A complaint root cause tracing method and device

Abstract

This application provides a method and apparatus for tracing the root causes of complaints. The method includes: acquiring complaint content data; processing the complaint content data using a preset causal extraction pattern matching model to obtain complaint keywords; extracting keywords from the complaint keywords to obtain complaint cause keywords; constructing a complaint graph based on the complaint keywords and complaint cause keywords; and tracing the root causes of complaints based on the complaint graph to obtain the root cause tracing results. It is evident that implementing this method can trace the root causes of complaints, identify the root problems that led to the complaints, and thus fundamentally resolve such complaint issues. This is beneficial for improving the efficiency and effectiveness of complaint management and effectively enhancing the consumer experience.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method and apparatus for tracing the root causes of complaints. Background Technology

[0002] Currently, the classification of bank credit card complaints follows a five-tier system set by the People's Bank of China and individual banks to meet their specific needs. This system is based on a banking business perspective, such as "bank card - credit card - credit card usage and repayment - interest and fees - interest and late fees," aiming to quickly identify and resolve similar complaints for customers from a business standpoint. However, in practice, it has been observed that some customers submit multiple complaints, each involving different types of transactions. This makes it impossible to definitively determine the cause of the complaint, hindering its resolution and negatively impacting user experience. Summary of the Invention

[0003] The purpose of this application is to provide a method and apparatus for tracing the root cause of complaints, which can trace the root cause of complaints, determine the root cause of the complaints, and thus fundamentally solve such complaints, thereby improving the efficiency and effectiveness of complaint management and effectively enhancing the consumer experience.

[0004] The first aspect of this application provides a method for tracing the root cause of a complaint, including:

[0005] Obtain complaint content data;

[0006] The complaint content data is processed by a preset causal extraction pattern matching model to obtain complaint keywords;

[0007] The complaint keywords are processed by keyword extraction to obtain the keywords for the reasons for the complaint;

[0008] Construct a complaint graph based on the complaint keywords and the complaint reason keywords;

[0009] The root cause of the complaint is traced based on the complaint map to obtain the root cause tracing results.

[0010] In the above implementation process, this method first acquires complaint content data; then, it processes the complaint content data using a pre-defined causal extraction pattern matching model to obtain complaint keywords; next, it extracts keywords from the complaint keywords to obtain complaint reason keywords; then, it constructs a complaint graph based on the complaint keywords and complaint reason keywords; finally, it traces the root cause of the complaint based on the complaint graph to obtain the root cause tracing results. It is evident that this method can trace the root cause of complaints, identify the root problem that caused the complaint, and thus fundamentally solve such complaint problems, which is conducive to improving the efficiency and effectiveness of complaint management and effectively enhancing consumer experience.

[0011] Furthermore, the keyword extraction process performed on the complaint keywords to obtain complaint reason keywords includes:

[0012] The complaint keywords are segmented to obtain a keyword set;

[0013] The keyword set is sorted by word frequency to obtain a keyword ranking sequence;

[0014] Select a preset number of keywords from the keyword sorting sequence to obtain the target keywords;

[0015] The target keywords are filtered based on a pre-built stop word dictionary to obtain keywords related to the reasons for complaints.

[0016] Furthermore, the step of constructing a complaint graph based on the complaint keywords and the complaint reason keywords includes:

[0017] The complaint keywords and the complaint reason keywords are filtered to obtain node data and edge data for constructing the complaint graph;

[0018] Obtain the weight configuration information for the node data and the edge data;

[0019] A complaint graph is constructed based on the node data, the edge data, and the weight configuration information.

[0020] Furthermore, the step of tracing the root causes of complaints based on the complaint map to obtain the root cause tracing results includes:

[0021] The mediator centrality of the complaint graph is calculated to obtain the root cause keyword mediator centrality.

[0022] The set of root cause keywords that intersects is determined based on the median centrality of the root cause keywords and the preset complaint categories;

[0023] The root cause keywords in the root cause keyword set are sorted according to the root cause keyword betweenness centrality to obtain a sorted keyword sequence;

[0024] The root cause tracing results of complaints are determined based on the sorted keyword sequence; wherein, the root cause tracing results of complaints include common root cause keywords, subjective root cause keywords, and objective root cause keywords.

[0025] Furthermore, the method also includes:

[0026] Based on the root cause tracing results of the complaints, subjective root cause keywords and objective root cause keywords are determined;

[0027] Based on the aforementioned subjective root cause keywords, determine the subjective root cause and objective root cause;

[0028] Obtain a first suggested solution for the subjective root cause and a second suggested solution for the objective root cause;

[0029] Output the first solution suggestion and the second solution suggestion.

[0030] A second aspect of this application provides a complaint root cause tracing device, the complaint root cause tracing device comprising:

[0031] The first acquisition unit is used to acquire complaint content data;

[0032] The first processing unit is used to process the complaint content data through a preset causal extraction pattern matching model to obtain complaint keywords;

[0033] The second processing unit is used to perform keyword extraction processing on the complaint keywords to obtain complaint reason keywords;

[0034] The construction unit is used to construct a complaint graph based on the complaint keywords and the complaint reason keywords;

[0035] The tracing unit is used to trace the root cause of complaints based on the complaint map and obtain the root cause tracing result.

[0036] In the above implementation process, the device can acquire complaint content data through the first acquisition unit; process the complaint content data through the first processing unit using a preset causal extraction pattern matching model to obtain complaint keywords; extract keywords from the complaint keywords through the second processing unit to obtain complaint reason keywords; construct a complaint graph based on the complaint keywords and complaint reason keywords through the construction unit; and finally, trace the root cause of the complaint based on the complaint graph through the tracing unit to obtain the root cause tracing result. It is evident that this device can trace the root cause of complaints, determine the root problem of the complaint, and thus fundamentally solve such complaint problems, which is conducive to improving the efficiency and effectiveness of complaint management and effectively enhancing consumer experience.

[0037] Further, the second processing unit includes:

[0038] The word segmentation subunit is used to segment the complaint keywords to obtain a keyword set;

[0039] The first sorting subunit is used to perform word frequency statistical sorting on the keyword set to obtain a keyword sorting sequence;

[0040] A sub-unit is selected to select a preset number of keywords from the keyword sorting sequence to obtain target keywords;

[0041] The filtering subunit is used to filter the target keywords according to a pre-built stop word dictionary to obtain keywords for the reasons for complaints.

[0042] Furthermore, the building unit includes:

[0043] The filtering subunit is used to filter the complaint keywords and the complaint reason keywords to obtain node data and edge data for constructing the complaint graph;

[0044] A sub-unit is used to acquire weight configuration information for the node data and the edge data;

[0045] A sub-unit is constructed to build a complaint graph based on the node data, the edge data, and the weight configuration information.

[0046] Furthermore, the traceability unit includes:

[0047] The calculation subunit is used to calculate the median centrality of the complaint graph to obtain the root cause keyword median centrality.

[0048] A subunit is defined to determine the set of root cause keywords that intersect based on the median centrality of the root cause keywords and the preset complaint category.

[0049] The second sorting subunit is used to sort the root keywords in the root keyword set according to the root keyword betweenness centrality to obtain a sorted keyword sequence.

[0050] The determining subunit is further configured to determine the root cause tracing result of the complaint based on the sorted keyword sequence; wherein the root cause tracing result of the complaint includes common root cause keywords, subjective root cause keywords, and objective root cause keywords.

[0051] Furthermore, the complaint root cause tracing device also includes:

[0052] The determining unit is used to determine subjective root cause keywords and objective root cause keywords based on the root cause tracing results of the complaint.

[0053] The determining unit is further configured to determine subjective root causes and objective root causes based on the subjective root cause keywords;

[0054] The second acquisition unit is used to acquire the first solution suggestion corresponding to the subjective root cause and the second solution suggestion corresponding to the objective root cause.

[0055] The output unit is used to output the first solution suggestion and the second solution suggestion.

[0056] A third aspect of this application provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor runs the computer program to cause the electronic device to perform the complaint root cause tracing method described in any one of the first aspects of this application.

[0057] A fourth aspect of this application provides a computer-readable storage medium storing computer program instructions, which, when read and executed by a processor, perform the complaint root cause tracing method described in any one of the first aspects of this application. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 A flowchart illustrating a method for tracing the root causes of complaints provided in this application embodiment;

[0060] Figure 2 A flowchart illustrating another method for tracing the root cause of a complaint provided in this application embodiment;

[0061] Figure 3 A schematic diagram of a complaint root cause tracing device provided in this application embodiment;

[0062] Figure 4 A schematic diagram of another complaint root cause tracing device provided in this application embodiment;

[0063] Figure 5 This is a schematic diagram illustrating an example architecture of a complaint graph provided in an embodiment of this application. Detailed Implementation

[0064] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0065] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0066] Example 1

[0067] Please refer to Figure 1 , Figure 1 This embodiment provides a flowchart illustrating a method for tracing the root causes of complaints. The method includes:

[0068] S101. Obtain complaint content data.

[0069] S102. The complaint content data is processed by a preset causal extraction pattern matching model to obtain complaint keywords.

[0070] S103. Extract keywords from the complaint keywords to obtain keywords for the reasons for the complaint.

[0071] S104. Construct a complaint map based on complaint keywords and complaint reason keywords.

[0072] S105. Based on the complaint map, trace the root causes of the complaints to obtain the root cause tracing results.

[0073] In this embodiment, the existing bank credit card complaint classification standard is a five-level classification standard. This standard, from the perspective of banking operations, aims to quickly locate and resolve similar complaints for customers from a business standpoint. However, recently, it has been found that some customers have filed multiple complaints, each with a different business type, but the root cause of the complaint is the same. For example, "Consumer A called on the 8th to complain about needing to apply for a bill extension and negotiate repayment due to unemployment, which is classified as a negotiated repayment issue according to the five-level complaint label; on the 21st of the same month, A called again due to unemployment, requesting fee reduction, which is classified as an interest and late payment penalty issue according to the five-level complaint label." This shows that there may be some overlap in banking operations, and it also illustrates that complaint management should not only focus on the business perspective but also explore the root causes of complaints from the customer's perspective. Only in this way can we better investigate, statistically analyze, and research banking financial consumer complaint data, facilitating financial management departments to collect and analyze industry-wide, comprehensive complaint data. This will help to identify, discover, warn, and address common problems and risks across the industry early, providing data support for relevant work decisions. To achieve the aforementioned effects, this method proposes a root cause tracing approach for bank credit card business in the financial field. This method aims to trace the root causes of customer complaints through graph-based methods, providing data insights for business complaint governance.

[0074] In this embodiment, the subject executing the method can be a computing device such as a computer or server, and no limitation is made in this embodiment.

[0075] In this embodiment, the subject executing the method can also be a smart device such as a smartphone or tablet, and no limitation is made in this embodiment.

[0076] As can be seen, the complaint root cause tracing method described in this embodiment, based on the logical structure of "research background—data preparation—graph construction—graph mining—conclusion presentation—policy recommendations," uses credit card business complaint information as a sample. It comprehensively utilizes pattern matching, NLP technology, and graph algorithm technology to study and analyze the hot issues in financial consumer complaints. From the perspective of financial consumers, it fully explores the complaint points, analyzes the root causes of complaints, and examines the connections between complaint points. By fully considering the problems and needs reflected in financial consumer complaints in business operations, internal control, and risk management, it facilitates in-depth root cause tracing and rectification, thereby continuously improving the quality and efficiency of complaint management and effectively enhancing consumer satisfaction.

[0077] Example 2

[0078] Please refer to Figure 2 , Figure 2 This embodiment provides a flowchart illustrating a method for tracing the root causes of complaints. The method includes:

[0079] S201. Obtain complaint content data.

[0080] S202. The complaint content data is processed by a preset causal extraction pattern matching model to obtain complaint keywords.

[0081] S203. Segment the complaint keywords to obtain a keyword set.

[0082] S204. Perform word frequency statistics to sort the keyword set and obtain the keyword sorting sequence.

[0083] S205. Select a preset number of keywords from the keyword sorting sequence to obtain the target keywords.

[0084] S206. Filter the target keywords according to the pre-built stop word dictionary to obtain the keywords for the reasons for the complaint.

[0085] In this embodiment, the method can use a "causal extraction pattern matching model" to obtain key complaint information. Specifically, the method uses a pattern matching-based causal event pair extraction scheme to construct a "causal extraction pattern matching model" to extract key complaint information from the "complaint content". The model input is the "complaint content" field of the work order; the output is the key complaint information extracted from the "complaint content", which includes the "cause" and "effect" of the complaint.

[0086] The specific pattern matching schemes are shown in the table below:

[0087]

[0088]

[0089] In this embodiment, the method can use a "keyword extraction model" to obtain keywords related to the reasons for complaints. Specifically, the method establishes a "keyword extraction model" based on more than 30,000 pieces of key complaint information. The steps for establishing the "keyword extraction model" include:

[0090] (1) First, the key information of each complaint is segmented to obtain the corresponding keyword set;

[0091] (2) Rank the keyword set by word frequency and select the top 100 verb and noun keywords;

[0092] (3) Based on business experience, a stop word dictionary was established. Through the stop word filtering method, more than 60 keywords with practical significance for the reasons for complaints were finally obtained.

[0093] S207. Filter the complaint keywords and complaint reason keywords to obtain node data and edge data for constructing the complaint graph.

[0094] S208. Obtain the weight configuration information for node data and edge data.

[0095] S209. Construct a complaint graph based on node data, edge data, and weight configuration information.

[0096] In this embodiment, the most basic data structure in graph computation consists of three factors: vertex V, edge E, and weight D.

[0097] In this embodiment, based on the data preprocessing results from the first part and the validity screening of other fields, the following data fields are selected as nodes and edges for constructing the credit card complaint graph, thus building the complete picture of the credit card complaint graph. Specifically, the vertex V, edge E, and weight D are as follows:

[0098] (1) Node V:

[0099] Node Name quantity Customer (This node's attributes include customer ID, gender, branch, etc.) 25886 Level 4 Complaint Tag 9 Level 5 Complaint Label 169 Root cause keywords 63

[0100] (2) Edge E:

[0101] The side name includes customer-level five complaint label; level four complaint label-level five complaint label; customer-complaint reason keywords.

[0102] (3) Weight D:

[0103] The edge weights for "Customer - Level 5 Complaint Tag" and "Customer - Complaint Reason Keywords" are set according to the month of the complaint ticket, as follows:

[0104] D = 1 + m ÷ 10;

[0105] Where m represents the month, i.e., m=1 for complaints in January and m=12 for complaints in December.

[0106] S210. Calculate the median centrality of the complaint graph to obtain the median centrality of the root cause keywords.

[0107] S211. Determine the set of root cause keywords that have intersection based on the median centrality of root cause keywords and the preset complaint categories.

[0108] S212. Sort the root keywords in the root keyword set according to the median centrality of the root keywords to obtain the sorted keyword sequence.

[0109] S213. Determine the root cause tracing results of the complaint based on the sorted keyword sequence; wherein, the root cause tracing results of the complaint include common root cause keywords, subjective root cause keywords and objective root cause keywords.

[0110] S214. Determine the subjective root cause keywords and objective root cause keywords based on the root cause tracing results of the complaint.

[0111] S215. Determine the subjective root cause and objective root cause based on the keywords of the subjective root cause.

[0112] S216. Obtain the first solution suggestion corresponding to the subjective root cause and the second solution suggestion corresponding to the objective root cause.

[0113] S217. Output suggestions for the first and second solutions.

[0114] In this embodiment, because each consumer's complaint form is highly personalized and complex, in order to summarize common root causes of complaints at the level of fourth and fifth-level complaint tags, and to achieve collaborative handling and joint governance of issues across different fourth-level complaint tags, as well as reduce multiple complaints from the same consumer, it is necessary to calculate and analyze the root cause keywords of each complaint in the graph. Taking the top three fifth-level categories with the highest proportion of credit card business complaints as an example, the basic graph framework for calculating the centrality of root cause keywords among them can refer to... Figure 5 As shown.

[0115] exist Figure 5 In the middle column, the set within the second rectangle (a square-like box) corresponds to three common root causes, while the first and third rectangles (the top and bottom rectangles) in the middle column correspond to two common root causes. The leftmost rectangle, the bottom rectangle, and the top right rectangle correspond to single-class specific root causes.

[0116] For example, the three common root causes must include at least the following root cause keywords: pandemic, repayment, reduction, interest, fees, bills, installment payments, income, application, overdue, personnel, credit limit reduction, outstanding debt, customer service, and negotiation.

[0117] In this embodiment, all five-level categories can obtain a set of root cause keywords that overlap. After ranking the root cause keywords in the set according to their relevance, it can be determined whether there are a large number of common root causes among different categories, what they are, which are subjective root causes, which are objective root causes, and whether the objective root causes can be resolved. If they can be resolved, multiple complaint categories can be addressed together, achieving multiple benefits with one effort.

[0118] In this embodiment, the subject executing the method can be a computing device such as a computer or server, and no limitation is made in this embodiment.

[0119] In this embodiment, the subject executing the method can also be a smart device such as a smartphone or tablet, and no limitation is made in this embodiment.

[0120] As can be seen, the complaint root cause tracing method described in this embodiment, based on the logical structure of "research background—data preparation—graph construction—graph mining—conclusion presentation—policy recommendations," uses credit card business complaint information as a sample. It comprehensively utilizes pattern matching, NLP technology, and graph algorithm technology to study and analyze the hot issues in financial consumer complaints, fully exploring the complaint points from the perspective of financial consumers, judging the root causes of complaints, and analyzing the connections between complaint points. It fully considers the problems and needs reflected in financial consumer complaints in business operations, internal control, and risk management, thereby...

[0121] This facilitates in-depth investigation and rectification, thereby continuously improving the quality and efficiency of complaint management and effectively enhancing consumer satisfaction.

[0122] Example 3

[0123] Please refer to Figure 3 , Figure 3 This embodiment provides a schematic diagram of the structure of a complaint root cause tracing device.

[0124] Figure. Figure 3 As shown, the root cause tracing device for complaints includes:

[0125] 0 First acquisition unit 310, used to acquire complaint content data;

[0126] The first processing unit 320 is used to process the complaint content data through a preset causal extraction pattern matching model to obtain complaint keywords;

[0127] The second processing unit 330 is used to extract keywords from the complaint keywords to obtain keywords for the reason for the complaint.

[0128] 5. Building unit 340, used to construct a complaint graph based on complaint keywords and complaint reason keywords;

[0129] The tracing unit 350 is used to trace the root cause of complaints based on the complaint map and obtain the root cause tracing results.

[0130] In this embodiment, the explanation of the complaint root cause tracing device can be referred to the description in Embodiment 1 or Embodiment 2, and will not be repeated here.

[0131] As can be seen from the example, the complaint root cause tracing device described in this embodiment can, based on the logical structure of "research background—data preparation—graph construction—graph mining—conclusion presentation—policy recommendations," use credit card business complaint information as a sample, and comprehensively utilize pattern matching, NLP technology, and graph algorithm technology to study and analyze the hot issues of financial consumer complaints, fully exploring the root cause of complaints from the perspective of financial consumers.

[0132] The key to resolving complaints lies in analyzing the root causes and the connections between the various points of contention. By fully considering the issues and needs reflected in financial consumer complaints in business operations, internal controls, and risk management, it becomes easier to conduct thorough root-cause analysis and rectification, thereby continuously improving the quality and efficiency of complaint management and effectively enhancing consumer satisfaction.

[0133] Example 4

[0134] Please refer to Figure 4 , Figure 4 This is a schematic diagram of a complaint root cause tracing device provided in this embodiment. Figure 4 As shown, the root cause tracing device for complaints includes:

[0135] The first acquisition unit 310 is used to acquire complaint content data;

[0136] The first processing unit 320 is used to process the complaint content data through a preset causal extraction pattern matching model to obtain complaint keywords;

[0137] The second processing unit 330 is used to extract keywords from the complaint keywords to obtain keywords for the reason for the complaint.

[0138] Building unit 340 is used to construct a complaint graph based on complaint keywords and complaint reason keywords;

[0139] The tracing unit 350 is used to trace the root cause of complaints based on the complaint map and obtain the root cause tracing results.

[0140] As an optional implementation, the second processing unit 330 includes:

[0141] Sub-unit 331 is used to segment complaint keywords to obtain a keyword set;

[0142] The first sorting subunit 332 is used to perform word frequency statistical sorting on the keyword set to obtain the keyword sorting sequence;

[0143] Select sub-unit 333 to select a preset number of keywords from the keyword sorting sequence to obtain the target keywords;

[0144] The filtering subunit 334 is used to filter the target keywords according to the pre-built stop word dictionary to obtain the keywords of the reason for the complaint.

[0145] As an optional implementation, the building unit 340 includes:

[0146] The filtering subunit 341 is used to filter complaint keywords and complaint reason keywords to obtain node data and edge data for constructing the complaint graph;

[0147] Obtain subunit 342, used to obtain weight configuration information for node data and edge data;

[0148] Subunit 343 is constructed to build a complaint graph based on node data, edge data, and weight configuration information.

[0149] As an optional implementation, the traceability unit 350 includes:

[0150] Calculation subunit 351 is used to calculate the median centrality of the complaint graph and obtain the root cause keyword median centrality.

[0151] Subunit 352 is defined to determine the set of root cause keywords that have intersection based on the median centrality of root cause keywords and the preset complaint categories;

[0152] The second sorting subunit 353 is used to sort the root keywords in the root keyword set according to the root keyword mediation centrality to obtain the sorted keyword sequence.

[0153] Subunit 352 is also used to determine the root cause tracing results of complaints based on the sorted keyword sequence; wherein, the root cause tracing results of complaints include common root cause keywords, subjective root cause keywords and objective root cause keywords.

[0154] As an optional implementation, the complaint root cause tracing device further includes:

[0155] Unit 360 is used to determine subjective root cause keywords and objective root cause keywords based on the root cause tracing results of complaints.

[0156] Unit 360 is also used to determine subjective root causes and objective root causes based on subjective root cause keywords;

[0157] The second acquisition unit 370 is used to acquire the first solution suggestion corresponding to the subjective root cause and the second solution suggestion corresponding to the objective root cause.

[0158] Output unit 380 is used to output the first solution suggestion and the second solution suggestion.

[0159] In this embodiment, the explanation of the complaint root cause tracing device can be referred to the description in Embodiment 1 or Embodiment 2, and will not be repeated here.

[0160] As can be seen, the complaint root cause tracing device described in this embodiment can, based on the logical structure of "research background—data preparation—graph construction—graph mining—conclusion presentation—policy recommendations," use credit card business complaint information as a sample to comprehensively utilize pattern matching, NLP technology, and graph algorithm technology to study and analyze the hot issues in financial consumer complaints. It fully explores the complaint points from the perspective of financial consumers, judges the root causes of complaints, and analyzes the connections between complaint points. By fully considering the problems and needs reflected in financial consumer complaints in business operations, internal control, and risk management, it facilitates in-depth root cause tracing and rectification, thereby continuously improving the quality and efficiency of complaint management and effectively increasing consumer satisfaction.

[0161] This application provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the complaint root cause tracing method in embodiment 1 or embodiment 2 of this application.

[0162] This application provides a computer-readable storage medium storing computer program instructions, which are read and executed by a processor to perform the complaint root cause tracing method in embodiment 1 or embodiment 2 of this application.

[0163] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0164] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0165] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a 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 a portion 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 a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0166] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application. It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0167] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0168] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A complaint root cause tracing method, characterized by, include: Obtain complaint content data; The complaint content data is processed by a preset causal extraction pattern matching model to obtain complaint keywords; The complaint keywords are processed by keyword extraction to obtain the keywords for the reasons for the complaint; Construct a complaint graph based on the complaint keywords and the complaint reason keywords; Based on the complaint map, the root cause of the complaint is traced to obtain the root cause tracing results. The step of tracing the root causes of complaints based on the complaint map to obtain the root cause tracing results includes: The mediator centrality of the complaint graph is calculated to obtain the root cause keyword mediator centrality. The set of root cause keywords that intersects is determined based on the median centrality of the root cause keywords and the preset complaint categories; The root cause keywords in the root cause keyword set are sorted according to the root cause keyword betweenness centrality to obtain a sorted keyword sequence; The root cause tracing results of complaints are determined based on the sorted keyword sequence; wherein, the root cause tracing results of complaints include common root cause keywords, subjective root cause keywords, and objective root cause keywords.

2. The method for tracing the root cause of complaints according to claim 1, characterized in that, The keyword extraction process for the complaint keywords yields complaint reason keywords, including: The complaint keywords are segmented to obtain a keyword set; The keyword set is sorted by word frequency to obtain a keyword ranking sequence; Select a preset number of keywords from the keyword sorting sequence to obtain the target keywords; The target keywords are filtered based on a pre-built stop word dictionary to obtain keywords related to the reasons for complaints.

3. The method for tracing the root cause of complaints according to claim 1, characterized in that, The step of constructing a complaint graph based on the complaint keywords and the complaint reason keywords includes: The complaint keywords and the complaint reason keywords are filtered to obtain node data and edge data for constructing the complaint graph; Obtain the weight configuration information for the node data and the edge data; A complaint graph is constructed based on the node data, the edge data, and the weight configuration information.

4. The method for tracing the root cause of complaints according to claim 1, characterized in that, The method further includes: Based on the root cause tracing results of the complaints, subjective root cause keywords and objective root cause keywords are determined; Based on the aforementioned subjective root cause keywords, determine the subjective root cause and objective root cause; Obtain a first suggested solution for the subjective root cause and a second suggested solution for the objective root cause; Output the first solution suggestion and the second solution suggestion.

5. A device for tracing the root cause of a complaint, characterized in that, The complaint root cause tracing device includes: The first acquisition unit is used to acquire complaint content data; The first processing unit is used to process the complaint content data through a preset causal extraction pattern matching model to obtain complaint keywords; The second processing unit is used to perform keyword extraction processing on the complaint keywords to obtain complaint reason keywords; The construction unit is used to construct a complaint graph based on the complaint keywords and the complaint reason keywords; The tracing unit is used to trace the root cause of complaints based on the complaint map and obtain the root cause tracing result. The traceability unit includes: The calculation subunit is used to calculate the median centrality of the complaint graph and obtain the root cause keyword median centrality. Determine sub-units to identify the set of root cause keywords that intersect based on the mediator centrality of root cause keywords and the preset complaint categories; The second sorting subunit is used to sort the root keywords in the root keyword set according to the root keyword betweenness centrality, so as to obtain the sorted keyword sequence. The determination of sub-units is also used to determine the root cause tracing results of complaints based on the sorted keyword sequence; wherein, the root cause tracing results of complaints include common root cause keywords, subjective root cause keywords, and objective root cause keywords.

6. The complaint root cause tracing device according to claim 5, characterized in that, The second processing unit includes: The word segmentation subunit is used to segment the complaint keywords to obtain a keyword set; The first sorting subunit is used to perform word frequency statistical sorting on the keyword set to obtain a keyword sorting sequence; A sub-unit is selected to select a preset number of keywords from the keyword sorting sequence to obtain target keywords; The filtering subunit is used to filter the target keywords according to a pre-built stop word dictionary to obtain keywords for the reasons for complaints.

7. The complaint root cause tracing device according to claim 5, characterized in that, The building unit includes: The filtering subunit is used to filter the complaint keywords and the complaint reason keywords to obtain node data and edge data for constructing the complaint graph; A sub-unit is used to acquire weight configuration information for the node data and the edge data; A sub-unit is constructed to build a complaint graph based on the node data, the edge data, and the weight configuration information.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor running the computer program to cause the electronic device to perform the complaint root cause tracing method according to any one of claims 1 to 4.

9. A readable storage medium, characterized in that, The readable storage medium stores computer program instructions, which, when read and executed by a processor, perform the complaint root cause tracing method according to any one of claims 1 to 4.

Images