Methods, devices, electronic equipment and storage media for predicting repeat complaint behavior

Abstract

This application provides a method, apparatus, electronic device, and storage medium for predicting repeat complaint behavior, relating to the field of computer technology, for predicting users with repeat complaint behavior. The method includes: acquiring historical user data features of multiple historical complaint users and a first feature and a second feature of a target user; the target user is any one of the multiple historical complaint users; a historical complaint user is a user who has filed a complaint once within a historical time period; the first feature reflects the user data features of the target user within a first preset time period before the complaint; the second feature reflects the user data features of the target user within a second preset time period after the complaint; determining the correlation between the first feature and the second feature to obtain a first correlation coefficient; determining the correlation between the historical user data features and the sum of the first and second features to obtain a second correlation coefficient; and determining the repeat complaint behavior of the target user based on the first correlation coefficient and the second correlation coefficient.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, electronic device, and storage medium for predicting recurring complaint behavior. Background Technology

[0002] Existing technologies for handling repeat complaints are lagging. When a user is dissatisfied with the outcome and files a second complaint, the operator can only confirm that the user is a repeat complainant by checking historical complaint records. This makes it impossible to predict and block repeat complainants, and thus impossible to take relevant measures in advance to prevent repeat complaint behavior from occurring. Therefore, there is an urgent need to design a method for predicting repeat complaint behavior. Summary of the Invention

[0003] This application provides a method, apparatus, electronic device, and storage medium for predicting repeat complaint behavior, used to predict users with repeat complaint behavior, enabling operators to take relevant measures to prevent repeat complaint behavior from occurring.

[0004] To achieve the above objectives, this application adopts the following technical solution:

[0005] Firstly, a method for predicting repeat complaint behavior is provided. The method includes: acquiring historical user data features of multiple historical complaining users and a first feature and a second feature of a target user; the target user is any one of the multiple historical complaining users; a historical complaining user is a user who has complained once within a historical time period; the first feature reflects the user data features of the target user within a first preset time period before the complaint; the second feature reflects the user data features of the target user within a second preset time period after the complaint; determining the correlation between the first feature and the second feature to obtain a first correlation coefficient; the first correlation coefficient is proportional to the probability of the target user engaging in repeat complaint behavior; determining the correlation between the historical user data features and the sum of the first feature and the second feature to obtain a second correlation coefficient; the second correlation coefficient is proportional to the probability of the target user engaging in repeat complaint behavior; and determining the repeat complaint behavior of the target user based on the first correlation coefficient and the second correlation coefficient.

[0006] Optionally, the repeated complaint behavior of the target user is determined based on the first correlation coefficient and the second correlation coefficient, including: determining the deviation between the first correlation coefficient and the second correlation coefficient; if the deviation is less than or equal to a preset deviation, weighting the first correlation coefficient and the second correlation coefficient to obtain a weighted result, and determining the repeated complaint behavior of the target user based on the weighted result; if the deviation is greater than the preset deviation, adjusting the first correlation coefficient and the second correlation coefficient until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and determining the repeated complaint behavior of the target user based on the adjusted first correlation coefficient and the second correlation coefficient.

[0007] Optionally, the first correlation coefficient and the second correlation coefficient are weighted to obtain a weighted result, including: obtaining the first weight of the first correlation coefficient and the second weight of the second correlation coefficient; weighting the first correlation coefficient according to the first weight to obtain a first weighted correlation coefficient, and weighting the second correlation coefficient according to the second weight to obtain a second weighted correlation coefficient; and determining the sum of the first weighted correlation coefficient and the second weighted correlation coefficient as the weighted result.

[0008] Optionally, the repeated complaint behavior of the target user can be determined based on the weighted result, including: determining that the target user will make repeated complaints if the weighted result is greater than a preset threshold; and determining that the target user will not make repeated complaints if the weighted result is less than or equal to the preset threshold.

[0009] Optionally, adjusting the first correlation coefficient and the second correlation coefficient includes: obtaining a third feature of the target user; the third feature is used to reflect the user data characteristics within a third preset time period before the target user makes a complaint; the third preset time period is shorter than the first preset time period; replacing the first feature with the third feature, and adjusting the first correlation coefficient and the second correlation coefficient according to the replaced first feature.

[0010] Secondly, a device for predicting repeat complaint behavior is provided. The device includes an acquisition unit and a determination unit. The acquisition unit is used to acquire historical user data features of multiple historical complaining users and a first feature and a second feature of a target user. The target user is any one of the multiple historical complaining users. A historical complaining user is a user who has complained once within a historical time period. The first feature reflects the user data features of the target user within a first preset time period before the complaint was made. The second feature reflects the user data features of the target user within a second preset time period after the complaint was made. The determination unit is used to determine the correlation between the first feature and the second feature to obtain a first correlation coefficient. The first correlation coefficient is proportional to the probability that the target user will make repeat complaints. The determination unit is used to determine the correlation between the historical user data features and the sum of the first feature and the second feature to obtain a second correlation coefficient. The second correlation coefficient is proportional to the probability that the target user will make repeat complaints. Based on the first correlation coefficient and the second correlation coefficient, the repeat complaint behavior of the target user is determined.

[0011] Optionally, the determining unit is specifically used for: determining the deviation between the first correlation coefficient and the second correlation coefficient; when the deviation is less than or equal to a preset deviation, weighting the first correlation coefficient and the second correlation coefficient to obtain a weighted result, and determining the target user's repeated complaint behavior based on the weighted result; when the deviation is greater than the preset deviation, adjusting the first correlation coefficient and the second correlation coefficient until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and determining the target user's repeated complaint behavior based on the adjusted first correlation coefficient and the second correlation coefficient.

[0012] Optionally, the determining unit is specifically used for: obtaining the first weight of the first correlation coefficient and the second weight of the second correlation coefficient; weighting the first correlation coefficient according to the first weight to obtain the first weighted correlation coefficient, and weighting the second correlation coefficient according to the second weight to obtain the second weighted correlation coefficient; and determining the sum of the first weighted correlation coefficient and the second weighted correlation coefficient as the weighted result.

[0013] Optionally, the determining unit is specifically used to: determine that the target user will file a repeat complaint if the weighted result is greater than a preset threshold; and determine that the target user will not file a repeat complaint if the weighted result is less than or equal to the preset threshold.

[0014] Optionally, the determining unit is specifically used for: obtaining a third feature of the target user; the third feature is used to reflect the user data characteristics within a third preset time period before the target user makes a complaint; the third preset time period is shorter than the first preset time period; replacing the first feature with the third feature, and adjusting the first correlation coefficient and the second correlation coefficient according to the replaced first feature.

[0015] Thirdly, an electronic device is provided, comprising: a processor and a memory for storing processor-executable instructions; wherein the processor is configured to execute instructions to implement the method for predicting recurring complaint behavior described in the first aspect above.

[0016] Fourthly, a computer-readable storage medium is provided, on which instructions are stored, such that when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is able to perform the method for predicting recurring complaint behavior as described in the first aspect above.

[0017] The technical solution provided in this application provides at least the following beneficial effects: The prediction device acquires historical user data features of multiple historical complaining users and a first feature and a second feature of the target user. The target user is any one of the multiple historical complaining users, and a historical complaining user is a user who has complained once within a historical time period. The first feature reflects the user data features within a first preset time period before the target user made a complaint, and the second feature reflects the user data features within a second preset time period after the target user made a complaint. Further, the prediction device determines the correlation between the first feature and the second feature to obtain a first correlation coefficient, and determines the correlation between the historical user data features and the sum of the first and second features to obtain a second correlation coefficient. This is equivalent to matching and calculating the complaint data before and after the repeated user complaint with the target user's behavioral habits. Since both the first and second correlation coefficients are proportional to the probability of the target user making repeated complaints, the larger the first and second correlation coefficients are, the greater the likelihood of the target user making repeated complaints. In related technologies, operators only determine which users are repeat complaining users after the repeated complaint behavior occurs, at which point the operator's reputation has already been affected. This application utilizes the characteristics of the first and second correlation coefficients to predict the repeat complaint behavior of target users, thereby enabling operators to take measures to prevent users from making repeat complaints, effectively resolve potential repeat complaint users, and thus improve the operator's reputation. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A schematic diagram of a wireless communication system structure is provided in an embodiment of this application;

[0020] Figure 2 A flowchart illustrating a method for predicting repeat complaint behavior provided in this application embodiment. Figure 1 ;

[0021] Figure 3 A flowchart illustrating a method for predicting repeat complaint behavior provided in this application embodiment. Figure 2 ;

[0022] Figure 4 A flowchart illustrating a method for predicting repeat complaint behavior provided in this application embodiment. Figure 3 ;

[0023] Figure 5 A flowchart illustrating a method for predicting repeat complaint behavior provided in this application embodiment. Figure 4 ;

[0024] Figure 6 This is a schematic diagram of the structure of a prediction device provided in an embodiment of this application;

[0025] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0028] It should also be noted that in the embodiments of this application, "of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be pointed out that when their differences are not emphasized, their meanings are consistent.

[0029] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0030] Before providing a detailed explanation of the embodiments of this application, some relevant technical terms and technologies involved in the embodiments of this application will be introduced first.

[0031] With the simultaneous growth of mobile network users and service volume, the demand for mobile data has exploded, and the requirements for telecommunications operators have evolved from simply increasing speed and reducing prices to ensuring both network performance and security. Users are becoming increasingly sensitive to network performance and more aware of their rights. When their mobile network experience is unsatisfactory, they will contact the operator's customer service hotline to file a complaint. If users are dissatisfied with the hotline's handling of their complaint, they will file a second complaint, resulting in what we call a "repeated complaint."

[0032] "Duplicate complaints" refer to a user who has already filed a complaint. If their complaint records within a pre-defined time period prior to this complaint are found to contain other complaints within that period, the user is considered a "duplicate complainant." Duplicate complaints negatively impact an operator's annual customer reputation evaluation, significantly affecting the operator's performance evaluation.

[0033] Existing technology involves operators classifying customer complaint information to determine the services a customer has subscribed to, performing retrospective matching to obtain historical service transaction information, generating service prediction work orders based on this historical information, and then using these work orders to predict user complaints. This method requires obtaining detailed information about the services subscribed to by the customer before model building and prediction can begin.

[0034] Existing technologies for handling repeat complaints are lagging. When a user is dissatisfied with the outcome and files a second complaint, the operator can only confirm that the user is a repeat complainant by checking historical complaint records. It is impossible to predict and block repeat complainants. Furthermore, existing technologies categorize complaints based on broad categories of issues, rather than predicting and blocking the behavior of individual customers.

[0035] In view of the above problems, this application provides a method for predicting repeat complaint behavior. The prediction device acquires the historical user data characteristics of multiple historical complainants, as well as the user data characteristics of the target user within a first preset time period before the complaint and the user data characteristics within a second preset time period after the complaint. The target user is any one of the multiple historical complainants. Based on the above data characteristics, a first correlation coefficient and a second correlation coefficient are determined to further determine the user's repeat complaint behavior, and then measures are taken to resolve potential repeat complaint users and improve the operator's reputation.

[0036] The communication method provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0037] The method for predicting repeat complaint behavior provided in this application can be applied to wireless communication systems. Figure 1 A schematic diagram of one structure of this wireless communication system is shown. For example... Figure 1As shown, the wireless communication system 10 includes a repeat complaint behavior prediction device (hereinafter referred to as the prediction device) 11 and a server 12. The prediction device 11 and the server 12 can be connected via a wired or wireless connection; this embodiment does not limit the connection in this way.

[0038] Server 12 can be a server from any mobile network operator. When a user files a service complaint with an operator, server 12 can record the user's identity information and the complaint information.

[0039] The prediction device 11 can obtain complaint information from any historical complainant from the server 12 and predict their repeat complaint behavior. Specific prediction methods can be found in the following embodiments, and will not be repeated here.

[0040] The prediction device 11 can be an electronic device with data processing capabilities. For example, the prediction device 11 can be a computer, tablet computer, ultra-mobile personal computer (UMPC), etc. The specific type of prediction device is not limited in the embodiments of this application.

[0041] Server 12 can be a single server or a server cluster. This application embodiment does not limit the specific type of server.

[0042] Figure 2 This is a flowchart illustrating a method for predicting repeat complaint behavior according to some exemplary embodiments. In some embodiments, the above-described method for predicting repeat complaint behavior can be applied to, for example... Figure 1 The prediction device shown can also be applied to other similar devices.

[0043] like Figure 2 As shown, the method for predicting repeat complaint behavior provided in this application includes the following steps S201-S204.

[0044] S201. Obtain historical user data features of multiple historical complaint users, as well as the first and second features of the target user.

[0045] The target user is any one of multiple historical complainants; a historical complainant is a user who has complained once within a historical time period; the first feature is used to reflect the user data characteristics within a first preset time period before the target user makes a complaint; the second feature is used to reflect the user data characteristics within a second preset time period after the target user makes a complaint.

[0046] One possible implementation involves the prediction device retrieving historical user data from multiple historical complainants within a given time period from the operator's server. From this data, the device selects the user who filed a complaint only once within that historical period as the target user. The prediction device acquires user data for a first preset time period before the target user filed a complaint, and user data for a second preset time period after the target user filed a complaint. Further, the prediction device performs feature encoding (embedding) on ​​the historical user data from multiple historical complainants to obtain historical user data features. It then performs feature encoding on the user data from the first preset time period before the target user filed a complaint to obtain a first feature, and performs feature encoding on the user data from the second preset time period after the target user filed a complaint to obtain a second feature.

[0047] Historical user data can include user ID, complaint time, complaint event number, services used by the user, user data usage, user signal quality, user network speed, user satisfaction, etc.

[0048] The user data within the first preset time period mentioned above may include the user's data usage, network speed, signal quality, etc., before the target user filed a complaint.

[0049] The user data within the aforementioned second preset time period may include user satisfaction, business processing status, data usage, network speed, signal quality, etc., after a target user files a complaint.

[0050] For example, the predictive device obtains specific historical complaint data from users through the customer service support system, the mobile network geographic information system (GIS), and mobile network complaint reports. Through the capability open platform system, it integrates data from four dimensions: user terminal, network, business, and service. This data is combined with perceived reputation information such as user complaints, appeals, Customer Effort Score (CES) satisfaction, Net Promoter Score (NPS), and Telecom Customer Satisfaction Index (TCSI) surveys to obtain historical user data from multiple users with historical complaints within a historical time period. Utilizing a multimodal learning model integrating images, videos, audio, and semantics, the historical user data is imported and repeatedly trained to generate historical user data features. The prediction device extracts user data within a first preset time period (e.g., within 90 days) prior to the occurrence of a target user complaint, and performs feature encoding on the user data within the first preset time period prior to the occurrence of the target user complaint to obtain a first feature. Extract user data within a second preset time period (e.g., within x days) after the target user's complaint occurs, and encode the user data within this second preset time period to obtain the second feature.

[0051] It should be noted that if a user who has already filed a complaint complains again within the first preset time period after filing a complaint, it constitutes a duplicate complaint. Therefore, this invention serves users who have only filed one complaint, preventing users from filing another complaint within the first preset time period after filing a complaint. To ensure data symmetry, data features from the first preset time period before the user's complaint occurred are selected to obtain the first feature K.

[0052] S202. Determine the degree of correlation between the first feature and the second feature to obtain the first correlation coefficient.

[0053] Among them, the first correlation coefficient is directly proportional to the probability that the target user will make repeated complaints.

[0054] As one possible approach, the prediction device uses the Canonical Correlation Analysis (CCA) algorithm to calculate the correlation between the first feature and the second feature, thereby obtaining the first correlation coefficient.

[0055] It should be noted that the CCA algorithm extracts two representative composite variables from two sets of variables and uses the correlation between these two composite variables to reflect the overall correlation between the two sets of indicators. The autocorrelation function describes the degree of correlation of a random signal at any two different times. The prediction device calculates the autocorrelation coefficient based on the correlation between the first feature and the second feature using the CCA algorithm.

[0056] For example, the prediction device calculates the autocorrelation coefficient ρ by using the CCA algorithm to calculate the correlation between the first feature K and the second feature T. 自 , ρ 自 The correlation coefficient is calculated by performing canonical correlation analysis on data from the 90 days prior to the target user's complaint (including the content of the complaint) and data from the x days after the complaint occurred.

[0057] Where COV(K, T) is the covariance of K and T, and μ k μ t Let K and T be the means, respectively, and D(K) and D(T) be the variances of K and T, respectively. The correlation coefficient ρ takes the value [-1, 1]. 自 For example, ρ 自The closer the absolute value of K is to 1, the higher the linear correlation between K and T; the closer it is to 0, the lower the linear correlation between K and T. The higher the correlation, the greater the likelihood that the target user will file a repeat complaint.

[0058] S203. Determine the correlation between the characteristics of historical user data and the sum of the first and second characteristics to obtain the second correlation coefficient.

[0059] The second correlation coefficient is directly proportional to the probability that the target user will make repeated complaints.

[0060] As one possible approach, the prediction device is based on the CCA algorithm, which calculates the correlation coefficient between historical user data features and the sum of the first and second features.

[0061] It should be noted that the cross-correlation function describes the degree of correlation between two different random signals at any two different times. The prediction device calculates the cross-correlation coefficient based on the CCA algorithm to determine the correlation between the historical user data features and the sum of the first and second features.

[0062] For example, the prediction device, based on the CCA algorithm, calculates the correlation coefficient ρ between historical user data feature M and the sum of the first feature K and the second feature T. 互 , ρ 互 The correlation coefficient is calculated by performing canonical correlation analysis on data from the 90+x days before and after the target user's complaint (including the content of the complaint) and historical user data characteristics.

[0063] S204. Based on the first correlation coefficient and the second correlation coefficient, determine the target user's repeated complaint behavior.

[0064] As one possible approach, if the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to a preset deviation, the prediction device performs weighted processing on the first correlation coefficient and the second correlation coefficient to obtain a weighted result, and determines the target user's repeated complaint behavior based on the weighted result.

[0065] In one design, to predict repeat complaint behavior of target users, such as... Figure 3 As shown, S204 may specifically include the following S2041-S2044.

[0066] S2041. Determine the deviation between the first correlation coefficient and the second correlation coefficient.

[0067] As one possible approach, the prediction device takes the absolute value of the difference between the first correlation coefficient and the second correlation coefficient, i.e., |ρ 自 -ρ 互| represents the deviation between the first and second correlation coefficients.

[0068] S2042. Determine whether this deviation is less than or equal to the preset deviation.

[0069] S2043. When the deviation is less than or equal to the preset deviation, the first correlation coefficient and the second correlation coefficient are weighted to obtain the weighted result, and the repeated complaint behavior of the target user is determined based on the weighted result.

[0070] As one possible approach, when the deviation is less than or equal to a preset deviation, the prediction device obtains a first weight for a first correlation coefficient and a second weight for a second correlation coefficient. It then weights the first correlation coefficient according to the first weight to obtain a first weighted correlation coefficient, and weights the second correlation coefficient according to the second weight to obtain a second weighted correlation coefficient. The sum of the first and second weighted correlation coefficients is determined as the weighted result. If the weighted result is greater than a preset threshold, the prediction device can determine that the target user is highly likely to file a repeat complaint; if the weighted result is less than or equal to the preset threshold, it can determine that the target user is unlikely to file a repeat complaint.

[0071] For example, such as Figure 4 As shown, the prediction device takes a preset deviation of 50%, when ρ 自 With ρ 互 When the deviation does not exceed 50%, i.e., |ρ 自 -ρ 互 When |≤0.5, it indicates that the data is accurate and has reference value; in this case, ρ is obtained. 自 The weights a and ρ 互 The weight b, where If we take a = b = 0.5, then the first weighted correlation coefficient is aρ. 自 =0.5ρ 自 The second weighted correlation coefficient is bρ 互 =0.5ρ 互 The weighted result ρ 总 =aρ 自 +bρ 互 =0.5ρ 自 +0.5ρ 互 Preset threshold ρ th The value is not specifically limited and can be set according to the actual needs of the scenario. Here, we take ρ. th =0.7, when ρ 总 >ρ th That is, 0.5ρ 自 +0.5ρ 互 When ρ > 0.7, it indicates that the target user is highly likely to file a repeat complaint. 总 ≤ρ thThat is, 0.5ρ 自 +0.5ρ 互 A value of ≤0.7 indicates that the target user is unlikely to file a repeat complaint.

[0072] S2044. If the deviation is greater than the preset deviation, adjust the first correlation coefficient and the second correlation coefficient until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and determine the target user's repeated complaint behavior based on the adjusted first correlation coefficient and the second correlation coefficient.

[0073] As one possible approach, when the deviation is greater than a preset deviation, the prediction device acquires the third feature of the target user, replaces the first feature with the third feature, and adjusts the first correlation coefficient and the second correlation coefficient according to the replaced first feature until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation. The repeated complaint behavior of the target user is then determined based on the adjusted first correlation coefficient and the second correlation coefficient.

[0074] The third feature is used to reflect the user data characteristics within a third preset time period before the target user makes a complaint. The third preset time period is shorter than the first preset time period.

[0075] For example, such as Figure 4 As shown, when ρ 自 With ρ 互 The deviation exceeds 50%, i.e., |ρ 自 -ρ 互 When |>0.5, it indicates that the predictive device's extraction of target user data features is not critical enough or representative enough, and the target user is likely to file a repeat complaint. In this case, the predictive device acquires user data from the y days prior to the target user's complaint (y<90) (this could include the user's data usage, network speed, signal quality, etc. within the y days before the complaint), performs feature encoding on this user data to obtain the third feature, and continuously reduces the y value until |ρ 自 -ρ 互 |≤0.5, at this time ρ 自 With ρ 互 The first feature K in the equation is replaced with the third feature to obtain a new ρ. 自 With ρ 互 At this point, for the new ρ 自 With ρ 互 After assigning weights a = 0.5 and b = 0.5 respectively, the sum is used to obtain the new ρ. 总 , will the new ρ 总 The system compares the data with a preset threshold to determine whether the target user will file a repeat complaint.

[0076] Understandable, such as Figure 5As shown, the prediction device acquires historical user data features M from multiple historical complainants, as well as data features K from the 90 days before the complaint, data features T from x days after the complaint, and data features K+T from (90+x) days before and after the complaint for the target user. Based on the correlation between K and T, ρ is obtained. 自 ρ is obtained based on the correlation between M and K+T. 互 The prediction device determines ρ 自 With ρ 互 Does the deviation exceed 50%? If the deviation does not exceed 50%, i.e., |ρ 自 -ρ 互 |≤0.5 when entering ρ 总 The comparison step with the preset threshold, if ρ 总 (ρ 总 =0.5ρ 自 +0.5ρ 互 ) greater than the preset threshold (take ρ) th If the deviation is less than or equal to 0.7, it is determined that the target user is highly likely to file another complaint; otherwise, it is determined that the target user is unlikely to file a repeat complaint. When the deviation exceeds 50%, the process of shortening the time and re-retrieval begins, which involves continuously shortening the 90-day period in the data feature K obtained from the 90 days prior to the target user's complaint until |ρ| is satisfied. 自 -ρ 互 If |≤0.5, then after obtaining the new data feature K, we obtain the new ρ. 自 With ρ 互 Then, from the judgment ρ 自 With ρ 互 If the deviation exceeds 50%, the process restarts until it is determined whether the target user will complain again.

[0077] The technical solution provided in this application provides at least the following beneficial effects: The prediction device acquires historical user data features of multiple historical complaining users and a first feature and a second feature of the target user. The target user is any one of the multiple historical complaining users, and a historical complaining user is a user who has complained once within a historical time period. The first feature reflects the user data features within a first preset time period before the target user made a complaint, and the second feature reflects the user data features within a second preset time period after the target user made a complaint. Further, the prediction device determines the correlation between the first feature and the second feature to obtain a first correlation coefficient, and determines the correlation between the historical user data features and the sum of the first and second features to obtain a second correlation coefficient. This is equivalent to matching and calculating the complaint data before and after the repeated user complaint with the target user's behavioral habits. Since both the first and second correlation coefficients are proportional to the probability of the target user making repeated complaints, the larger the first and second correlation coefficients are, the greater the likelihood of the target user making repeated complaints. In related technologies, operators only determine which users are repeat complaining users after the repeated complaint behavior occurs, at which point the operator's reputation has already been affected. This application utilizes the characteristics of the first and second correlation coefficients to predict the repeat complaint behavior of target users, thereby enabling operators to take measures to prevent users from making repeat complaints, effectively resolve potential repeat complaint users, and thus improve the operator's reputation.

[0078] The above embodiments mainly describe the solutions provided by the embodiments of this application from the perspective of an apparatus (device). It is understood that, in order to implement the above methods, the apparatus or device includes hardware structures and / or software modules corresponding to the execution of each method flow. These hardware structures and / or software modules corresponding to the execution of each method flow can constitute a material information determination apparatus. Those skilled in the art should readily recognize that, in conjunction with the algorithm steps of the various examples described in the embodiments of the invention herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software-driven hardware manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0079] This application embodiment can divide the device or equipment into functional modules according to the above method examples. For example, the device or equipment can be divided into functional modules corresponding to each function, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0080] Figure 6 This is a schematic diagram illustrating the structure of a prediction device according to an exemplary embodiment. (Refer to...) Figure 6 As shown, the prediction device 30 provided in this application embodiment includes an acquisition unit 301 and a determination unit 302.

[0081] The acquisition unit 301 is used to acquire historical user data features of multiple historical complaining users and the first and second features of the target user; the target user is any one of the multiple historical complaining users; a historical complaining user is a user who has complained once within a historical time period; the first feature is used to reflect the user data features within a first preset time period before the target user made a complaint; the second feature is used to reflect the user data features within a second preset time period after the target user made a complaint; the determination unit 302 is used to determine the correlation between the first and second features to obtain a first correlation coefficient; determine the correlation between the historical user data features and the sum of the first and second features to obtain a second correlation coefficient; and determine the repeated complaint behavior of the target user based on the first and second correlation coefficients; the first correlation coefficient is proportional to the probability of the target user making repeated complaint behavior; the second correlation coefficient is also proportional to the probability of the target user making repeated complaint behavior.

[0082] Optionally, the determining unit 302 is specifically used for: determining the deviation between the first correlation coefficient and the second correlation coefficient; when the deviation is less than or equal to a preset deviation, weighting the first correlation coefficient and the second correlation coefficient to obtain a weighted result, and determining the target user's repeated complaint behavior based on the weighted result; when the deviation is greater than the preset deviation, adjusting the first correlation coefficient and the second correlation coefficient until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and determining the target user's repeated complaint behavior based on the adjusted first correlation coefficient and the second correlation coefficient.

[0083] Optionally, the determining unit 302 is specifically used to: obtain the first weight of the first correlation coefficient and the second weight of the second correlation coefficient; weight the first correlation coefficient according to the first weight to obtain the first weighted correlation coefficient, and weight the second correlation coefficient according to the second weight to obtain the second weighted correlation coefficient; and determine the sum of the first weighted correlation coefficient and the second weighted correlation coefficient as the weighted result.

[0084] Optionally, the determining unit 302 is specifically used to: determine that the target user will file a repeated complaint if the weighted result is greater than a preset threshold; and determine that the target user will not file a repeated complaint if the weighted result is less than or equal to the preset threshold.

[0085] Optionally, the determining unit 302 is specifically used to: obtain the third feature of the target user; the third feature is used to reflect the user data characteristics within a third preset time period before the target user makes a complaint; the third preset time period is shorter than the first preset time period; replace the first feature with the third feature, and adjust the first correlation coefficient and the second correlation coefficient according to the replaced first feature.

[0086] Figure 7 This is a schematic diagram of the structure of an electronic device provided in this application. For example... Figure 7 The electronic device 40 may include at least one processor 401 and a memory 402 for storing processor-executable instructions, wherein the processor 401 is configured to execute the instructions in the memory 402 to implement the recurring complaint behavior prediction method in the above embodiments.

[0087] In addition, the electronic device 40 may also include a communication bus 403 and at least one communication interface 404.

[0088] Processor 401 may be a processor (central processing unit, CPU), microprocessor unit, ASIC, or one or more integrated circuits for controlling the execution of programs according to the present application.

[0089] The communication bus 403 may include a path for transmitting information between the aforementioned components.

[0090] Communication interface 404 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0091] The memory 402 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory may exist independently and be connected to the processor 401 via a bus. The memory may also be integrated with the processor 401.

[0092] The memory 402 stores instructions for executing the scheme of this application, and the processor 401 controls the execution. The processor 401 executes the instructions stored in the memory 402 to realize the functions of the method of this application.

[0093] As an example, combined Figure 6 The functions implemented by the acquisition unit 301 and the determination unit 302 in the prediction device 30 are the same as those of the acquisition unit 301 and the determination unit 302. Figure 7 The processor 401 in it has the same function.

[0094] In a specific implementation, as one example, processor 401 may include one or more CPUs, for example... Figure 7 CPU0 and CPU1 in the CPU.

[0095] In a specific implementation, as one example, the electronic device 40 may include multiple processors, such as... Figure 7 Processors 401 and 407 are described herein. Each of these processors may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. A processor here may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0096] In a specific implementation, as one embodiment, the electronic device 40 may further include an output device 405 and an input device 406. The output device 405 communicates with the processor 401 and can display information in various ways. For example, the output device 405 may be a liquid crystal display (LCD), a light-emitting diode (LED) display device, a cathode ray tube (CRT) display device, or a projector, etc. The input device 406 communicates with the processor 401 and can accept input from user objects in various ways. For example, the input device 406 may be a mouse, keyboard, touchscreen device, or sensing device, etc.

[0097] Those skilled in the art will understand that Figure 7 The structure shown does not constitute a limitation on the electronic device 40, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0098] In addition, this application also provides a computer-readable storage medium that, when the instructions in the computer-readable storage medium are executed by the processor of an electronic device, enables the electronic device to perform the recurring complaint behavior prediction method provided in the above embodiments.

[0099] In addition, this application also provides a computer program product, including computer instructions, which, when executed on an electronic device, cause the electronic device to perform the recurring complaint behavior prediction method provided in the above embodiments.

[0100] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention described herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not invented herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

Claims

1. A method for predicting repeat complaint behavior, characterized in that, The method includes: The system acquires historical user data features of multiple historical complainants and first and second features of a target user; the target user is any one of the multiple historical complainants; a historical complainant is a user who has filed a complaint once within a historical time period; the first feature reflects the user data features of the target user within a first preset time period before the complaint was filed; the second feature reflects the user data features of the target user within a second preset time period after the complaint was filed. The correlation between the first feature and the second feature is determined to obtain a first correlation coefficient; the first correlation coefficient is proportional to the probability that the target user will make repeated complaint behavior. The correlation between the historical user data features and the sum of the first feature and the second feature is determined to obtain a second correlation coefficient; the second correlation coefficient is proportional to the probability that the target user will make repeated complaint behavior. Based on the first correlation coefficient and the second correlation coefficient, the repeated complaint behavior of the target user is determined.

2. The method according to claim 1, characterized in that, The step of determining the target user's repeated complaint behavior based on the first correlation coefficient and the second correlation coefficient includes: Determine the deviation between the first correlation coefficient and the second correlation coefficient; If the deviation is less than or equal to a preset deviation, the first correlation coefficient and the second correlation coefficient are weighted to obtain a weighted result, and the repeated complaint behavior of the target user is determined based on the weighted result. If the deviation is greater than the preset deviation, the first correlation coefficient and the second correlation coefficient are adjusted until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and the repeated complaint behavior of the target user is determined based on the adjusted first correlation coefficient and the second correlation coefficient.

3. The method according to claim 2, characterized in that, The weighting process for the first correlation coefficient and the second correlation coefficient to obtain the weighted result includes: Obtain the first weight of the first correlation coefficient and the second weight of the second correlation coefficient; The first correlation coefficient is weighted according to the first weight to obtain a first weighted correlation coefficient, and the second correlation coefficient is weighted according to the second weight to obtain a second weighted correlation coefficient. The sum of the first weighted correlation coefficient and the second weighted correlation coefficient is determined as the weighted result.

4. The method according to claim 3, characterized in that, The step of determining the target user's repeated complaint behavior based on the weighted result includes: If the weighted result is greater than a preset threshold, it is determined that the target user will file repeated complaints. If the weighted result is less than or equal to a preset threshold, it is determined that the target user will not file a repeat complaint.

5. The method according to claim 2, characterized in that, The adjustment of the first correlation coefficient and the second correlation coefficient includes: A third feature of the target user is obtained; the third feature is used to reflect the user data characteristics within a third preset time period before the target user files a complaint; the third preset time period is shorter than the first preset time period. The first feature is replaced with the third feature, and the first correlation coefficient and the second correlation coefficient are adjusted according to the replaced first feature.

6. A device for predicting repeat complaint behavior, characterized in that, The prediction device includes an acquisition unit and a determination unit; The acquisition unit is used to acquire historical user data features of multiple historical complaint users and a first feature and a second feature of a target user; the target user is any one of the multiple historical complaint users; a historical complaint user is a user who has filed a complaint once within a historical time period; the first feature is used to reflect the user data features of the target user within a first preset time period before the complaint was filed; the second feature is used to reflect the user data features of the target user within a second preset time period after the complaint was filed. The determining unit is used to determine the degree of correlation between the first feature and the second feature to obtain a first correlation coefficient; the first correlation coefficient is proportional to the probability that the target user will make repeated complaint behavior; The determining unit is further configured to determine the degree of correlation between the historical user data features and the sum of the first feature and the second feature, and obtain a second correlation coefficient; the second correlation coefficient is proportional to the probability that the target user will generate repeated complaint behavior; The determining unit is further configured to determine the repeated complaint behavior of the target user based on the first correlation coefficient and the second correlation coefficient.

7. The prediction device according to claim 6, characterized in that, The determining unit is specifically configured to: determine the deviation between the first correlation coefficient and the second correlation coefficient; if the deviation is less than or equal to a preset deviation, perform weighted processing on the first correlation coefficient and the second correlation coefficient to obtain a weighted result, and determine the repeated complaint behavior of the target user based on the weighted result; if the deviation is greater than the preset deviation, adjust the first correlation coefficient and the second correlation coefficient until the deviation between the first correlation coefficient and the second correlation coefficient is less than or equal to the preset deviation, and determine the repeated complaint behavior of the target user based on the adjusted first correlation coefficient and the second correlation coefficient.

8. The prediction device according to claim 7, characterized in that, The determining unit is specifically configured to: obtain a first weight of the first correlation coefficient and a second weight of the second correlation coefficient; weight the first correlation coefficient according to the first weight to obtain a first weighted correlation coefficient, and weight the second correlation coefficient according to the second weight to obtain a second weighted correlation coefficient; and determine the sum of the first weighted correlation coefficient and the second weighted correlation coefficient as the weighted result.

9. The prediction device according to claim 8, characterized in that, The determining unit is specifically used to: determine that the target user will file a repeat complaint if the weighted result is greater than a preset threshold; and determine that the target user will not file a repeat complaint if the weighted result is less than or equal to the preset threshold.

10. The prediction device according to claim 7, characterized in that, The determining unit is specifically used to: obtain a third feature of the target user; the third feature is used to reflect the user data characteristics within a third preset time period before the target user makes a complaint; the third preset time period is shorter than the first preset time period; replace the first feature with the third feature, and adjust the first correlation coefficient and the second correlation coefficient according to the replaced first feature.

11. An electronic device, characterized in that, include: A processor and a memory for storing instructions executable by the processor; wherein the processor is configured to execute instructions to implement the method for predicting recurring complaint behavior as described in any one of claims 1-5.

12. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is enabled to perform the method for predicting recurring complaint behavior as described in any one of claims 1-5.

Images