 is used to segment the knowledge base data set to obtain at least two knowledge base data subsets, wherein the difference in data volume between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold.

[0410] Parallel unit 806 is used to execute each data reconciliation task corresponding to each knowledge base data subset in parallel and obtain data reconciliation results. The data reconciliation task is used to compare the data differences between the knowledge base data subset and the source knowledge base, and the data reconciliation result is used to represent the data differences between the customer service knowledge base and the source knowledge base.

[0411] Alignment unit 808 is used to align the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark.

[0412] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0413] As an optional solution, the parallel unit 806 includes:

[0414] The parallel module is used to execute various data reconciliation tasks in parallel and obtain the execution results of each data reconciliation task.

[0415] The integration module is used to combine the results of various task executions to obtain data reconciliation results.

[0416] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0417] As an optional solution, the device also includes:

[0418] The first acquisition module is used to acquire the first question corresponding to the target reconciliation task during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each task corresponding to each data reconciliation task. The target reconciliation task is used to compare the data differences between the target data subset and the source knowledge base. Each data reconciliation task includes the target reconciliation task, and each knowledge base data subset includes the target data subset.

[0419] The second acquisition module is used to acquire a subset of target data, which is the first answer to the first question, during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each data reconciliation task.

[0420] The third acquisition module is used to acquire the source knowledge base for the second answer to the first question during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each data reconciliation task.

[0421] The fourth acquisition module is used to compare the first answer and the second answer during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each data reconciliation task to obtain the information difference of the first answer. The execution result of the target reconciliation task includes the information difference of the first answer.

[0422] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0423] As an optional solution, the device also includes:

[0424] The fifth acquisition module is used to acquire the second question corresponding to the target reconciliation task during the process of executing each data reconciliation task in parallel and acquiring the execution results of each data reconciliation task. The second question is related to the first question.

[0425] The sixth acquisition module is used to acquire a subset of target data during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each data reconciliation task, which is the third answer to the second question.

[0426] The seventh acquisition module is used to acquire the source knowledge base for the fourth answer to the second question during the parallel execution of various data reconciliation tasks and the acquisition of the execution results of each data reconciliation task.

[0427] The eighth acquisition module is used to acquire the correlation and coherence information corresponding to the target reconciliation task based on the first answer, the second answer, the third answer, and the fourth answer during the parallel execution of various data reconciliation tasks and the acquisition of the task execution results corresponding to each data reconciliation task. The correlation and coherence information is used to represent the coherence of the target data subset and the source knowledge base in the business logic of related questions and answers. The task execution results corresponding to the target reconciliation task include the correlation and coherence information.

[0428] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0429] As an optional solution, the device also includes:

[0430] The second acquisition unit is used to acquire a data reconciliation request before acquiring the knowledge base data set of the customer service knowledge base in response to a data reconciliation request, provided that at least one reconciliation triggering condition is met, wherein at least one reconciliation triggering condition includes at least one of the following: reconciliation frequency condition, data change condition of the customer service knowledge base, and customer service interaction condition of the customer service knowledge base.

[0431] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0432] As an optional solution, the device also includes:

[0433] The first determining unit is configured to, before obtaining a data reconciliation request, determine that a first triggering condition is met if the amount of data added or modified in the customer service knowledge base within a preset time period is greater than or equal to a first threshold. The first triggering condition includes data change conditions in the customer service knowledge base, and at least one reconciliation triggering condition includes the first triggering condition; or...

[0434] The second determining unit is used to determine whether the second triggering condition is met before obtaining the data reconciliation request, provided that the reconciliation time interval of the customer service knowledge base is greater than or equal to a second threshold and the number of customer service interactions in the customer service knowledge base is greater than or equal to a third threshold. The second triggering condition includes the reconciliation frequency condition and the customer service interaction condition of the customer service knowledge base, and at least one reconciliation triggering condition includes the second triggering condition.

[0435] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0436] As an optional solution, the apparatus further includes: a third determining unit, used to determine the data segmentation method based on the data type of the knowledge base data set before segmenting the knowledge base data set to obtain at least two knowledge base data subsets;

[0437] The segmentation unit 804 includes a segmentation module, used to segment the knowledge base data set according to the data segmentation method to obtain at least two knowledge base data subsets.

[0438] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0439] As an optional solution, the module is divided into:

[0440] The first segmentation submodule is used to segment the knowledge base data set based on a hash value modulo segmentation algorithm when the amount of data of different data types in the knowledge base data set is greater than or equal to the fourth threshold, so as to obtain at least two knowledge base data subsets.

[0441] The second segmentation submodule is used to segment the knowledge base dataset based on a time range-based segmentation algorithm when the amount of data of different data types in the knowledge base dataset is less than the fourth threshold, so as to obtain at least two knowledge base dataset subsets.

[0442] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0443] As an optional solution, the device also includes: a fourth determining unit, used to determine the data synchronization method based on the data reconciliation results before aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark;

[0444] Alignment unit 808 includes: an alignment module, used to align the data between the customer service knowledge base and the source knowledge base according to the data synchronization method and based on the data in the source knowledge base.

[0445] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0446] As an optional solution, the alignment module includes:

[0447] The first alignment submodule is used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to the fifth threshold.

[0448] The second alignment submodule is used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base by using a full data synchronization method when the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold.

[0449] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0450] As an optional solution, the device also includes:

[0451] The third acquisition unit is used to acquire various types of data in the customer service knowledge base in response to the data preprocessing request before acquiring the knowledge base data set of the customer service knowledge base in response to the data reconciliation request.

[0452] The preprocessing unit is used to preprocess various types of data before obtaining the knowledge base data set of the customer service knowledge base in response to the data reconciliation request, so as to obtain multiple target data, wherein the knowledge base data set includes multiple target data.

[0453] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0454] As an optional solution, the device also includes:

[0455] The verification unit is used to perform format verification on various types of data during the preprocessing of various types of data to obtain multiple target data.

[0456] The repair unit is used to mark the first data in the data category that fails the format validation during the preprocessing of various types of data to obtain multiple target data, and to record the first data in the error log and repair the format of the first data.

[0457] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0458] As an optional solution, the device also includes:

[0459] The cleaning unit is used to clean the various types of data during the preprocessing of various types of data to obtain multiple target data, and to remove the second data in each type of data, where the second data is duplicate or invalid data.

[0460] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0461] As an optional solution, the device also includes:

[0462] The fourth acquisition unit is used to acquire the amount of data that has changed during the data alignment process between the customer service knowledge base and the source knowledge base, based on the data reconciliation results and using the data in the source knowledge base as a benchmark.

[0463] The display unit is used to display an error message when the amount of data change is greater than or equal to the sixth threshold after aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark. The error message is used to indicate that there is an error in the data reconciliation process of the customer service knowledge base.

[0464] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0465] As an optional solution, the display unit includes at least one of the following:

[0466] The first display module is used to display the first prompt information, which is used to indicate the degree of impact of the abnormal situation on the customer service knowledge base related business.

[0467] The second display module is used to display the second prompt information, wherein the second prompt information is used to indicate the scope of the impact of the abnormal situation on the customer service knowledge base related business;

[0468] The third display module is used to display the third prompt information, which is used to prompt the description of the abnormal data covered by the abnormal situation;

[0469] The fourth display module is used to display the fourth prompt message, which indicates the reason for the abnormal situation.

[0470] The fifth display module is used to display the fifth prompt message, which is a suggested handling measure for abnormal situations.

[0471] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0472] As an optional solution, the device also includes:

[0473] The correction module is used to correct the erroneous data according to a preset correction method after the error message is displayed, if the error message indicates that the error is caused by incorrect data entry.

[0474] The sixth display module is used to display auxiliary prompts after the error message is displayed, in cases where the error message indicates that the abnormal situation is caused by changes in business rules. The auxiliary prompts are used to indicate the relevant business rules.

[0475] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0476] As an optional solution, the device also includes:

[0477] The fifth acquisition unit is used to acquire all the data that needs to be changed in the customer service knowledge base during the process of aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results and using the data in the source knowledge base as a benchmark.

[0478] The change unit is used to make orderly changes to each piece of data in all the data based on the data in the source knowledge base, until all the data is aligned with the data in the source knowledge base. During the orderly change process, it is possible to trace the data version of any piece of data in all the data.

[0479] For specific implementation examples, please refer to the example shown in the data reconciliation method of the customer service knowledge base above. These examples will not be repeated here.

[0480] According to another aspect of the embodiments of this application, an electronic device for implementing the above-described customer service knowledge base data reconciliation method is also provided. This electronic device may, but is not limited to, [providing a specific function or capability]. Figure 1 The user equipment 102 or server 112 shown in the figure, in this embodiment, is taken as an example of an electronic device, namely user equipment 102. Further, as shown in the figure... Figure 9 As shown, the electronic device includes a memory 902 and a processor 904. The memory 902 stores a computer program, and the processor 904 is configured to execute the steps of any of the above method embodiments through the computer program.

[0481] In an optional embodiment, the aforementioned electronic device may be located in at least one of a plurality of network devices in a computer network.

[0482] In an optional embodiment, the processor described above may be configured to perform the following steps via a computer program:

[0483] S1, in response to the data reconciliation request, obtains the knowledge base data set of the customer service knowledge base, where the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base;

[0484] S2, divide the knowledge base data set to obtain at least two knowledge base data subsets, wherein the difference in data volume between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold.

[0485] S3, execute each data reconciliation task corresponding to each knowledge base data subset in parallel, and obtain the data reconciliation results. The data reconciliation task is used to compare the data differences between the knowledge base data subset and the source knowledge base, and the data reconciliation results are used to represent the data differences between the customer service knowledge base and the source knowledge base.

[0486] S4, based on the data reconciliation results, aligns the data between the customer service knowledge base and the source knowledge base using the data in the source knowledge base as a benchmark.

[0487] Alternatively, as those skilled in the art will understand, Figure 9The structure shown is for illustrative purposes only. Figure 9 This does not limit the structure of the aforementioned electronic devices. For example, the electronic device may also include components that are more... Figure 9 The more or fewer components shown (such as network interfaces, etc.), or having the same Figure 9 The different configurations shown.

[0488] The memory 902 can be used to store software programs and modules, such as the program instructions / modules corresponding to the customer service knowledge base data reconciliation method and apparatus in this embodiment. The processor 904 executes various functional applications and data processing by running the software programs and modules stored in the memory 902, thereby realizing the aforementioned customer service knowledge base data reconciliation method. The memory 902 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 902 may further include memory remotely located relative to the processor 904, and these remote memories can be connected to electronic devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. Specifically, the memory 902 may be used, but is not limited to, to store information such as knowledge base data sets, subsets of knowledge base data, and data reconciliation results. As an example, such as... Figure 9 As shown, the memory 902 may include, but is not limited to, the first acquisition unit 802, the segmentation unit 804, the parallel unit 806, and the alignment unit 808 of the customer service knowledge base data reconciliation device. Furthermore, it may include, but is not limited to, other module units of the customer service knowledge base data reconciliation device, which will not be elaborated upon in this example.

[0489] Optionally, the transmission device 906 described above is used to receive or send data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 906 includes a Network Interface Controller (NIC), which can be connected to other network devices and routers via a network cable to communicate with the Internet or a local area network. In another example, the transmission device 906 is a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0490] In addition, the aforementioned electronic device also includes: a display 908 for displaying information such as the aforementioned knowledge base data set, knowledge base data subset, and data reconciliation results; and a connection bus 910 for connecting the various module components in the aforementioned electronic device.

[0491] In other embodiments, the aforementioned user equipment or server can be a node in a distributed system, wherein the distributed system can be a blockchain system, which is a distributed system formed by connecting multiple nodes through network communication. The nodes can form a peer-to-peer network, and any form of computing device, such as a server, user equipment, or other electronic device, can become a node in the blockchain system by joining this peer-to-peer network.

[0492] According to one aspect of this application, a computer program product is provided, comprising a computer program / instructions containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions provided in embodiments of this application.

[0493] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0494] It should be noted that the computer system of the electronic device is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0495] A computer system includes a Central Processing Unit (CPU), which performs various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) or loaded from RAM. ROM also stores various programs and data required for system operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output interfaces (I / O interfaces) are also connected to the bus.

[0496] The following components are connected to the input / output interface: input sections including keyboards, mice, etc.; output sections including cathode ray tubes (CRTs), liquid crystal displays (LCDs), and speakers; storage sections including hard drives; and communication sections including network interface cards such as LAN cards and modems. The communication section performs communication processing via a network such as the Internet. Drives are also connected to the input / output interface as needed. Removable media, such as disks, optical discs, magneto-optical discs, semiconductor memories, etc., are installed on the drive as needed so that computer programs read from them can be installed into the storage section as required.

[0497] Specifically, according to embodiments of this application, the processes described in the various method flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions defined in the system of this application.

[0498] According to one aspect of this application, a computer-readable storage medium is provided, wherein a processor of a computer device reads computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations described above.

[0499] In an optional embodiment, the computer-readable storage medium described above may be configured to store a computer program for performing the following steps:

[0500] S1, in response to the data reconciliation request, obtains the knowledge base data set of the customer service knowledge base, where the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base;

[0501] S2, divide the knowledge base data set to obtain at least two knowledge base data subsets, wherein the difference in data volume between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold.

[0502] S3, execute each data reconciliation task corresponding to each knowledge base data subset in parallel, and obtain the data reconciliation results. The data reconciliation task is used to compare the data differences between the knowledge base data subset and the source knowledge base, and the data reconciliation results are used to represent the data differences between the customer service knowledge base and the source knowledge base.

[0503] S4, based on the data reconciliation results, aligns the data between the customer service knowledge base and the source knowledge base using the data in the source knowledge base as a benchmark.

[0504] Optionally, in embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0505] In optional embodiments, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware of an electronic device. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0506] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0507] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods of the various embodiments of this application.

[0508] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0509] In the several embodiments provided in this application, it should be understood that the disclosed user equipment can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0510] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0511] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0512] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A data reconciliation method for a customer service knowledge base, characterized in that, include: In response to a data reconciliation request, the customer service knowledge base data set is obtained, wherein the customer service knowledge base is a knowledge base that provides data support for customer service accounts, and the data reconciliation request is used to request the reconciliation of data between the customer service knowledge base and the source knowledge base; The knowledge base dataset is divided to obtain at least two knowledge base data subsets, wherein the difference in data volume between each knowledge base data subset in the at least two knowledge base data subsets is less than or equal to a preset threshold. The data reconciliation tasks corresponding to each subset of knowledge base data are executed in parallel to obtain data reconciliation results. The data reconciliation tasks are used to compare the data differences between the subset of knowledge base data and the source knowledge base, and the data reconciliation results are used to represent the data differences between the customer service knowledge base and the source knowledge base. Based on the data reconciliation results, and using the data in the source knowledge base as a benchmark, the data between the customer service knowledge base and the source knowledge base is aligned.

2. The method according to claim 1, characterized in that, The parallel execution of each data reconciliation task corresponding to each subset of knowledge base data to obtain data reconciliation results includes: The data reconciliation tasks are executed in parallel, and the execution results of each data reconciliation task are obtained. The results of each task execution are integrated to obtain the data reconciliation result.

3. The method according to claim 2, characterized in that, In the process of executing the various data reconciliation tasks in parallel and obtaining the execution results of each corresponding data reconciliation task, the method further includes: Obtain the first question corresponding to the target reconciliation task, wherein the target reconciliation task is used to compare the data differences between the target data subset and the source knowledge base, each data reconciliation task includes the target reconciliation task, and each knowledge base data subset includes the target data subset; Obtain the target data subset and the first answer to the first question; Obtain the source knowledge base for the second answer to the first question; The first answer and the second answer are compared to obtain the first answer information difference, wherein the task execution result corresponding to the target reconciliation task includes the first answer information difference.

4. The method according to claim 3, characterized in that, In the process of executing the various data reconciliation tasks in parallel and obtaining the execution results of each data reconciliation task, the method further includes: Obtain the second question corresponding to the target reconciliation task, wherein the second question is related to the first question; Obtain the target data subset for the third answer to the second question; Obtain the source knowledge base for the fourth answer to the second question; Based on the first answer, the second answer, the third answer, and the fourth answer, the associated coherence information corresponding to the target reconciliation task is obtained. The associated coherence information is used to represent the business logic coherence between the target data subset and the source knowledge base in related questions and answers. The task execution result corresponding to the target reconciliation task includes the associated coherence information.

5. The method according to claim 1, characterized in that, Before responding to the data reconciliation request and obtaining the knowledge base data set of the customer service knowledge base, the method further includes: The data reconciliation request is obtained when at least one reconciliation trigger condition is met, wherein the at least one reconciliation trigger condition includes at least one of the following: reconciliation frequency condition, data change condition of the customer service knowledge base, and customer service interaction condition of the customer service knowledge base.

6. The method according to claim 5, characterized in that, Prior to obtaining the data reconciliation request, the method further includes: If the amount of data added or modified in the customer service knowledge base within a preset time period is greater than or equal to a first threshold, the first triggering condition is determined to be met. The first triggering condition includes data change conditions in the customer service knowledge base, and the at least one reconciliation triggering condition includes the first triggering condition; or, If the reconciliation time interval of the customer service knowledge base is greater than or equal to a second threshold, and the number of customer service interactions in the customer service knowledge base is greater than or equal to a third threshold, it is determined that a second triggering condition is met. The second triggering condition includes the reconciliation frequency condition and the customer service interaction condition of the customer service knowledge base, and the at least one reconciliation triggering condition includes the second triggering condition.

7. The method according to claim 1, characterized in that, Before dividing the knowledge base data set to obtain at least two knowledge base data subsets, the method further includes: determining a data segmentation method based on the data type of the knowledge base data set; The step of dividing the knowledge base data set to obtain at least two knowledge base data subsets includes: dividing the knowledge base data set according to the data segmentation method to obtain the at least two knowledge base data subsets.

8. The method according to claim 7, characterized in that, The step of dividing the knowledge base data set according to the data segmentation method to obtain at least two knowledge base data subsets includes: If the amount of data of different data types in the knowledge base data set is greater than or equal to the fourth threshold, the knowledge base data set is divided based on the hash value modulo splitting algorithm to obtain at least two knowledge base data subsets. If the amount of data of different data types in the knowledge base data set is less than the fourth threshold, the knowledge base data set is divided by a time range-based segmentation algorithm to obtain at least two knowledge base data subsets.

9. The method according to claim 1, characterized in that, Before aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results, the method further includes: determining the data synchronization method based on the data reconciliation results; The step of aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results includes: aligning the data between the customer service knowledge base and the source knowledge base according to the data synchronization method, using the data in the source knowledge base as a benchmark.

10. The method according to claim 9, characterized in that, The step of aligning the customer service knowledge base with the source knowledge base, based on the data in the source knowledge base, according to the data synchronization method, includes: If the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is greater than or equal to the fifth threshold, an incremental data synchronization method is used to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base. If the data reconciliation result indicates that the data difference between the customer service knowledge base and the source knowledge base is less than the fifth threshold, a full data synchronization method is adopted to supplement the difference data in the source knowledge base to the corresponding knowledge base data subset in the customer service knowledge base.

11. The method according to claim 1, characterized in that, Before responding to the data reconciliation request and obtaining the knowledge base data set of the customer service knowledge base, the method further includes: In response to a data preprocessing request, various types of data are obtained from the customer service knowledge base; The various types of data are preprocessed to obtain multiple target data, wherein the knowledge base data set includes multiple target data.

12. The method according to claim 11, characterized in that, In the process of preprocessing the various types of data to obtain multiple target data, the method further includes: Perform format validation on the aforementioned data types; If the first piece of data in the various data types fails the format verification, the first piece of data is marked and recorded in the error log, and the format of the first piece of data is repaired.

13. The method according to claim 11, characterized in that, In the process of preprocessing the various types of data to obtain multiple target data, the method further includes: A cleaning operation is performed on the various types of data to remove the second data, which is either duplicate or invalid data.

14. The method according to claim 1, characterized in that, After aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results, using the data in the source knowledge base as a benchmark, the method further includes: The amount of data that has changed during the process of aligning the customer service knowledge base with the source knowledge base is obtained. If the amount of data that has been changed is greater than or equal to the sixth threshold, an error message will be displayed. The error message is used to indicate that there is an anomaly in the data reconciliation process of the customer service knowledge base.

15. The method according to claim 14, characterized in that, The displayed error message includes at least one of the following: Display a first prompt message, wherein the first prompt message is used to indicate the degree of impact of the abnormal situation on the customer service knowledge base related business; Display a second prompt message, wherein the second prompt message is used to indicate the scope of the impact of the abnormal situation on the customer service knowledge base related business; Display a third prompt message, wherein the third prompt message is used to indicate the description of the abnormal data covered by the abnormal situation; The fourth prompt message is displayed, which is used to indicate the reason for the abnormal situation. The fifth prompt message is displayed, which is used to suggest handling measures for the abnormal situation.

16. The method according to claim 14, characterized in that, After displaying the error message, the method further includes: If the error message indicates that the error is caused by incorrect data entry, the erroneous data will be corrected according to a preset correction method. When the error message indicates that the abnormal situation is caused by a change in business rules, auxiliary prompts are displayed, wherein the auxiliary prompts are used to indicate the relevant business rules.

17. The method according to any one of claims 1 to 16, characterized in that, In the process of aligning the data between the customer service knowledge base and the source knowledge base based on the data reconciliation results, using the data in the source knowledge base as a benchmark, the method further includes: Based on the data reconciliation results, and using the data in the source knowledge base as a benchmark, all data in the customer service knowledge base that needs to be changed is obtained. Based on the data in the source knowledge base, each piece of data in all the data is changed in an orderly manner until all the data is aligned with the data in the source knowledge base. During the orderly change process, it is allowed to trace the data version of any piece of data in all the data.

18. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program is executed by an electronic device to perform the method according to any one of claims 1 to 17.

19. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1 to 17.

20. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 17 through the computer program.