Quality difference user determination method, apparatus, device, medium, and product

Abstract

Embodiments of the present application provide a poor quality user determination method, device, equipment, medium and product, comprising: obtaining first network quality data of a plurality of first users in a target area within a first preset time period; clustering the plurality of first network quality data based on a preset quality difference reason and a preset clustering radius to obtain a plurality of second network quality data; determining the similarity of each two second network quality data in the plurality of second network quality data and a correction coefficient for adjusting the similarity according to the plurality of second network quality data; adjusting the similarity of each two second network quality data in the plurality of second network quality data based on the correction coefficient to obtain a target score for evaluating user network quality; and determining a first user as a poor quality user in a case where a user score of the first user is less than the target score. Embodiments of the present application can improve the accuracy of determining poor quality users.

Description

Technical Field

[0001] This application belongs to the field of broadband user quality evaluation technology, and in particular relates to a method, apparatus, equipment, medium and product for identifying poor quality users. Background Technology

[0002] With the increasing prevalence of broadband services and the surge in broadband users, maintaining network quality has become particularly important. Therefore, assessing the quality level of broadband users is crucial for accurately identifying potential low-quality users, facilitating early repairs, and improving user experience. However, current technologies for identifying low-quality users have relatively low accuracy. Summary of the Invention

[0003] This application provides a method, apparatus, device, medium, and product for identifying poor-quality users, which can improve the accuracy of identifying poor-quality users.

[0004] In a first aspect, embodiments of this application provide a method for determining users with poor network quality, the method comprising: acquiring first network quality data of multiple first users within a target area within a first preset time period;

[0005] Based on the preset reasons for poor quality and the preset clustering radius, multiple first network quality data are clustered to obtain multiple second network quality data, and the number of first network quality data is greater than that of second network quality data.

[0006] Based on multiple second network quality data, determine the similarity between every two second network quality data, and the correction coefficient used to adjust the similarity;

[0007] Based on the correction coefficient, the similarity between every two second network quality data in multiple second network quality data is adjusted to obtain the target score for evaluating the user's network quality;

[0008] If the user rating of the first user is lower than the target rating, the first user is identified as a poor-quality user, and the user rating is set based on the first user's first network quality data and the preset rating rules.

[0009] In an optional implementation of the first aspect, based on a preset cause of poor quality and a preset clustering radius, multiple first network quality data are clustered to obtain multiple second network quality data, including:

[0010] Based on multiple first network quality data, the user distance between the multiple first network quality data is calculated, so as to determine multiple third network quality data from the multiple first network quality data according to the user distance. The number of third network quality data is greater than the number of second network quality data. The user distance is characterized as the correlation between the multiple first network quality data.

[0011] Based on the preset reasons for poor quality and the preset clustering radius, multiple third network quality data are clustered to determine the network quality data whose user distance from the cluster center point is less than or equal to the preset clustering radius. These are then designated as second network quality data, thus obtaining multiple second network quality data.

[0012] In an optional implementation of the first aspect, calculating user distances between the multiple first network quality data based on multiple first network quality data, to determine multiple third network quality data from the multiple first network quality data based on the user distances, includes:

[0013] Divide the N first network quality data into M first quality data and NM second quality data;

[0014] Calculate the user distances between the Mth first quality data point and the NM second quality data points to obtain the NM Kth user distances;

[0015] Determine the second quality data corresponding to the largest user distance among the NM distances to the Kth user, and use it as the (M+1)th first quality data, thus obtaining the M+1 first quality data and NM-1 second quality data;

[0016] Calculate the user distances between the (M+1)th first quality data point and the (NM-1)th second quality data points to obtain the (NM-1)th (K+1)th user distance;

[0017] The second quality data corresponding to the distance to the target user is determined as the M+2th first quality data, and M+2 first quality data and NM-2 second quality data are obtained to obtain a preset number of first quality data. Among them, the target user distance is the smallest user distance among NM-1 Kth user distances and the largest user distance among the smallest user distances among NM-1 K+1 user distances.

[0018] Based on a preset number of first quality data, multiple third network quality data are determined.

[0019] The number of first network quality data is N, where N is a positive integer greater than 3, M = K = 1, and both the first quality data and the second quality data are first network quality data.

[0020] In one optional implementation of the first aspect, a plurality of third network quality data are determined based on a preset number of first quality data, including:

[0021] Using each first quality data point in a preset quantity of first quality data as the center and a preset distance as the radius, a first region is divided to obtain multiple first regions. Each of the multiple first regions includes multiple first quality data points.

[0022] Determine a first preset quantity of first quality data for each of the multiple first regions, which constitutes multiple third network quality data.

[0023] In an optional implementation of the first aspect, determining the similarity between every two second network quality data points based on a plurality of second network quality data points, and a correction coefficient for adjusting the similarity, includes:

[0024] Similarity calculations are performed on the multiple dimensions of the second network quality data to obtain multiple first similarities with each dimension of the data.

[0025] Based on the first preset weight, the first similarity corresponding to each dimension data in the multiple dimensions data is weighted and summed to obtain the similarity between each pair of second network quality data in the multiple second network quality data.

[0026] Based on multiple service quality data included in multiple second network quality data sets, a correction coefficient is calculated to adjust the similarity.

[0027] In an optional implementation of the first aspect, similarity calculations are performed on the multiple dimensions of the second network quality data to obtain multiple first similarities corresponding to each dimension of the multiple dimensions, including:

[0028] The similarity of each dimension of the second network quality data is calculated separately to obtain multiple second similarities for each dimension of the data.

[0029] In the multi-dimensional data, the first similarity is determined by identifying the first similarity among the multiple second similarities corresponding to each dimension data that is greater than the preset similarity threshold corresponding to each dimension data. This results in multiple first similarities corresponding to each dimension data in the multi-dimensional data.

[0030] In an optional implementation of the first aspect, the similarity between every two second network quality data points in a plurality of second network quality data is adjusted based on a correction coefficient to obtain a target score for evaluating user network quality, including:

[0031] The similarity between each pair of second network quality data is adjusted by multiplying the correction coefficient with the similarity between each pair of second network quality data in multiple second network quality data, resulting in multiple adjusted similarities.

[0032] If the target similarity group meets the preset conditions, multiple fourth network quality data are determined based on the target similarity group so that the target similarity group does not meet the preset conditions. The target similarity group includes two similarities determined sequentially from multiple adjusted similarities in a preset order.

[0033] Calculate the variance of the network quality data for each dimension based on the network quality data included in each of the multiple fourth network quality data sets.

[0034] The target score is determined by multiplying the variance by the second preset weight.

[0035] In one optional implementation of the first aspect, the preset condition includes that the difference between the reciprocals of the two similarity groups included in the target similarity group is not greater than a preset threshold.

[0036] Secondly, embodiments of this application provide a device for determining users with poor quality, the device comprising:

[0037] The acquisition module is used to acquire the first network quality data of multiple first users within a first preset time period in the target area;

[0038] The clustering module is used to cluster multiple first network quality data based on preset reasons for poor quality and preset clustering radius to obtain multiple second network quality data, wherein the number of first network quality data is greater than the number of second network quality data.

[0039] The first determining module is used to determine the similarity between every two second network quality data in the multiple second network quality data, and to determine the correction coefficient for adjusting the similarity based on the multiple second network quality data.

[0040] The adjustment module is used to adjust the similarity of every two second network quality data in multiple second network quality data based on the correction coefficient, so as to obtain a target score for evaluating the user's network quality.

[0041] The second determining module is used to determine the first user as a poor-quality user if the user rating of the first user is less than the target rating. The user rating is set based on the first user's first network quality data and preset rating rules.

[0042] Thirdly, an electronic device is provided, comprising: a memory for storing computer program instructions; and a processor for reading and executing the computer program instructions stored in the memory to perform a poor quality user determination method provided by any optional embodiment of the first and second aspects.

[0043] Fourthly, a computer storage medium is provided, on which computer program instructions are stored, wherein when the computer program instructions are executed by a processor, the poor quality user determination method provided by any optional implementation of the first and second aspects is implemented.

[0044] Fifthly, a computer program product is provided, wherein instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform a poor quality user determination method provided by any optional implementation of the first and second aspects.

[0045] In this embodiment, by acquiring first network quality data of multiple first users within a target area over a first preset time period, and then clustering these first network quality data based on preset reasons for poor quality and preset clustering radii, multiple second network quality data are obtained. Based on these second network quality data, the similarity between any two second network quality data points and a correction coefficient for adjusting the similarity can be determined. Furthermore, based on the correction coefficient, the similarity between any two second network quality data points can be adjusted to obtain a target score for evaluating user network quality. This allows a first user to be identified as a user with poor quality if their score, based on their first network quality data and a preset scoring rule, is less than the target score. Thus, by clustering the acquired first network quality data and calculating correction coefficients to adjust the similarity between any two second network quality data points obtained from the clustering process, the problem of uneven distribution of user network quality data is eliminated, thereby improving the accuracy of identifying users with poor quality. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating a method for determining poor-quality users provided in an embodiment of this application;

[0048] Figure 2 This is a flowchart illustrating another method for determining poor-quality users provided in an embodiment of this application;

[0049] Figure 3 This is a flowchart illustrating another method for determining poor-quality users provided in an embodiment of this application;

[0050] Figure 4 This is a flowchart illustrating another method for determining poor-quality users provided in an embodiment of this application;

[0051] Figure 5 This is a schematic diagram of the structure of a poor quality user determination device provided in an embodiment of this application;

[0052] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0053] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0054] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0055] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0056] With the increasing popularity of broadband services and the surge in broadband users, maintaining user network quality has become particularly important. Therefore, it is crucial to assess user network quality and identify users with poor network performance for proactive repairs.

[0057] However, current technologies typically determine whether a user is a poor-performing user by collecting a single metric. But a single metric is insufficient to accurately represent a user's true perception, leading to low accuracy in identifying poor-performing users. Alternatively, objective evaluation of historical user data can be used, but this objective evaluation significantly impacts the identification of poor-performing users, making accurate identification difficult.

[0058] In summary, to address the issue of low accuracy in identifying poor-quality users in existing technologies, this application provides a method, apparatus, device, medium, and product for identifying poor-quality users. This method acquires first network quality data of multiple first users within a target area over a first preset time period. Then, based on preset reasons for poor quality and a preset clustering radius, the multiple first network quality data are clustered to obtain multiple second network quality data. Based on these second network quality data, the similarity between any two second network quality data points and a correction coefficient for adjusting the similarity can be determined. Furthermore, based on the correction coefficient, the similarity between any two second network quality data points can be adjusted to obtain a target score for evaluating user network quality. This allows the first user to be identified as a poor-quality user if the user score based on the first user's first network quality data and a preset scoring rule is less than the target score. In this way, by clustering the acquired first network quality data and calculating the correction coefficient, the similarity between every two second network quality data in the clustered second network quality data can be adjusted, thereby eliminating the problem of uneven distribution of user network quality data and improving the accuracy of identifying poor quality users.

[0059] The method for determining poor-quality users provided in this application can be executed by a device for determining poor-quality users, or by a control module within that device for executing the method. This application uses the execution of the method by a device for determining poor-quality users as an example to illustrate the method provided in this application.

[0060] The method for determining poor-quality users provided in this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] Figure 1 This is a flowchart illustrating a method for determining poor-quality users provided in an embodiment of this application.

[0062] like Figure 1 As shown, the execution device for this method of determining poor-quality users is a device for determining poor-quality users, which may specifically include the following steps:

[0063] S110, acquire the first network quality data of multiple first users in the target area within a first preset time period.

[0064] The device for identifying poor-quality users can acquire first network quality data of multiple first users within a first preset time period in a scenario where the network quality of broadband users is being evaluated to identify poor-quality users, so as to obtain multiple first network quality data.

[0065] The target area can be a pre-defined area based on actual needs, such as an area divided by residential community or an area managed by the operator's server; no specific limitation is made here. The first user can be a broadband user within the target area. The first preset time period can be a pre-set time period based on actual needs or experience, such as one week or one month; no specific limitation is made here. The first network quality data can include at least two of the following: home network data, bearer network data, service quality data, and service quality data.

[0066] It should be noted that home network data may include at least one of the following indicators: gateway continuous runtime, percentage of high-performance periods for the Central Processing Unit (CPU), percentage of high-performance periods for memory, user offline frequency, optical network unit (ONU) optical power, gateway port speed, downstream device status, strong wireless-fidelity (WiFi) interference, and weak wireless-fidelity (WiFi) coverage. Bearer network data may include at least one of the following indicators: peak bandwidth utilization of the optical line terminal (OLT) uplink, peak bandwidth utilization of the Passive Optical Network (PON) port, frequent short-term online / offline cycles, and service unavailability duration. Service quality data may include Transmission Control Protocol (TCP) latency exceeding limits rate, packet loss exceeding limits rate, percentage of stuttering time, TV playback success rate, Electronic Program Guide (EPG) response latency compliance rate, MOS value, wireless access quality, and Central Processing Unit (CPU) performance. Unit / Processor (CPU) memory usage high or abnormal disconnection frequency, and service quality data may include at least one of the following indicators: repeated complaints, abnormal complaints, complaint satisfaction, and installation satisfaction.

[0067] S120, based on the preset reasons for poor quality and the preset clustering radius, cluster multiple first network quality data to obtain multiple second network quality data.

[0068] The preset reasons for poor network quality can be based on the various dimensions of the actual network quality data and pre-defined causes that may lead to poor network quality for broadband users. For example, it could be the frequency of TCP latency exceeding a preset value. The preset clustering radius can be a clustering radius pre-set based on actual conditions or empirical values. The quantity of the first network quality data is greater than that of the second network quality data. The data in each dimension of the second network quality data is consistent with the data in each dimension of the first network quality data. This is not specifically limited here.

[0069] Specifically, after obtaining multiple sets of first network quality data, the device can cluster these sets of first network quality data based on preset reasons for poor quality and preset clustering radii. This clustering process identifies network quality data sets that satisfy the preset clustering radii as second network quality data, thus generating multiple sets of second network quality data. In this way, clustering multiple sets of first network quality data eliminates uneven distribution of user network quality data, improving the accuracy of identifying users with poor quality.

[0070] S130, Based on multiple second network quality data, determine the similarity between every two second network quality data, and a correction coefficient for adjusting the similarity.

[0071] The device for identifying poor-quality users can, after obtaining multiple sets of second network quality data, determine the similarity between every two sets of second network quality data to obtain multiple similarity scores. Furthermore, it can determine correction coefficients for adjusting the similarity scores based on the obtained second network quality data. This allows for subsequent adjustments to the similarity between every two sets of second network quality data based on the correction coefficients, thereby improving the accuracy of identifying poor-quality users.

[0072] Similarity characterizes the correlation between any two second-network quality data points; the higher the similarity, the stronger the correlation, and vice versa. Further details are omitted here. The correction coefficient can be used to adjust the similarity between any two second-network quality data points across multiple datasets.

[0073] S140, based on the correction coefficient, adjusts the similarity between every two second network quality data in multiple second network quality data to obtain a target score for evaluating user network quality.

[0074] The poor-quality user identification device can adjust the similarity between any two second network quality data points based on a calculated correction coefficient to obtain a target score for evaluating user network quality. Thus, by determining the target score for evaluating user network quality, it is possible to subsequently determine whether a broadband user is a poor-quality user based on the target score. The target score can be a standard value used to evaluate the network quality of a broadband user.

[0075] S150: If the user rating of the first user is lower than the target rating, the first user is identified as a poor-quality user.

[0076] The user rating can be based on the first user's initial network quality data and preset rating rules. The preset rating rules can be pre-set based on actual conditions or experience values ​​to determine the user rating.

[0077] The device for determining poor-quality users can determine whether the user score of the first user is less than the target score after setting the user score of the first user based on the first network quality data of the first user and the preset scoring rules. If the user score of the first user is less than the target score, the device can determine the first user as a poor-quality user.

[0078] In this embodiment, by acquiring first network quality data of multiple first users within a target area over a first preset time period, and then clustering these first network quality data based on preset reasons for poor quality and preset clustering radii, multiple second network quality data are obtained. Based on these second network quality data, the similarity between any two second network quality data points and a correction coefficient for adjusting the similarity can be determined. Furthermore, based on the correction coefficient, the similarity between any two second network quality data points can be adjusted to obtain a target score for evaluating user network quality. This allows a first user to be identified as a user with poor quality if their score, based on their first network quality data and a preset scoring rule, is less than the target score. Thus, by clustering the acquired first network quality data and calculating correction coefficients to adjust the similarity between any two second network quality data points obtained from the clustering process, the problem of uneven distribution of user network quality data is eliminated, thereby improving the accuracy of identifying users with poor quality.

[0079] In one embodiment, to eliminate the problem of uneven distribution of user network quality data, multiple second network quality data can be accurately acquired so that users with poor network quality can be accurately identified subsequently. Figure 2 As shown, the S120 mentioned above can specifically include the following steps:

[0080] S210, based on multiple first network quality data, calculate the user distance between the multiple first network quality data, so as to determine multiple third network quality data from the multiple first network quality data based on the user distance.

[0081] The amount of third network quality data is greater than that of second network quality data. The data in each dimension of the third network quality data is consistent with that in the first network quality data, and will not be elaborated further here. User distance is represented by the correlation between multiple first network quality data; the greater the user distance, the smaller the correlation between the first network quality data, and vice versa.

[0082] The device for determining poor network quality can determine multiple third network quality data from multiple first network quality data by calculating the user distance between the multiple first network quality data based on the user distance.

[0083] S220, based on the preset reasons for poor quality and the preset clustering radius, by performing clustering processing on multiple third network quality data, the network quality data whose user distance from the cluster center point is less than or equal to the preset clustering radius is determined as second network quality data, so as to obtain multiple second network quality data.

[0084] Here, the cluster center point can be the center point of multiple first network quality data after clustering. For example, the cluster center point can be a certain indicator data in a certain first network quality data, without specific limitations.

[0085] Specifically, after determining multiple third network quality data, the device can perform clustering processing on the multiple third network quality data based on preset reasons for poor quality and preset clustering radius to obtain multiple clustering results. Each clustering result can include multiple third network quality data and the user distance between each of the multiple third network quality data and the cluster center point. Then, based on the user distance between the multiple third network quality data and the cluster center point in each clustering result, the device can determine the network quality data in each clustering result whose user distance between the multiple third network quality data and the cluster center point is less than or equal to the preset clustering radius, and designate them as second network quality data, so as to obtain multiple second network quality data.

[0086] In one example, during the process of clustering multiple third-party network quality data based on preset reasons for poor quality and preset clustering radii, it is assumed that a certain clustering result is obtained with X as the clustering factor. i (k) represents the clustering process performed on the cluster centers, where X i (k) represents the k-th metric in the i-th third network quality data, dist(X) i (k), X j (k) represents the user distance between the k index of the i-th network quality data and the k index of the j-th third network quality data. If the user distance is less than the preset clustering radius, the third network quality data corresponding to the distance can be determined as the second network quality data.

[0087] In this way, multiple third network quality data can be determined from multiple first network quality data based on the user distance between multiple first network quality data, and the multiple third network quality data can be clustered to eliminate the problem of uneven user network quality data, thereby facilitating the subsequent improvement of the accuracy of identifying users with poor quality.

[0088] In one embodiment, in order to accurately determine multiple third network quality data, such as Figure 2 As shown, the steps described above, which involve calculating the user distance between multiple first network quality data points to determine multiple third network quality data points from the multiple first network quality data points based on the user distance, can specifically include the following steps:

[0089] Divide the N first network quality data into M first quality data and NM second quality data;

[0090] Calculate the user distances between the Mth first quality data point and the NM second quality data points to obtain the NM Kth user distances;

[0091] Determine the second quality data corresponding to the largest user distance among the NM distances to the Kth user, and use it as the (M+1)th first quality data, thus obtaining the M+1 first quality data and NM-1 second quality data;

[0092] Calculate the user distances between the (M+1)th first quality data point and the (NM-1)th second quality data points to obtain the (NM-1)th (K+1)th user distance;

[0093] The second quality data corresponding to the distance to the target user is determined as the M+2th first quality data, and M+2 first quality data and NM-2 second quality data are obtained to obtain a preset number of first quality data;

[0094] Based on a preset number of first quality data, multiple third network quality data are determined.

[0095] The first network quality data consists of N data points, where N is a positive integer greater than 3, and M = K = 1. Both the first and second quality data points are considered part of the first network quality data set. The target user distance is the minimum user distance among NM-1 Kth user distances and the maximum user distance among the minimum user distances among NM-1 K+1th user distances. The preset number is based on practical experience and is not further specified here.

[0096] Specifically, the poor quality user determination device can, after acquiring N first network quality data points, divide these N first network quality data points into M first quality data points (i.e., 1 first quality data point) and NM second quality data points (i.e., N-1 second quality data points). Then, it can calculate the user distances between each of these M first quality data points and each of the N-1 second quality data points to obtain the N-1th user distances. This allows it to determine the second quality data point corresponding to the largest user distance among the N-1th user distances, which is the (M+1)th first quality data point (i.e., the 2nd first quality data point), resulting in 2 first quality data points and N-2 second quality data points. Thus, it can calculate the user distances between the (M+1)th first quality data point and each of the NM-1th second quality data points, i.e., calculate the user distances between the 2nd first quality data point and each of the N-2th second quality data points, to obtain the N-2th user distances. Based on this, the poor quality user determination device can select the smallest K-th user distance from NM-1 K-th user distances, and select the smallest K+1 user distance from NM-1 K+1 user distances. This allows it to determine the smallest K-th user distance. The maximum user distance between the smallest K+1 user distance and the smallest K+1 user distance is the target user distance. The second quality data corresponding to this target user distance is then determined as the M+2nd first quality data, i.e., the third first quality data, to obtain M+3 first quality data and NM-2 second quality data, until a preset number of first quality data is obtained. After obtaining the preset number of first quality data, multiple third network quality data are determined based on this preset number of first quality data.

[0097] In this way, the above process can be used to identify multiple third network quality data with low correlation among multiple first network quality data, so as to avoid the low accuracy of identifying poor quality users in the subsequent process of identifying poor quality users due to the high correlation between the acquired network quality data. This can effectively improve the accuracy of identifying poor quality users.

[0098] In one embodiment, to more accurately acquire multiple third-party network quality data points, thereby facilitating the subsequent improvement of the accuracy in identifying users with poor network quality, the aforementioned S260 may include the following steps:

[0099] Using each first mass data point in a preset quantity of first mass data as the center and a preset distance as the radius, divide the first region to obtain multiple first regions;

[0100] Determine a first preset quantity of first quality data for each of the multiple first regions, which constitutes multiple third network quality data.

[0101] Specifically, after obtaining a preset number of first quality data, the poor quality user determination device can divide the data into first regions with each first quality data point as the center and a preset distance as the radius, thus obtaining a preset number of first regions. Furthermore, a first preset number of first quality data points can be determined within each of these first regions, resulting in multiple third network quality data points. The preset distance can be a distance pre-set based on actual conditions or empirical values. The first preset number is a number pre-set based on actual conditions or empirical values.

[0102] In one example, the preset number of first quality data is y first quality data. These data have low correlation with each other and are more representative of the characteristics of users in a region. Therefore, the device for determining poor quality users can use each of the y first quality data as the center point and a preset distance as the radius to divide the region into multiple first regions. Then, multiple first quality data can be selected in each of the multiple first regions. For example, if g first quality data are selected in each first region, then y×g first quality data can be determined as the third network quality data.

[0103] In this embodiment, after determining a preset number of first quality data, the poor quality user determination device can divide multiple first regions with each first quality data as the center and a preset distance as the radius. Then, based on the first preset number of first quality data included in each first region, multiple third network quality data can be determined. In this way, by determining a preset number of first quality data with low correlation to each other, regions can be divided with each quality data as the center, and the acquired multiple first quality data can be used as third network quality data, which facilitates the improvement of the accuracy of determining poor quality users in the future.

[0104] In one embodiment, in order to accurately determine the similarity between any two second network quality data points after obtaining multiple second network quality data points, and to adjust the similarity using a correction coefficient, such as... Figure 3As shown, the steps involved in S130 above may include the following:

[0105] S310, perform similarity calculations on the multiple dimensions of the second network quality data to obtain multiple first similarities with each dimension of the data.

[0106] Specifically, after multiple second network quality data, the poor quality user determination device can perform similarity calculations for each dimension of the multiple dimensions included in the second network quality data. For example, by calculating the cosine of the angle between each dimension and the preset origin, the similarity of that dimension can be obtained, and thus multiple first similarities can be obtained corresponding to each dimension of the multiple dimensions.

[0107] It should be noted that the first similarity scores are calculated for any one dimension of the data in a multi-dimensional dataset. This can be achieved using a matrix. To illustrate, for example, regarding the dimension of home network data, a 23 This indicates the similarity between the home network data included in the second set of second network quality data and the home network data included in the third set of second network quality data. If a... 23 =0 means that the similarity between the home network data included in the second second network quality data and the home network data included in the third second network quality data has been removed.

[0108] S320, according to the first preset weight, the multiple first similarities corresponding to each dimension data in the multiple dimension data are weighted and summed to obtain the similarity between every two second network quality data in the multiple second network quality data.

[0109] Specifically, after determining multiple first similarities corresponding to each dimension of data in multiple dimensions, the poor quality user determination device can, based on a first preset weight, perform a weighted summation of the obtained multiple first similarities corresponding to each dimension of data in multiple dimensions to obtain the similarity between every two second network quality data in multiple second network quality data.

[0110] In one example, when the second network quality data includes multiple dimensions such as home network quality data, service quality data, and bearer network data, the similarity between any two second network quality data can be calculated using the following formula:

[0111] M(i,j)=w1M1(i,j)+w2M2(i,j)+w3M3(i,j) (1)

[0112] Where i and j represent the numbers of the second network quality data, w1, w2, and w3 are weights, M1(i, j) is a matrix including multiple first similarities for home network data, M2(i, j) is a matrix including multiple first similarities for service quality data, and M3(i, j) is a matrix including multiple first similarities for bearer network data.

[0113] S330, calculate the correction coefficient for adjusting similarity based on the multiple service quality data included in the multiple second network quality data.

[0114] The device for identifying users with poor service quality can calculate a correction coefficient for adjusting similarity based on on-site service data from multiple service quality data included in multiple second network quality data.

[0115] Specifically, for any two sets of second network quality data, where the service quality data in one set indicates that the corresponding user has a history of on-site service, while the service quality data in the other set indicates that the corresponding user does not have a history of on-site service, the correction coefficient for adjusting similarity can be calculated using the following formula:

[0116]

[0117] Here, A and B represent the number of people who received on-site service and the number who did not, respectively. If both sets of second network quality data include service quality data indicating a history of on-site service, the correction factor is X2; if both sets of second network quality data include service quality data indicating no history of on-site service, the correction factor is X3. X2 and X3 are identical in terms of service quality description similarity and do not require adjustment. For example, X1 can be calculated as 0.5, X2 as 1, and X3 as 1.

[0118] In this embodiment, similarity calculations can be performed on the multiple dimensions of data included in the second network quality data to obtain multiple first similarities corresponding to each dimension. Then, based on a first preset weight, the multiple first similarities corresponding to the number of each dimension in the multiple dimensions can be weighted and summed to obtain the similarity between every two pieces of second network quality data. Furthermore, based on every two pieces of service quality data included in the multiple second network quality data, correction coefficients for adjusting the similarity can be calculated to obtain multiple correction coefficients. In this way, the similarity between every two pieces of network quality data in the multiple second network quality data, as well as the corresponding correction coefficients, can be accurately obtained, facilitating the accurate identification of users with poor service quality.

[0119] In one embodiment, the above-mentioned step S310 may include the following steps:

[0120] The similarity of each dimension of the second network quality data is calculated separately to obtain multiple second similarities for each dimension of the data.

[0121] In the multi-dimensional data, the first similarity is determined by identifying the first similarity among the multiple second similarities corresponding to each dimension data that is greater than the preset similarity threshold corresponding to each dimension data. This results in multiple first similarities corresponding to each dimension data in the multi-dimensional data.

[0122] The second similarity is represented by the correlation between every two second network quality data. The preset similarity threshold is a threshold pre-set based on practical experience, which will not be elaborated on here.

[0123] In one example, during the calculation of multiple first similarities for the dimension of home network data, multiple second similarities can be calculated first for multiple home network data. Then, based on a pre-set first threshold, if the cosine of the angle between a certain second network quality data's home network data and a preset origin is less than the first threshold, the home network data of that second network quality data can be treated as noise and removed, thus obtaining multiple first similarities.

[0124] In another example, during the calculation of the first similarity between service quality data in multiple second network quality data sets, multiple second similarities can be calculated for the multiple service quality data sets. Noise data can be removed based on a pre-set second threshold. For example, the similarity can be calculated based on the service overlap included in each service quality data set. Assuming the second threshold is 0.87, if service quality data A contains 5 service lists and service quality data B contains 6 service lists, and the 5 service lists included in service data A are identical to the 5 service lists included in the 6 service lists included in service quality data B, their service overlap is 0.83, which is less than the second threshold of 0.87. Therefore, this overlap can be treated as noise data and removed. In this way, multiple first similarities for the service quality data can be obtained.

[0125] It should be noted that the thresholds for the various dimensions of data included in the second network quality data can be the same or different; no specific restrictions are imposed here.

[0126] In this way, after performing preliminary similarity calculations on each dimension of the multi-dimensional data to obtain multiple second similarities for each dimension, and based on a preset similarity threshold for each dimension, the similarities among the second similarities for each dimension that are greater than the preset similarity threshold are determined as first similarities, thus obtaining multiple first similarities for each dimension. This improves the accuracy of the determined first similarities.

[0127] In one embodiment, to determine a more accurate target score, and to subsequently determine whether a user is a poor-performing user based on the target score, thereby improving the accuracy of identifying poor-performing users, the above-mentioned step S140 may include the following steps:

[0128] S410, adjust the similarity of each pair of second network quality data according to the product of the correction coefficient and the similarity of each pair of second network quality data in multiple second network quality data, and obtain multiple adjusted similarities.

[0129] After obtaining the correction coefficient for each pair of second network quality data, the similarity of each pair of second network quality data can be adjusted based on the correction coefficient and the product of the similarity of each pair of second network quality data corresponding to the correction coefficient. That is, the correction coefficient and the product of the similarity of each pair of second network quality data corresponding to the correction coefficient are determined as the new similarity of each pair of the above-mentioned second network quality data.

[0130] S420, if the target similarity group meets the preset conditions, determine multiple fourth network quality data according to the target similarity group so that the target similarity group does not meet the preset conditions.

[0131] The target similarity group comprises two similarities determined sequentially from a set of adjusted similarities, following a preset order. This preset order can be either descending or ascending order of similarities. The preset conditions are conditions pre-set based on practical experience; for example, a preset condition could include that the difference between the reciprocals of the two similarities in the target similarity group is no greater than a preset threshold. The preset threshold can be a threshold pre-set based on actual circumstances or experience.

[0132] Specifically, the poor quality user determination device can select two similarities from multiple adjusted similarities in ascending order, and then determine whether the difference between the reciprocals of the two selected similarities is not greater than a preset threshold. If it is not greater than the preset threshold, the two second network quality data corresponding to the two similarities are updated according to the average value until the difference between the reciprocals of the two selected similarities is greater than the preset threshold, so as to obtain multiple fourth network quality data.

[0133] S430, Calculate the variance of the network quality data for each dimension based on the network quality data for each of the multiple fourth network quality data.

[0134] Specifically, the poor quality user determination device can calculate the variance of the network quality data for each dimension based on the network quality data included in each of the multiple fourth network quality data.

[0135] In one example, suppose the fourth network quality data can include five dimensions of data. For the A dimension data, according to the preset scoring rules, the A dimension data included in the fourth network quality data will be scored, resulting in multiple scores for the A dimension data. The variance of the A dimension data can then be calculated based on these scores.

[0136] S440, determine the target score based on the product of variance and the second preset weight.

[0137] After obtaining multiple variances, the quality defect user determination device can determine the target score based on the product of the multiple variances and a second preset weight. That is, the target score is obtained by multiplying the multiple variances and the second weight. The second preset weight can be a weight pre-set based on actual experience, which will not be elaborated on here.

[0138] In this embodiment, the quality defect determination device can adjust the similarity of every two second network quality data points in a plurality of second network quality data points based on a correction coefficient. It can then sequentially select two similarities from the adjusted similarities according to a preset order to determine whether they meet preset conditions. Based on the selected two similarities, multiple fourth network quality data points are obtained. Furthermore, based on the multiple dimensions of network quality data included in each of the multiple fourth network quality data points, the variance of the network quality data in each dimension is calculated. The target score is then determined by multiplying the variance by a second preset weight. This allows for a more accurate determination of the target score, improving the accuracy of identifying users with poor quality.

[0139] Based on the same inventive concept, embodiments of this application also provide a device for determining users with poor quality. (Specifically combined with...) Figure 5 The device for determining poor quality users provided in the embodiments of this application will be described in detail.

[0140] Figure 5 This is a schematic diagram of a device for determining poor user quality provided in an embodiment of this application.

[0141] like Figure 5 As shown, the poor quality user determination device 500 may include: an acquisition module 510, a clustering module 520, a first determination module 530, an adjustment module 540, and a second determination module 550.

[0142] The acquisition module 510 is used to acquire the first network quality data of multiple first users within a first preset time period in the target area.

[0143] The clustering module 520 is used to cluster multiple first network quality data based on preset reasons for poor quality and preset clustering radius to obtain multiple second network quality data, wherein the number of first network quality data is greater than the number of second network quality data.

[0144] The first determining module 530 is used to determine the similarity between every two second network quality data in the multiple second network quality data, and to adjust the similarity by a correction coefficient.

[0145] The adjustment module 540 is used to adjust the similarity of every two second network quality data in multiple second network quality data based on the correction coefficient to obtain a target score for evaluating the user's network quality.

[0146] The second determining module 550 is used to determine the first user as a poor-quality user when the user score of the first user is less than the target score. The user score is set based on the first user's first network quality data and preset scoring rules.

[0147] In one embodiment, the clustering module may include a first calculation submodule and a first determination submodule.

[0148] The first calculation submodule is used to calculate the user distance between multiple first network quality data based on multiple first network quality data, so as to determine multiple third network quality data from multiple first network quality data based on the user distance. The number of third network quality data is greater than the number of second network quality data. The user distance is represented as the correlation between multiple first network quality data.

[0149] The first determination submodule is used to determine the network quality data that is less than or equal to the preset cluster radius by performing clustering processing on multiple third network quality data based on preset reasons for poor quality and preset cluster radius, so as to obtain multiple second network quality data.

[0150] In one embodiment, the first calculation submodule mentioned above may include: a partitioning unit, a first calculation unit, a first determination unit, a second calculation unit, a second determination unit, and a third determination unit.

[0151] A partitioning unit is used to divide N first network quality data into M first quality data and NM second quality data.

[0152] The first calculation unit is used to calculate the user distances between the Mth first quality data and the NM second quality data, and obtain the NM Kth user distances.

[0153] The first determining unit is used to determine the second quality data corresponding to the largest user distance among the NM distances to the Kth user, which is the M+1th first quality data, thus obtaining M+1 first quality data and NM-1 second quality data.

[0154] The second calculation unit is used to calculate the user distance between the (M+1)th first quality data and the NM-1th second quality data, and obtain the NM-1th (K+1)th user distance.

[0155] The second determining unit is used to determine the second quality data corresponding to the distance to the target user as the M+2th first quality data, to obtain M+2 first quality data and NM-2 second quality data, so as to obtain a preset number of first quality data, wherein the target user distance is the smallest user distance among NM-1 Kth user distances and the largest user distance among the smallest user distances among NM-1 K+1 user distances.

[0156] The third determining unit is used to determine multiple third network quality data based on a preset number of first quality data.

[0157] The number of first network quality data is N, where N is a positive integer greater than 3, M = K = 1, and both the first quality data and the second quality data are first network quality data.

[0158] In one embodiment, the third determining unit may include a dividing subunit and a first determining subunit.

[0159] The sub-unit is used to divide the first region with each first quality data in the first quality data of a preset quantity as the center and a preset distance as the radius, to obtain multiple first regions. Each of the multiple first regions includes multiple first quality data.

[0160] The first determining sub-unit determines multiple first quality data within multiple regions, which are then converted into multiple third network quality data.

[0161] In one embodiment, the first determining module may include: a similarity calculation submodule, a weighted summation module, and a second calculation submodule.

[0162] The similarity calculation submodule is used to perform similarity calculations on the multiple dimensions of the second network quality data, and obtain multiple first similarities with each dimension of the data.

[0163] The weighted summation module is used to perform weighted summation on multiple first similarities corresponding to each dimension of data in multiple dimensions of data according to a first preset weight, so as to obtain the similarity between every two second network quality data in multiple second network quality data.

[0164] The second calculation submodule is used to calculate the correction coefficient for adjusting similarity based on the multiple service quality data included in the multiple second network quality data.

[0165] In one embodiment, the similarity calculation submodule may include a similarity calculation unit and a fourth determination unit.

[0166] The similarity calculation unit is used to perform similarity calculations on the data of each dimension included in the second network quality data, and to obtain multiple second similarities corresponding to each dimension of the data.

[0167] The fourth determining unit is used to determine the first similarity among the multiple second similarities corresponding to each dimension of data in the multiple dimensions of data, which is greater than the preset similarity threshold corresponding to each dimension of data, and obtain multiple first similarities corresponding to each dimension of data in the multiple dimensions of data.

[0168] In one embodiment, the adjustment module may include: an adjustment submodule, a second determination submodule, and a third calculation submodule.

[0169] The adjustment submodule is used to adjust the similarity of each pair of second network quality data based on the product of the correction coefficient and the similarity of each pair of second network quality data in multiple second network quality data, so as to obtain multiple adjusted similarities.

[0170] The second determining submodule is used to determine multiple fourth network quality data based on the target similarity group when the target similarity group meets the preset conditions, so that the target similarity group does not meet the preset conditions. The target similarity group includes two similarities determined sequentially from multiple adjusted similarities in a preset order.

[0171] The third calculation submodule is used to calculate the variance of the network quality data for each dimension based on the network quality data for each of the multiple dimensions of the network quality data included in the multiple fourth network quality data.

[0172] The third determination submodule is used to determine the target score based on the product of the variance and the second preset weight.

[0173] In one embodiment, the preset condition includes that the difference between the reciprocals of the two similarity groups included in the target similarity group is not greater than a preset threshold.

[0174] In this embodiment, by acquiring first network quality data of multiple first users within a target area over a first preset time period, and then clustering these first network quality data based on preset reasons for poor quality and preset clustering radii, multiple second network quality data are obtained. Based on these second network quality data, the similarity between any two second network quality data points and a correction coefficient for adjusting the similarity can be determined. Furthermore, based on the correction coefficient, the similarity between any two second network quality data points can be adjusted to obtain a target score for evaluating user network quality. This allows a first user to be identified as a user with poor quality if their score, based on their first network quality data and a preset scoring rule, is less than the target score. Thus, by clustering the acquired first network quality data and calculating correction coefficients to adjust the similarity between any two second network quality data points obtained from the clustering process, the problem of uneven distribution of user network quality data is eliminated, thereby improving the accuracy of identifying users with poor quality.

[0175] The various modules in the poor quality user determination device provided in this application embodiment can achieve... Figures 1 to 4 The method steps of any of the embodiments shown can achieve the corresponding technical effects, and for the sake of brevity, they will not be described in detail here.

[0176] Figure 6 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0177] An electronic device may include a processor 601 and a memory 602 storing computer program instructions.

[0178] Specifically, the processor 601 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0179] Memory 602 may include mass storage for data or instructions. For example, and not limitingly, memory 602 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 602 may include removable or non-removable (or fixed) media. Where appropriate, memory 602 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 602 is non-volatile solid-state memory.

[0180] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the methods according to one aspect of this disclosure.

[0181] The processor 601 implements any of the poor quality user determination methods in the above embodiments by reading and executing computer program instructions stored in the memory 602.

[0182] In one example, the electronic device may also include a communication interface 603 and a bus 610. For example, Figure 6 As shown, the processor 601, memory 602, and communication interface 603 are connected through bus 610 and complete communication with each other.

[0183] The communication interface 603 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0184] Bus 610 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 610 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0185] Furthermore, in conjunction with the method for determining poor-quality users in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement the method for determining poor-quality users provided in this application embodiment.

[0186] This application also provides a computer program product, in which instructions are executed by the processor of an electronic device, causing the electronic device to perform the scientific and technological innovation achievement evaluation method provided in this application.

[0187] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0188] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0189] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0190] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable poor-quality user determination device to produce a machine such that these instructions, executable via the processor of the computer or other programmable poor-quality user determination device, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0191] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for identifying users with poor quality of service, characterized in that, The method includes: Obtain first network quality data of multiple first users within a first preset time period within the target area; Based on preset reasons for poor quality and preset clustering radius, the multiple first network quality data are clustered to obtain multiple second network quality data, wherein the number of first network quality data is greater than the number of second network quality data. Based on the plurality of second network quality data, determine the similarity between every two second network quality data in the plurality of second network quality data, and a correction coefficient for adjusting the similarity; Based on the correction coefficient, the similarity between every two second network quality data in the plurality of second network quality data is adjusted to obtain a target score for evaluating the user's network quality; If the user rating of the first user is less than the target rating, the first user is determined to be a poor quality user. The user rating is set based on the first user's first network quality data and a preset rating rule. The step of determining the similarity between any two second network quality data points based on the plurality of second network quality data points, and the correction coefficient used to adjust the similarity, includes: Similarity calculations are performed on the multiple dimensions of data included in the second network quality data to obtain multiple first similarities corresponding to each dimension of data. Based on the first preset weight, the multiple first similarities corresponding to each dimension data in the multiple dimension data are weighted and summed to obtain the similarity between every two second network quality data in the multiple second network quality data. Based on the multiple service quality data included in the multiple second network quality data, a correction coefficient for adjusting similarity is calculated; The step of adjusting the similarity between every two second network quality data points based on the correction coefficient to obtain a target score for evaluating user network quality includes: The similarity of each pair of second network quality data is adjusted by multiplying the correction coefficient with the similarity of each pair of second network quality data in the plurality of second network quality data, resulting in a plurality of adjusted similarities; If the target similarity group meets the preset conditions, multiple fourth network quality data are determined according to the target similarity group so that the target similarity group does not meet the preset conditions. The target similarity group includes two similarities determined sequentially from the multiple adjusted similarities in a preset order. Based on the network quality data of each of the multiple fourth network quality data, which includes multiple dimensions of network quality data, calculate the variance of the network quality data of each dimension. The target score is determined by multiplying the variance by the second preset weight.

2. The method according to claim 1, characterized in that, Based on preset causes of poor quality and preset clustering radii, the plurality of first network quality data are clustered to obtain a plurality of second network quality data, including: Based on the plurality of first network quality data, a user distance is calculated between the plurality of first network quality data to determine a plurality of third network quality data from the plurality of first network quality data based on the user distance, wherein the number of third network quality data is greater than the number of second network quality data, and the user distance characterizes the correlation between the plurality of first network quality data; Based on the preset reasons for poor quality and the preset clustering radius, the multiple third network quality data are clustered to determine the network quality data whose user distance from the cluster center point is less than or equal to the preset clustering radius. These are then designated as second network quality data, thus obtaining multiple second network quality data.

3. The method according to claim 2, characterized in that, The step of calculating the user distance between the plurality of first network quality data based on the plurality of first network quality data, and determining a plurality of third network quality data from the plurality of first network quality data based on the user distance, includes: The N first network quality data are divided into M first quality data and NM second quality data; Calculate the user distances between the Mth first quality data point and the NM second quality data points to obtain the NM Kth user distances; The second quality data corresponding to the largest user distance among the NM Kth user distances is determined as the M+1th first quality data, thus obtaining M+1 first quality data and NM-1 second quality data; Calculate the user distances between the (M+1)th first quality data point and the (NM-1)th second quality data points to obtain the (NM-1)th (K+1)th user distance; The second quality data corresponding to the distance to the target user is determined as the M+2th first quality data, and M+2 first quality data and NM-2 second quality data are obtained to obtain a preset number of first quality data. Among them, the target user distance is the smallest user distance among NM-1 Kth user distances and the largest user distance among the smallest user distances among NM-1 K+1 user distances. Based on a preset number of first quality data, multiple third network quality data are determined; Wherein, the quantity of the first network quality data is N, where N is a positive integer greater than 3, M=K=1, and both the first quality data and the second quality data are first network quality data.

4. The method according to claim 3, characterized in that, The step of determining multiple third network quality data based on a preset number of first quality data includes: Using each first quality data point in a preset quantity of first quality data as the center and a preset distance as the radius, a first region is divided to obtain multiple first regions. Each of the multiple first regions includes multiple first quality data points. Determine a first preset quantity of first quality data for each of the multiple first regions, which constitutes multiple third network quality data.

5. The method according to claim 1, characterized in that, The similarity calculation is performed on the multiple dimensions of the second network quality data to obtain multiple first similarities corresponding to each dimension of the multiple dimensions of data, including: The similarity is calculated for each dimension of the second network quality data to obtain multiple second similarities for each dimension of the data. The first similarity is determined by identifying the first similarity among the multiple second similarities corresponding to each dimension of the data that is greater than the preset similarity threshold corresponding to each dimension. This results in multiple first similarities corresponding to each dimension of the data.

6. The method according to claim 1, characterized in that, The preset condition includes that the difference between the reciprocals of the two similarity groups included in the target similarity group is not greater than a preset threshold.

7. A device for identifying users with poor quality, characterized in that, The device includes: The acquisition module is used to acquire the first network quality data of multiple first users within a first preset time period in the target area; The clustering module is used to cluster the multiple first network quality data based on a preset cause of poor quality and a preset clustering radius to obtain multiple second network quality data, wherein the number of first network quality data is greater than the number of second network quality data. The first determining module is used to determine the similarity between every two second network quality data in the plurality of second network quality data based on the plurality of second network quality data, and to adjust the similarity by a correction coefficient. The adjustment module is used to adjust the similarity of every two second network quality data in the plurality of second network quality data based on the correction coefficient, so as to obtain a target score for evaluating the user's network quality. The second determining module is used to determine the first user as a poor-quality user when the user rating of the first user is less than the target rating, wherein the user rating is set based on the first user's first network quality data and a preset rating rule; The first determining module includes: The similarity calculation submodule is used to perform similarity calculation on the multiple dimensions of data included in the second network quality data, and obtain multiple first similarities with each dimension of data. The weighted summation module is used to perform a weighted summation of the multiple first similarities corresponding to each dimension data in the multiple dimension data according to a first preset weight, so as to obtain the similarity between every two second network quality data in the multiple second network quality data. The second calculation submodule is used to calculate a correction coefficient for adjusting similarity based on the multiple service quality data included in the multiple second network quality data. The adjustment module includes: The adjustment submodule is used to adjust the similarity of each pair of second network quality data according to the product of the correction coefficient and the similarity of each pair of second network quality data in the plurality of second network quality data, so as to obtain a plurality of adjusted similarities; The second determining submodule is used to determine multiple fourth network quality data based on the target similarity group when the target similarity group meets the preset conditions, so that the target similarity group does not meet the preset conditions. The target similarity group includes two similarities determined sequentially from the multiple adjusted similarities in a preset order. The third calculation submodule is used to calculate the variance of the network quality data for each dimension based on the network quality data for each of the multiple fourth network quality data. The third determining submodule is used to determine the target score based on the product of the variance and the second preset weight.

8. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; The processor reads and executes the computer program instructions to implement the poor quality user determination method as described in any one of claims 1-6.

9. A computer storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the poor quality user determination method as described in any one of claims 1-6.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device performs the poor quality user determination method as described in any one of claims 1-6.

Images