Malicious complaint behavior identification method and device and readable storage medium
By matching keyword information in the recognition complaint text in the knowledge base, the problems of low efficiency and insufficient accuracy of malicious complaint recognition in the prior art are solved, and near-real-time recognition and more comprehensive handling solutions are achieved.
Patent Information
- Application Number
- CN202411045469.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2025-02-21
AI Technical Summary
The prior art has problems of low identification efficiency and insufficient accuracy in identifying and handling malicious complaints, resulting in long processing cycles and limited processing solutions.
By obtaining pending complaint text, extracting keyword information, and matching and identifying it in a knowledge base built on historical malicious complaint incidents and legal and regulatory information, we will determine the identification results and handling plans for malicious complaint behavior.
It realizes near-real-time identification of malicious complaints, shortens the identification time, and improves the comprehensiveness and accuracy of the handling plan.
Smart Images
Figure CN119494341A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of Internet technology, and in particular to a method, device and readable storage medium for identifying malicious complaint behavior. Background Art
[0002] As companies pay more attention to and improve their service quality and implement their social responsibilities, when companies file complaints against customers, in some cases, they are also more concerned about the illegal issues of the complainants.
[0003] Currently, the industry focuses on the identification of illegal complaints. For example, when making a complaint, a person with extreme ideas threatens to take illegal actions against others, or makes a malicious complaint to achieve an impure purpose. For such complaints, when the customer service department of the company receives the complaint, it is necessary to accurately identify whether the complaint is a malicious complaint or illegal behavior in real time, and give a specific solution based on the characteristics of the incident. The current main identification and processing methods are usually carried out in the following ways:
[0004] When a complaint that may be malicious or illegal is received and the agent is unable to identify it, it will be reported step by step and ultimately received and handled by a specialist for this incident. If the specialist is still unable to determine whether the complaint is illegal at the time of the complaint, communication among multiple levels of personnel is required, and identification is made based on past experience of individuals and companies, or other resources may be sought through various channels to identify whether the complaint is illegal at the time of the complaint, and ultimately determine what response measures should be taken.
[0005] However, the existing processing methods have certain limitations. Therefore, how to improve the efficiency and accuracy of identifying malicious complaint behavior has become a problem that needs to be solved. Summary of the invention
[0006] The technical problem to be solved by the present application is to provide a method, device and readable storage medium for identifying malicious complaint behavior in order to solve the problems existing in the prior art.
[0007] In a first aspect, the present application provides a method for identifying malicious complaint behavior, the method comprising:
[0008] S1. Obtain a complaint text to be processed, and obtain keyword information according to the complaint text to be processed;
[0009] S2. Based on the keyword information, matching and identifying are performed in a knowledge base to obtain matching results, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information;
[0010] S3. Obtain identification results of malicious complaint behavior based on the matching results.
[0011] In some embodiments, in S1, keyword information is obtained according to the complaint text to be processed, including:
[0012] S11, performing word segmentation processing on the complaint text to be processed to obtain a first processing result;
[0013] S12. Perform keyword extraction on the first processing result to obtain a second processing result, wherein the second processing result includes keyword information of the complaint text to be processed.
[0014] In some embodiments, S12 includes:
[0015] Performing keyword extraction on the first processing result through a deep learning model to obtain the second processing result; the input of the deep learning model includes keywords of a preset type, and the output of the deep learning model includes a priority order corresponding to the keywords of the preset type;
[0016] The keywords of the preset type include at least one of the following:
[0017] The words whose percentage of occurrence in the current preset time period is greater than the preset threshold;
[0018] Words whose historical occurrence percentage is greater than the preset threshold;
[0019] The words whose percentage of occurrence in the knowledge base during the current preset time period is greater than the preset threshold;
[0020] Words whose historical occurrence percentage in the knowledge base is greater than the preset threshold.
[0021] In some embodiments, the process of building the knowledge base includes:
[0022] Obtain historical malicious complaint incidents and legal and regulatory information;
[0023] Filter data based on the historical malicious complaint events and legal and regulatory information to obtain knowledge data related to malicious complaint behaviors;
[0024] Classify the data based on the knowledge data related to the malicious complaint behavior and establish a corresponding index;
[0025] The knowledge base is constructed according to the knowledge data and the corresponding index.
[0026] In some embodiments, in S2, the matching result includes: identification matching information of the keyword information and the index; the identification matching information includes historical event matching information and / or legal and regulatory matching information;
[0027] The historical event matching information includes: search keywords, first result words retrieved based on the keywords, and historical event information associated with the first result words, wherein the historical event information at least includes event description content, processing method, and processing status;
[0028] The legal and regulatory matching information includes: search keywords, second result words retrieved based on the keywords, and legal and regulatory information associated with the second result words. The legal and regulatory information includes at least legal names, legal clauses, and legal contents.
[0029] In some embodiments, S3 includes:
[0030] If the matching result is that at least one of the historical event matching information and the legal and regulatory matching information is not empty, it is determined that malicious complaint behavior exists.
[0031] In some embodiments, it also includes:
[0032] S4. If it is determined that there is malicious complaint behavior, the corresponding identification results and corresponding processing solutions are output.
[0033] In a second aspect, the present application provides a malicious complaint behavior identification device, the device comprising:
[0034] A keyword acquisition module, which is configured to acquire a complaint text to be processed and obtain keyword information according to the complaint text to be processed;
[0035] An identification and matching module, which is configured to perform matching and identification in a knowledge base based on the keyword information to obtain a matching result, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information;
[0036] A malicious complaint identification module is configured to obtain an identification result of malicious complaint behavior based on the matching result.
[0037] In a third aspect, the present application provides a malicious complaint behavior identification device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the malicious complaint behavior identification method described in the first aspect above.
[0038] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for identifying malicious complaint behavior described in the first aspect above is implemented.
[0039] The present application provides a method, device and readable storage medium for identifying malicious complaint behaviors, the method comprising: obtaining a complaint text to be processed, obtaining keyword information based on the complaint text to be processed; performing matching and identification in a knowledge base based on the keyword information to obtain a matching result, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information; obtaining the identification result of malicious complaint behaviors based on the matching result. The present application provides a method for identifying malicious complaint behaviors, which has the following beneficial effects: 1. Near-real-time identification, shortening the identification time: since the complaint content is manually entered, the method and device can capture the complaint text in real time, and achieve near-real-time identification and processing, which can greatly shorten the processing cycle. 2. The processing scheme is more comprehensive: the identification basis for identifying whether it is illegal when complaining comes from the input of current legal provisions, and the identification result is more accurate. After the corresponding identification of illegality when complaining, the corresponding processing scheme comes from the input of complaint events that have occurred internally and externally and the input of relevant network information, and the processing scheme given accordingly is more comprehensive. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0041] Figure 1 A schematic diagram of a malicious complaint behavior identification system provided in an embodiment of the present application;
[0042] Figure 2 A schematic diagram of a process for expanding the content of the knowledge base module 30 provided in an embodiment of the present application;
[0043] Figure 3 A flowchart of a method for identifying malicious complaint behavior provided in an embodiment of the present application;
[0044] Figure 4 A schematic diagram of the structure of a malicious complaint behavior identification device provided in an embodiment of the present application;
[0045] Figure 5 A schematic diagram of the structure of another malicious complaint behavior identification device provided in an embodiment of the present application.
[0046] Description of Figure Numbers:
[0047] 10. Real-time complaint information preprocessing module; 11. Complaint information collection component; 12. Word segmentation component; 13. Keyword extraction component; 20. Knowledge growth module; 21. Knowledge expansion component; 22. Business phrase configuration component; 30. Knowledge base module; 31. Complaint event library; 32. Legal knowledge base; 40. Result aggregation module; 41. Result identification component; 42. Processing solution response component; 50. Notification module; 51. External communication interface calling component; 110. Keyword acquisition module; 120. Identification and matching module; 130. Malicious complaint identification module; 220. Memory; 210. Processor.
[0048] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION
[0049] In order to enable those skilled in the art to better understand the technical solution of the present application, the implementation mode of the present application will be further described in detail below with reference to the accompanying drawings.
[0050] It should be understood that the specific embodiments and drawings described herein are only used to explain the present application, rather than to limit the present application.
[0051] It can be understood that, in the absence of conflict, the various embodiments in the present application and the various features in the embodiments can be combined with each other.
[0052] It can be understood that, for the convenience of description, the drawings of the present application only show the parts related to the present application, while the parts unrelated to the present application are not shown in the drawings.
[0053] It can be understood that each unit and module involved in the embodiments of the present application may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.
[0054] It can be understood that the terms "first", "second", etc. in the embodiments of the present application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.
[0055] It is understandable that, in the absence of conflict, the functions and steps marked in the flowcharts and block diagrams of the present application may occur in an order different from that marked in the drawings.
[0056] It is understood that the flowcharts and block diagrams of the present application illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present application. Among them, each box in the flowchart or block diagram may represent a unit, module, program segment, code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or by a combination of hardware and computer instructions.
[0057] It can be understood that the units and modules involved in the embodiments of the present application can be implemented by software or hardware, for example, the units and modules can be located in a processor.
[0058] The existing approach to handling malicious complaints has the following problems:
[0059] 1. The current identification cycle of illegality during complaints is long: Currently, the identification of whether a complaint is illegal is usually based on communication between personnel. This process consumes more time and cannot identify and take corresponding measures in real time when a real complaint is illegal;
[0060] 2. The current solution for handling illegal complaints has limitations: the current identification process is mainly based on personal experience and corporate experience, supplemented by temporary network and legal information, communication and coordination. The subjectivity and cognitive level of the identification personnel have a great impact on the identification results. For example, a novice customer service representative will inevitably be less skilled than an experienced customer service representative in handling problems.
[0061] In order to solve the problems of long identification cycle and limitations of handling solutions in existing technical solutions for violations at the time of complaints, this application proposes to identify the current complaint text to be processed based on the knowledge of laws and regulations, internal and external complaints that have occurred, and relevant network information, to determine whether it is a violation of the law at the time of the complaint, and to provide an immediate handling solution.
[0062] Compared with the prior art, the advantages of the solution proposed in this application are:
[0063] 1. Near real-time recognition, shortening the recognition time: Since the complaint content is manually entered, the method and device can capture the complaint text in real time, and achieve near real-time recognition and processing, which can greatly shorten the processing cycle.
[0064] 2. The processing plan is more comprehensive: The basis for identifying whether the complaint is illegal comes from the input of current legal provisions, so the result of the identification is more accurate. After the corresponding complaint is identified as illegal, the corresponding processing plan comes from the input of internal and external complaint events and related network information, so the processing plan given is more comprehensive.
[0065] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.
[0066] In order to facilitate the real-time identification and processing of malicious complaint behaviors, this application first proposes a malicious complaint behavior identification system.
[0067] Figure 1 A schematic diagram of a malicious complaint behavior identification system provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the malicious complaint behavior identification system includes five modules and ten components therein.
[0068] refer to Figure 1 The five modules and ten components of the system include:
[0069] (1) Real-time complaint information preprocessing module 10, as well as internal complaint information collection component 11, word segmentation component 12, and keyword extraction component 13;
[0070] (2) a knowledge growth module 20, and an internal knowledge expansion component 21 and a business phrase configuration component 22;
[0071] (3) Knowledge base module 30, as well as internal complaint event database 31 and legal knowledge database 32;
[0072] (4) a result aggregation module 40, and an internal result identification component 41 and a processing solution response component 42;
[0073] (5) Notification module 50, and internal external communication interface calling component 51.
[0074] refer to Figure 1 , wherein the real-time complaint information preprocessing module 10 is used to capture the complaint text to be processed, and after processing it accordingly, output the keywords to be processed to the knowledge base module 30;
[0075] Among them, the knowledge growth module 20 is used to receive legal and regulatory knowledge input, internal and external complaint incidents that have occurred input and related network information input, and after matching and filtering according to the input business phrase content, finally output the relevant legal and regulatory knowledge, internal and external complaint incidents that have occurred knowledge and other related knowledge to the knowledge base module 30.
[0076] The knowledge base module 30 is used to receive the knowledge output by the knowledge growth module 20 and maintain two index libraries internally. It also receives the keywords to be processed output by the real-time complaint information preprocessing module 10 and uses the keywords to be processed to match the recognition results and processing solutions in the two internal index libraries respectively. The recognition results and processing solutions are encapsulated as array types, and finally the two arrays are encapsulated into a two-dimensional array, which is output to the result aggregation module 40.
[0077] The result aggregation module 40 is used to receive the two-dimensional array output by the knowledge base module 30 , obtain the most suitable recognition result and processing solution in the array, and package the final result and output it to the notification module 50 .
[0078] The notification module 50 receives the recognition results and processing solutions output by the result aggregation module 40 and notifies the relevant processors via external communication.
[0079] The following are the details of each component in each module:
[0080] refer to Figure 1 The real-time complaint information preprocessing module 10 includes a complaint information collection component 11, a word segmentation component 12, and a keyword extraction component 13.
[0081] The complaint information collection component 11 receives the complaint text to be processed input from the outside, and outputs the text to the word segmentation component 12. The word segmentation component 12 is used to perform word segmentation on the text, remove stop words, and output the remaining sentences to the keyword extraction component 13. Keyword extraction mainly involves the accurate identification of real-time keywords. Then, for the deep learning model of keyword extraction, the input parameters of the algorithm model are: "words with a percentage of occurrence greater than the preset threshold in the current preset time period", "words with a percentage of historical occurrence greater than the preset threshold", "words with a percentage of occurrence greater than the preset threshold in the knowledge base in the current preset time period", "words with a percentage of historical occurrence greater than the preset threshold in the knowledge base". The output is the priority order of the keywords in the text content.
[0082] The following is an explanation of these four keywords:
[0083] (1) "Words whose percentage of occurrence in the current preset time period is greater than the preset threshold": select several words with a higher percentage of occurrence in the past month (or other time period);
[0084] (2) “Words with a historical occurrence percentage greater than a preset threshold”: select several words with a higher percentage of occurrence in the past three years (or other time period);
[0085] (3) "Words whose percentage of occurrence in the knowledge base in the current preset time period is greater than the preset threshold": select several words with a higher percentage of occurrence in the knowledge base in the past month (or other time period);
[0086] (4) "Words whose percentage of historical appearance in the knowledge base is greater than a preset threshold": select several words with a higher percentage of appearance in the knowledge base in the past three years (or other time period).
[0087] The preset thresholds corresponding to the above four keywords may be the same or different, and may be set according to actual conditions.
[0088] The following is a simple implementation of the deep learning process of the algorithm:
[0089] Convolutional layer 1
[0090] Input: words whose occurrence percentage in the current preset time period is greater than the preset threshold
[0091] Number of filters: 32
[0092] Filter size: (3, 3)
[0093] Activation function: ReLu
[0094] Pooling layer: 2x2 max pooling
[0095] Convolutional layer 2:
[0096] Input: words whose historical occurrence percentage is greater than the preset threshold
[0097] Number of filters: 64
[0098] Filter size: (3, 3)
[0099] Activation function: ReLu
[0100] Pooling layer: 2x2 max pooling
[0101] Convolutional layer 3:
[0102] Input: words whose percentage of occurrence in the knowledge base during the current preset time period is greater than the preset threshold
[0103] Number of filters: 128
[0104] Filter size: (3, 3)
[0105] Activation function: ReLU
[0106] Pooling layer: 2x2 max pooling
[0107] Volume layer 4:
[0108] Input: words whose historical occurrence percentage in the knowledge base is greater than the preset threshold
[0109] Number of filters: 256
[0110] Filter size: (3, 3)
[0111] Activation function: ReLu
[0112] Pooling layer: 2x2 max pooling
[0113] Fully connected layer 1:
[0114] Number of neurons: 256
[0115] Activation function: ReLu
[0116] Dropout layer: dropout rate of 0.5
[0117] Output Layer
[0118] Number of neurons: 1 (because it is a binary classification task)
[0119] Activation function: Sigmoid (output priority percentage)
[0120] Through the above deep learning algorithm, a result judgment for the four arrays can be obtained. By resetting the weight of each element, a priority confirmation relationship for the four arrays, namely real-time keywords, can be obtained. The keyword extraction component 13 finally extracts keywords and outputs them to the knowledge base module 30.
[0121] refer to Figure 1 , the knowledge growth module 20 includes a knowledge expansion component 21 and a business phrase configuration component 22.
[0122] The knowledge expansion component 21 will periodically capture external knowledge and filter it according to the business phrases configured in the business phrase configuration component 22 , and then output the result to the knowledge base module 30 to expand the content of the knowledge base module 30 .
[0123] refer to Figure 1 , the knowledge base module 30 includes a complaint event base 31 and a legal knowledge base 32.
[0124] Figure 2 A flow chart of expanding the content of the knowledge base module 30 provided in the embodiment of the present application is as follows: Figure 2 As shown, the expansion mainly includes the following steps:
[0125] (1) Define a list of data source addresses: including defining the network addresses of relevant laws and regulations, defining the addresses of internal and external complaint platforms, and defining relevant network addresses.
[0126] (2) Regularly capture data from the three types of addresses defined in step (1).
[0127] (3) Filter the data captured in the above step (2), and finally retain only the knowledge data containing the business phrases configured in the business phrase configuration component 22.
[0128] (4) Classify the data after filtering and cleaning in step (3). The principle of classification is: data that complies with laws and regulations proceeds to step (6), and other data, such as business and complaint-related data, proceeds to step (5).
[0129] (5) Legal knowledge content will enter the legal knowledge base 32 in the knowledge base module 30 and create an index.
[0130] (6) The content of the complaint event will enter the complaint event library 31 in the knowledge base module 30 and create an index.
[0131] The classification method here is that if the address defined in step (1) is a legal and regulatory network address, the captured data is data that complies with laws and regulations, and other addresses are uniformly entered into the complaint event library 31.
[0132] In the present application, the index data of the legal knowledge base 32 is derived from the legal knowledge content output by the knowledge growth module 20. The index is retrieved by the keywords output by the real-time complaint information preprocessing module 10, and the recognition result array arra[n] is output, and the value of n can usually be set.
[0133] Each element in the recognition result array contains the following information, and the sample content is shown in Table 1 below:
[0134]
[0135] Table 1
[0136] Wherein: search_word: the keyword to be searched this time output by the real-time complaint information preprocessing module 10.
[0137] find_word: the result word retrieved from the legal knowledge base 32 according to the keywords to be searched this time.
[0138] crime_name: The crime name in the laws and regulations associated with the retrieved words.
[0139] legal_name: The name of the law to which the crime belongs.
[0140] legal_index: The number of the law in which the crime is located.
[0141] legal_provision: The original content of the law.
[0142] In this application, the index data of the complaint event library 31 is derived from the complaint event content output by the knowledge growth module 20. The index will be retrieved by the keywords output by the real-time complaint information preprocessing module 10 in 1 above, and the processing solution array arrb[n] is output, and the value of n can usually be set.
[0143] Each element in the solution array contains the following information list, and the sample content is shown in Table 2 below:
[0144]
[0145] Table 2
[0146] Wherein, search_word: the keyword to be searched this time output by the real-time complaint information preprocessing module 10 .
[0147] find_word: the result word retrieved from the complaint event database 31 according to the keyword to be searched this time.
[0148] complainted: Based on the retrieved words, the description of the complaint incident that has occurred is associated.
[0149] action_of_complained: The method adopted by the complainee to defend his or her legitimate rights and interests in this complaint incident.
[0150] status: The processing status of the historical complaint event.
[0151] The recognition result array arra[n] and the processing solution array arrb[n] described above are combined into a two-dimensional array arr[i][n] (i=0,1) in this module and output to the result aggregation module 40. The two-dimensional array satisfies the following conditions:
[0152]
[0153] refer to Figure 1 The result aggregation module 40 includes a result identification component 41 and a processing solution response component 42 .
[0154] The result recognition component 41 obtains arr[0][n] in the two-dimensional array arr[i][n] (i=0, 1) output by the knowledge base module 30, and obtains a result that best meets the conditions.
[0155] The processing solution response component 42 obtains arr[1][n] in the two-dimensional array arr[i][n] (i=0,1) output by the knowledge base module 30, and obtains a result that best meets the conditions.
[0156] If the content of the above two-dimensional array arr[i][n] (i=0,1) is empty, or one of them is empty, a default value can be given.
[0157] The identification result array and the processing solution array are finally aggregated in the processing solution response component 42 , and the result is output to the notification module 50 .
[0158] refer to Figure 1 The notification module 50 includes an external communication interface calling component 51, which specifically obtains the content including the recognition result and the processing solution output by the result aggregation module 40, and calls the external communication interface to output the recognition result and the processing solution.
[0159] This application provides a malicious complaint behavior identification system, which has the following beneficial effects:
[0160] 1. Near real-time recognition, shortening the recognition time: Since the complaint content is manually entered, the method and device can capture the complaint text in real time, and achieve near real-time recognition and processing, which can greatly shorten the processing cycle.
[0161] 2. The processing plan is more comprehensive: The basis for identifying whether the complaint is illegal comes from the input of current legal provisions, so the result of the identification is more accurate. After the corresponding complaint is identified as illegal, the corresponding processing plan comes from the input of internal and external complaint events and related network information, so the processing plan given is more comprehensive.
[0162] Based on the above-mentioned malicious complaint behavior identification system, the present application provides a malicious complaint behavior identification method. In addition to the above-mentioned system, the working process of this method can also be implemented through electronic devices, such as computers, handheld smart terminals, etc.
[0163] Figure 3 A schematic diagram of a method for identifying malicious complaint behavior provided in an embodiment of the present application, such as Figure 3 As shown, the present application provides a method for identifying malicious complaint behavior, the method includes S1-S3, which are as follows:
[0164] S1. Obtain a complaint text to be processed, and obtain keyword information according to the complaint text to be processed;
[0165] In some embodiments, in S1, keyword information is obtained according to the complaint text to be processed, including:
[0166] S11, performing word segmentation processing on the complaint text to be processed to obtain a first processing result;
[0167] S12. Perform keyword extraction on the first processing result to obtain a second processing result, wherein the second processing result includes keyword information of the complaint text to be processed.
[0168] In some embodiments, S12 includes:
[0169] Performing keyword extraction on the first processing result through a deep learning model to obtain the second processing result; the input of the deep learning model includes keywords of a preset type, and the output of the deep learning model includes a priority order corresponding to the keywords of the preset type;
[0170] The keywords of the preset type include at least one of the following:
[0171] The words whose percentage of occurrence in the current preset time period is greater than the preset threshold;
[0172] Words whose historical occurrence percentage is greater than the preset threshold;
[0173] The words whose percentage of occurrence in the knowledge base during the current preset time period is greater than the preset threshold;
[0174] Words whose historical occurrence percentage in the knowledge base is greater than the preset threshold.
[0175] S2. Based on the keyword information, matching and identifying are performed in a knowledge base to obtain matching results, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information;
[0176] In some embodiments, the process of building the knowledge base includes:
[0177] Obtain historical malicious complaint incidents and legal and regulatory information;
[0178] Filter data based on the historical malicious complaint events and legal and regulatory information to obtain knowledge data related to malicious complaint behaviors;
[0179] Classify the data based on the knowledge data related to the malicious complaint behavior and establish a corresponding index;
[0180] The knowledge base is constructed according to the knowledge data and the corresponding index.
[0181] In some embodiments, in S2, the matching result includes: identification matching information of the keyword information and the index; the identification matching information includes historical event matching information and / or legal and regulatory matching information;
[0182] The historical event matching information includes: search keywords, first result words retrieved based on the keywords, and historical event information associated with the first result words, wherein the historical event information at least includes event description content, processing method, and processing status;
[0183] The legal and regulatory matching information includes: search keywords, second result words retrieved based on the keywords, and legal and regulatory information associated with the second result words. The legal and regulatory information includes at least legal names, legal clauses, and legal contents.
[0184] S3. Obtain identification results of malicious complaint behavior based on the matching results.
[0185] In some embodiments, S3 includes:
[0186] If the matching result is that at least one of the historical event matching information and the legal and regulatory matching information is not empty, it is determined that malicious complaint behavior exists.
[0187] In some embodiments, it also includes:
[0188] S4. If it is determined that there is malicious complaint behavior, the corresponding identification results and corresponding processing solutions are output.
[0189] Regarding the limitation of the method for identifying malicious complaint behavior, please refer to the limitation of the system for identifying malicious complaint behavior in the above embodiments of the present application, which will not be repeated in this embodiment.
[0190] This application provides a method for identifying malicious complaint behavior, which has the following beneficial effects:
[0191] 1. Near real-time recognition, shortening the recognition time: Since the complaint content is manually entered, the method and device can capture the complaint text in real time, and achieve near real-time recognition and processing, which can greatly shorten the processing cycle.
[0192] 2. The processing plan is more comprehensive: The basis for identifying whether the complaint is illegal comes from the input of current legal provisions, so the result of the identification is more accurate. After the corresponding complaint is identified as illegal, the corresponding processing plan comes from the input of internal and external complaint events and related network information, so the processing plan given is more comprehensive.
[0193] It should be understood that, although the various steps in the flowcharts in the above-described embodiments are sequentially displayed according to the indications of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and they can be executed in other orders. Moreover, at least a portion of the steps in the figure may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily to be carried out sequentially, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
[0194] Figure 4 A schematic diagram of a malicious complaint behavior identification device provided in an embodiment of the present application, such as Figure 4 As shown, the present application provides a malicious complaint behavior identification device, the device comprising:
[0195] A keyword acquisition module 110, which is configured to acquire a complaint text to be processed and obtain keyword information according to the complaint text to be processed;
[0196] An identification and matching module 120, which is configured to perform matching and identification in a knowledge base based on the keyword information to obtain a matching result, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information;
[0197] The malicious complaint identification module 130 is configured to obtain an identification result of the malicious complaint behavior based on the matching result.
[0198] Regarding the limitation of the malicious complaint behavior identification device, please refer to the limitation of the malicious complaint behavior identification method in the above embodiments of the present application, and this embodiment will not be repeated here.
[0199] Figure 5 Another schematic diagram of a malicious complaint behavior identification device provided for an embodiment of the present application is shown in Figure y. In some embodiments, the present application provides a malicious complaint behavior identification device, including a memory 220 and a processor 210, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the malicious complaint behavior identification method in the above-mentioned embodiments of the present application.
[0200] The memory is connected to the processor, the memory may be a flash memory or a read-only memory or other memory, and the processor may be a central processing unit or a single-chip microcomputer.
[0201] In some embodiments, the present application provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for identifying malicious complaint behavior in the above-mentioned embodiments of the present application is implemented.
[0202] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.
[0203] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present application, but the present application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of the present application, and these modifications and improvements are also considered to be within the scope of protection of the present application.
Claims
1. A method for identifying malicious complaint behavior, characterized in that: The method comprises: S1. Obtain a complaint text to be processed, and obtain keyword information according to the complaint text to be processed; S2. Based on the keyword information, matching and identifying are performed in a knowledge base to obtain matching results, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information; S3. Obtain identification results of malicious complaint behavior based on the matching results.
2. The method for identifying malicious complaint behavior according to claim 1 is characterized in that: In S1, keyword information is obtained according to the complaint text to be processed, including: S11, performing word segmentation processing on the complaint text to be processed to obtain a first processing result; S12. Perform keyword extraction on the first processing result to obtain a second processing result, wherein the second processing result includes keyword information of the complaint text to be processed.
3. The method for identifying malicious complaint behavior according to claim 2 is characterized in that: S12, including: Performing keyword extraction on the first processing result through a deep learning model to obtain the second processing result; the input of the deep learning model includes keywords of a preset type, and the output of the deep learning model includes a priority order corresponding to the keywords of the preset type; The keywords of the preset type include at least one of the following: The words whose percentage of occurrence in the current preset time period is greater than the preset threshold; Words whose historical occurrence percentage is greater than the preset threshold; The words whose percentage of occurrence in the knowledge base during the current preset time period is greater than the preset threshold; Words whose historical occurrence percentage in the knowledge base is greater than the preset threshold.
4. The method for identifying malicious complaint behavior according to claim 1 is characterized in that: The process of constructing the knowledge base includes: Obtain historical malicious complaint incidents and legal and regulatory information; Filter data based on the historical malicious complaint events and legal and regulatory information to obtain knowledge data related to malicious complaint behaviors; Classify the data based on the knowledge data related to the malicious complaint behavior and establish a corresponding index; The knowledge base is constructed according to the knowledge data and the corresponding index.
5. The method for identifying malicious complaint behavior according to claim 4 is characterized in that: In S2, the matching result includes: identification matching information between the keyword information and the index; the identification matching information includes historical event matching information and / or legal and regulatory matching information; The historical event matching information includes: search keywords, first result words retrieved based on the keywords, and historical event information associated with the first result words, wherein the historical event information at least includes event description content, processing method, and processing status; The legal and regulatory matching information includes: search keywords, second result words retrieved based on the keywords, and legal and regulatory information associated with the second result words. The legal and regulatory information includes at least legal names, legal clauses, and legal contents.
6. The method for identifying malicious complaint behavior according to claim 5 is characterized in that: S3, including: If the matching result is that at least one of the historical event matching information and the legal and regulatory matching information is not empty, it is determined that malicious complaint behavior exists.
7. The method for identifying malicious complaint behavior according to any one of claims 1 to 6, characterized in that: Also includes: S4. If it is determined that there is malicious complaint behavior, the corresponding identification results and corresponding processing solutions are output.
8. A malicious complaint behavior identification device, characterized in that: The device comprises: A keyword acquisition module, which is configured to acquire a complaint text to be processed and obtain keyword information according to the complaint text to be processed; An identification and matching module, which is configured to perform matching and identification in a knowledge base based on the keyword information to obtain a matching result, wherein the knowledge base is constructed based on historical malicious complaint events and legal and regulatory information; A malicious complaint identification module is configured to obtain an identification result of malicious complaint behavior based on the matching result.
9. A malicious complaint behavior identification device, characterized in that: It includes a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to implement the malicious complaint behavior identification method as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method for identifying malicious complaint behavior as described in any one of claims 1-7 is implemented.