CN113099475B - Network quality detection method, device, electronic equipment and readable storage medium - Google Patents

Abstract

The present application discloses a network quality detection method, device, electronic device and readable storage medium, which belongs to the field of communication technology. The network quality detection method includes: in response to a user's network quality complaint request, obtaining the user's target data information, the target data information is used to locate the user; inputting the data information into a network status prediction model to obtain the user's network problem, the network status prediction model is a model that predicts the user's network status based on the time factors, regional scenarios and network factors that affect the network status.

Description

Network quality detection method, device, electronic device and readable storage medium

Technical Field

The present application relates to the field of communication technology, and in particular to a network quality detection method, device, electronic device and readable storage medium.

Background technique

In the operation of wireless communication networks, equipment failures often occur. At this time, mobile phone users at a certain location cannot communicate with the base station, or the communication with the base station is not smooth, which is specifically manifested as no signal on the user's mobile phone, unable to dial, dropped calls, etc.

Typically, wireless communication network operators provide complaint services for mobile phone users. Through complaints from mobile phone users, network operators can discover problems in the network and solve them. Currently, network operators usually receive user complaints through complaint telephones and enter user complaints into the system for further processing by back-end network maintenance and optimization personnel.

Existing network problems are difficult to predict, and the network problem-solving method based on user complaints is relatively passive and lagging, and the speed of solving network problems is slow.

Summary of the invention

The purpose of the embodiments of the present application is to provide a network quality detection method, device, electronic device and readable storage medium to at least solve the problem that existing network problems are difficult to predict.

The technical solution of this application is as follows:

According to a first aspect of an embodiment of the present application, a network quality detection method is provided, and the method may include:

In response to a user's network quality complaint request, obtaining the user's target data information, the target data information is used to locate the user;

The data information is input into the network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on the time factors, regional scenarios and network factors that affect the network status.

Furthermore, the target data information may be user location information and/or user telephone number information.

Furthermore, before obtaining the user's target data information in response to the user's network quality complaint request, the method may further include:

Obtain network complaint data at daily granularity;

Extract quantifiable scenario complaint evaluation index data from network complaint data to obtain a complaint evaluation index set;

Delete the holiday data in the complaint evaluation index set, and mark the complaint evaluation index set according to the time series characteristics to obtain sample data;

The machine learning model is trained using sample data to obtain a network status prediction model.

Furthermore, quantifiable scenario complaint evaluation index data is extracted from the network complaint data to obtain a complaint evaluation index set, which may include:

Normalize the abnormal data in the network complaint data to obtain regular indicator data;

Extract quantifiable scenario complaint evaluation index data from conventional index data to obtain a complaint evaluation index set.

Furthermore, quantifiable scenario complaint evaluation index data may include: KQI, KPI, AOI scenario, environmental index, abnormal index and soft complaint index.

Furthermore, the machine learning model is trained and tested using sample data to obtain a network status prediction model, which may include:

Performing feature enhancement processing on the sample data to obtain enhanced sample data;

The machine learning model is trained and tested using enhanced sample data to obtain a network status prediction model.

Furthermore, after inputting the data information into the network status prediction model to obtain the user's network problem, the above method may further include:

Generate appeal explanations based on network issues;

Send an explanation of the request to the user.

According to a second aspect of an embodiment of the present application, a method for determining a target user is provided, and the method may include:

Using the network status prediction model of the first aspect of the embodiment to predict an area where the probability of network problems exceeding a preset value occurs;

Collect network-side big data from users in areas with high risk of network problems;

Extract voice performance indicators and Internet quality indicators from the big data on the network side;

Target users are determined based on voice performance indicators and Internet quality indicators, combined with users' traffic usage, average user income, and B-domain economic data.

According to a third aspect of an embodiment of the present application, a network quality detection device is provided, which may include:

An acquisition device, used to respond to a user's network quality complaint request and acquire the user's target data information, the target data information being used to locate the user;

The prediction module is used to input data information into the network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on time factors, regional scenarios and network factors that affect the network status.

According to a fourth aspect of an embodiment of the present application, a target user determination device is provided, and the device may include:

A regional prediction module, used to predict a region where a probability of a network problem occurring exceeds a preset value by using the network status prediction model of the first aspect of the embodiment;

The collection module is used to collect network-side big data of users in areas with high risk of network problems;

An extraction module is used to extract voice performance indicators and Internet quality indicators from the big data on the network side;

The determination module is used to determine the target user based on the voice performance index and the Internet quality index, combined with the user's traffic usage, the user's average income and the B-domain economic data.

According to a fifth aspect of an embodiment of the present application, an electronic device is provided, which may include:

processor;

a memory for storing processor-executable instructions;

The processor is configured to execute instructions to implement the network quality detection method shown in any one of the embodiments of the first aspect or the target user determination method shown in any one of the embodiments of the second aspect.

According to the sixth aspect of the embodiments of the present application, a storage medium is provided. When the instructions in the storage medium are executed by a processor of an information processing device or a server, the information processing device or the server implements the method shown in any one of the embodiments of the first aspect or the target user determination shown in the embodiments of the second aspect.

The technical solution provided by the embodiments of the present application brings at least the following beneficial effects:

The embodiment of the present application obtains the user's target data information, and the target data information is used to locate the user; the data information is input into a network status prediction model to obtain the user's network problem. The network status prediction model used by the method comprehensively considers the time factors, regional scenarios and network factors that affect the network status, and the factors that affect the network problem are considered more comprehensively. Therefore, the model predicts the network problem more accurately, and then the complaining user can be predicted according to the network status, which can solve the network complaint problem in a more targeted manner.

It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application, and do not constitute improper limitations on the present application.

FIG1 is a flow chart of a network quality detection method according to an exemplary embodiment;

FIG2 is a flow chart of a method for intelligent prediction of network complaints according to a specific embodiment;

FIG3 is a flowchart of data collection and extraction according to a specific embodiment;

FIG4 is a schematic diagram of a box plot recognition method according to a specific embodiment;

FIG5 is a schematic diagram of a time series decomposition according to a specific embodiment;

FIG6 is a flowchart of establishing a high-risk scenario complaint prediction model according to a specific embodiment;

FIG7 is a schematic diagram of sample iteration according to a specific embodiment;

FIG8 is a first schematic diagram of a prediction result according to a specific embodiment;

FIG9 is a second schematic diagram of prediction results according to a specific embodiment;

FIG10 is a schematic diagram of a prediction process according to a specific embodiment;

FIG11 is a schematic diagram showing the structure of an electronic device according to an exemplary embodiment;

Fig. 12 is a schematic diagram showing a hardware structure of an electronic device according to an exemplary embodiment.

Detailed ways

In order to enable ordinary persons in the art to better understand the technical solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims.

The network quality detection method provided in the embodiment of the present application is described in detail below through specific embodiments and their application scenarios in conjunction with the accompanying drawings.

As shown in FIG. 1 , in a first aspect of an embodiment of the present application, a network quality detection method is provided, and the method may include:

Step 110: In response to the user's network quality complaint request, obtain the user's target data information, the target data information is used to locate the user;

Step 120: Input the data information into a network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on time factors, regional scenarios, and network factors that affect the network status.

The network status prediction model used in the above-mentioned embodiment method comprehensively considers the time factors, regional scenarios and network factors that affect the network status, and considers the factors affecting network problems more comprehensively. Therefore, the model predicts network problems more accurately, and can predict complaining users based on the network status, so as to solve network complaint problems in a more targeted manner.

The specific implementation of the above steps will be described in detail below.

First, step 110 is introduced, in response to the user's network quality complaint request, the user's target data information is obtained, and the target data information is used to locate the user.

In this step, the user may be located based on the location information provided by the user or based on the user's phone number.

Then, step 120 is introduced, in which data information is input into a network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on time factors, regional scenarios, and network factors that affect the network status.

The network status prediction model in this step can quickly identify user scenarios, user network perception status, locate current and historical network problems, generate explanations, quickly respond to customer demands, and improve customer network experience.

In some optional embodiments of the present application, the target data information may be user location information and/or user telephone number information.

In some optional embodiments of the present application, before obtaining the user's target data information in response to the user's network quality complaint request, the method may further include:

Obtain network complaint data at daily granularity;

Extract quantifiable scenario complaint evaluation index data from network complaint data to obtain a complaint evaluation index set;

Delete the holiday data in the complaint evaluation index set, and mark the complaint evaluation index set according to the time series characteristics to obtain sample data;

The machine learning model is trained using sample data to obtain a network status prediction model.

In some optional embodiments of the present application, extracting quantifiable scenario complaint evaluation index data from network complaint data to obtain a complaint evaluation index set may include:

Normalize the abnormal data in the network complaint data to obtain regular indicator data;

Extract quantifiable scenario complaint evaluation index data from conventional index data to obtain a complaint evaluation index set.

In some optional embodiments of the present application, quantifiable scenario complaint evaluation index data may include: KQI, KPI, AOI scenario, environmental index, abnormal index and soft complaint index.

In some optional embodiments of the present application, training and testing a machine learning model using sample data to obtain a network status prediction model may include:

Performing feature enhancement processing on the sample data to obtain enhanced sample data;

The machine learning model is trained and tested using enhanced sample data to obtain a network status prediction model.

In some optional embodiments of the present application, after inputting the data information into the network status prediction model to obtain the user's network problem, the above method may further include:

Generate appeal explanations based on network issues;

Send an explanation of the request to the user.

According to a second aspect of an embodiment of the present application, a method for determining a target user is provided, and the method may include:

Using the network status prediction model of the first aspect of the embodiment to predict an area where the probability of network problems exceeding a preset value occurs;

Collect network-side big data from users in areas with high risk of network problems;

Extract voice performance indicators and Internet quality indicators from the big data on the network side;

Target users are determined based on voice performance indicators and Internet quality indicators, combined with users' traffic usage, average user income, and B-domain economic data.

The target user in this embodiment may be a high-quality user who has filed a network complaint.

As shown in FIG2 , in a specific embodiment of the present application, a network complaint intelligent prediction method is provided, the method comprising:

Step 100: Collect and extract network optimization and Internet data, and build a quantifiable scenario complaint evaluation index set

Step 200: Build a feature training library based on time series

Step 300: Establish a high-risk scenario complaint prediction model based on the CatBoost algorithm

Step 400: Quantitative evaluation of user perceived experience in high-risk areas

Step 500: Complaint warning empowers online customer service support

Among them, step 100: collecting and extracting network optimization, Internet, and classified data, as shown in FIG3 , may include:

Step 101: Data is collected at a daily granularity and aggregated to a geographic location;

Data related to complaint indicators is collected at a daily granularity, and the collection time is recorded. The collection time is specific to the daily granularity and is collected based on the existing network base station dimension. The collected data includes KPI, KQI, QoE, network environment, abnormal events and other data, as shown in the following table:

Step 102, abnormal data cleaning;

Outliers are unreasonable values in the data, which may be sudden increases, sudden decreases, or missing values. The box plot analysis method is used to detect outliers. The box plot determines outliers based on quartiles and interquartile ranges. Quartiles are very robust, and 25% of the data can be arbitrarily far values without interfering with the quartile values, so outliers do not affect this standard. Therefore, the box plot recognition method is more objective in identifying outliers and has certain advantages. The box plot diagram is shown in Figure 4.

Among them, the upper quartile Q3 indicates that the number greater than this value accounts for 1/4 of the total; the lower quartile Q1 indicates that the number less than this value accounts for 1/4 of the total; IQR = Q3-Q1; the upper limit represents the maximum value within the non-abnormal range, upper limit = Q3+1.5IQR; the lower limit represents the minimum value within the non-abnormal range, lower limit = Q1-1.5IQR; therefore, the range of abnormal values is greater than the upper limit (Q3+1.5IQR) and less than the lower limit (Q1-1.5IQR).

After identifying the sudden increase and sudden decrease abnormal values in the data, if it is a sudden increase point, the maximum data value in this range is used as the replacement value of this position; if it is a sudden decrease value, the minimum flow value in this range is used as the replacement value of this position. For a small number of missing values, the data value at the same time point in the previous week is used for filling.

Step 103, constructing a set of quantifiable scenario complaint evaluation indicators;

Select the indicators collected in step 100 to establish a complaint evaluation indicator set. The selected indicators are shown in the following table:

Step 200: Build a time series model training library;

Since the wireless network usage during the weekend is very different from that during the weekday, separate predictions for weekends and weekdays will make the prediction results more accurate. Therefore, the network indicators of weekends and weekdays under the same base station are used as two independent time series to construct training samples based on the collection time. And because holidays have a great impact on the distribution of people and the use of wireless networks, and cannot follow the normal cycle law, the data during holidays in the traffic data is deleted and not used as training sample data. The data storage format in the training library is shown in the following table:

Week Index Collection time Scene Type Scene Name Complaint evaluation index set Index value

Among them, the week index value is weekday or weekend; the collection time is specific to the hour granularity; the scene types include primary and secondary schools, universities, commercial areas, office buildings, residential areas, hospitals and tourist areas; the scene name is the specific school name or shopping mall name, etc.

Since the changes in the collected indicator data series are jointly affected by four factors: long-term trends, seasonal changes, cyclical cycles, and irregular changes, the time series can be expressed as the following additive structural model:

in,

(1) represents the predicted value;

(2) T(t) represents the long-term trend factor: it reflects the long-term development trend, which can be expressed as a continuous upward or downward or stable trend that is approximately linear over a fairly long period of time;

(3) S(t) represents the seasonal variation factor: it is a periodic fluctuation with a fixed length and amplitude caused by the seasonal variation of flow. It is manifested as a periodic change of flow with the change of time sequence within a year or shorter period of time.

(4) C(t) represents the cyclical variation factor: The cyclical variation factor is also called the cyclical variation factor. It is affected by factors such as the economic environment and forms a long-term ups and downs fluctuation;

(5) I(t) represents the irregular variation factor: Irregular variation is also called random variation, which is a non-periodic and non-trend irregular fluctuation caused by sudden, accidental events or some unknown reasons.

The time series decomposition method is used to build a time series feature library. The decomposition method can decompose each factor, and then simply estimate the impact of this factor on the time series, which can overcome the interference of other factors. The decomposition steps of the time series are shown in Figure 5.

Step 300: Establish a high-risk scenario complaint prediction model based on the CatBoost algorithm, as shown in FIG6 , including:

Step 301: type feature processing;

Based on the scene complaint evaluation index set obtained in step 100 and the time series training library obtained in step 200, in the data preprocessing stage, the values of these features are converted into numbers and counted by time period. Generally, categorical features will be converted into one or more numerical features (one-hot encoding, etc.), and the accuracy will be improved by gradient.

When randomly permuting the data set and calculating the average label value of samples with the same category value, the label value of the sample before this sample is included in the calculation. That is, when the feature of each sample is converted to a numerical value, the average is taken based on the category label before this sample, and the priority and priority weight coefficient are added. This approach can reduce the noise caused by low-frequency features in the category features.

Randomly permute the input sample data set.

Convert the label to integer and discretize the target value. There is a Binarization process in CatBoost. At the beginning, there is a parameter K. The target value will be discretized into K+1 buckets, and each value will only be in one of the buckets. In this way, a label is obtained again, with a value from 0 to K.

Features are converted into numerical features: CatBoost chooses which method to use based on the parameters set at the beginning. Let’s review what we said in the previous article. There is an observation data set D = (Xi, Yi) i = 1…n, and after random permutation, we get a sequence σ = (σ1,…,σn). We traverse the random sequence from σ1 to σn in turn.

Use this formula to convert the nominal value into a numerical value, and calculate i∈[0,k-1] according to the bucket to which the label belongs.

The formula used is the same as Borders, and i∈[0,k] is calculated according to the bucket to which the label belongs (K+1 buckets are created), but the parameters have different meanings.

Where totalCount and prior have the same meaning. countInClass is a ratio, where the nominal features in σ1 to σk-1 are the sum of the converted class values of the samples with α, divided by the largest label value k after conversion.

Counter. This calculation method does not depend on the label value. The nominal values of the training set and the validation set are calculated separately.

Use the following formula to calculate the numerical value corresponding to the category attribute value.

Assume that the attribute value of the currently calculated category feature is α.

In the training set, curCount is the number of category features α. maxCount is the maximum number of times the category feature appears in the training set. prior is a given constant.

In the test set, prior is a given constant. The calculation methods of the other two parameters are divided into three cases:

1. PrefixTest, curCount is equal to curCount in the training set plus the number of samples with the same category feature value as of the current test sample. maxCount is the maximum number of times the category feature appears in the new set consisting of the training set and the current test sample.

2. FullTest, take the entire data set (training data set plus test data set), curCount is the number of category features α. maxCount is the maximum number of times the category feature appears in the training set.

3. SkipTest, ignore the test set and use curCount and maxCount calculated in the training set.

Step 301: Orderly enhancement;

To solve the prediction shift mentioned above, the method is as follows:

Randomly generate a training sample of permutation σ of [1,n];

Support n different models M1,…,M2, so that Mi is learned only using the first i samples in the permutation.

At each iteration, we can obtain the j-th sample residual through the model Mj-1, as shown in Figure 7.

Step 301: sample processing and tuning;

1. Experimental setup;

Sample set division: 80% of the sample table is used as training samples, and the remaining 20% is used as validation samples.

Feature attribute screening: scene type, scene area, number of scene cells, number of alarms affecting services, duration of alarms, and other indicators.

Default value setting: For type feature attributes, the default value is set to a specific value (such as the default value of scene type: residential area); for numerical feature attributes, the default value is set to 0.

2. Sample data is shown in the following table:

3. Model parameter tuning

The CatBoost algorithm runs parameters on preprocessed training data and tuned training data. The best results are obtained by continuously optimizing the program running parameters.

learning_rate:Log-uniform distribution[e-7,1]

random_strength: a discrete uniform distribution over the set {1,20}

one_hot_max_size: Discrete uniform distribution over a set {0,25}

l2_leaf_reg: Log-uniform distribution [1,10]

bagging_temperature: uniform [0,1]

gradient_iterations: a discrete uniform distribution over the set {1,10}

4. Complaint prediction results;

Prediction results: The high-risk scenario prediction model can ultimately output the predicted scenario name, scenario type, predicted number of complaints in the scenario, and the deviation from the actual complaint result. The model predicts the number of complaints in a certain area by scenario, and can effectively locate network faults in the scenario area, actively prevent user complaints, solve problems before users complain, and continuously improve user perception. The prediction results are shown in Figure 8-9.

A graphical representation of the prediction process is shown in Figure 10.

Step 400: Quantitative evaluation of user perceived experience in high-risk areas;

Step 401: Collecting user XDR/MR/S1U-HTTP and other network-side big data;

Step 402: extracting key indicators of voice performance and Internet quality respectively;

Step 403: Count the poor quality events of voice call problems and Internet access problems that occur at the hourly and daily granularity of the user;

Step 405: Identify high-value potential complaint users by combining B-domain economic data such as the user's traffic usage, ARPU value (average revenue per user), and the number of months the user has been online.

Algorithm for analyzing and identifying potential high-probability and high-value complaint users:

1. Convert the indicators of user voice quality, such as retention, mobility, access, delay, registration and quality, as well as the indicators of the three Internet sub-services, into corresponding scores.

The total score of potential complaint users = user voice perception score + Internet perception score.

In the formula, w1~w9 are the weights of various indicators, K1 is the proportion of voice service duration of the user in a certain time period to the total service duration, and K2 is the proportion of Internet service duration of the user in a certain time period to the total service duration.

2. Arrange the user perception scores obtained in high-probability scenarios in order from low to high. The top 20% of users are high-probability complaint users.

3. The B-domain economic data of access users is used to distinguish the value of potential complaint users based on their traffic usage, online time, ARPU and DOU. The following is the B-domain economic data of access. APRPU, DOU, customer star rating, and customer online months are selected as indicators to measure the importance of customers, and the total customer traffic is arranged from high to low to identify high-value users, as shown in the following table:

The calculation method of customer importance is shown in the following table:

Among them, value score > 75, importance level one; 60 < score <= 75, importance level two; 45 < score <= 60, importance level three; score <= 45, importance level four.

4. Finally, high-risk and high-value customers in high-probability complaint scenarios are output;

The data result table is shown in the following table:

Step 500: Complaint warning empowers online customer service support;

The high-risk area complaint user results of step 400 are converted into information resources for front-end customer service support. By inputting the user location/user number, the user scenario and user network perception status can be identified with one click, current and historical network problems can be intelligently located, explanations can be automatically generated, customer demands can be quickly responded to, and customer network experience can be improved.

The predicted impact range of the high complaint volume area (i.e. the name of the property point), the predicted number of cases, the cause of the problem and the explanation will be pushed to the operator to reduce the number of complaints recorded.

Push rules: Set the predicted complaint threshold based on the POI scenario and area size, and push to areas with a daily predicted complaint volume of ≥5. For critical scenarios, push as long as complaints are predicted; for other scenarios, set the threshold based on the size of the POI area.

Push method: The complaint prediction results will be automatically pushed to the complaint forwarding system. If the prevention is successful, the push of prediction information will be closed.

It should be noted that the execution subject of the network quality detection method provided in the embodiment of the present application may be a network quality detection device, or a control module in the network quality detection device for executing the method of network quality detection. In the embodiment of the present application, the method of executing network quality detection by the network quality detection device is taken as an example to illustrate the network quality detection device provided in the embodiment of the present application.

According to a third aspect of an embodiment of the present application, a network quality detection device is provided, which may include:

An acquisition device, used to respond to a user's network quality complaint request and acquire the user's target data information, the target data information being used to locate the user;

The prediction module is used to input data information into the network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on time factors, regional scenarios and network factors that affect the network status.

According to a fourth aspect of an embodiment of the present application, a target user determination device is provided, and the device may include:

A regional prediction module, used to predict a region where a probability of a network problem occurring exceeds a preset value by using the network status prediction model of the first aspect of the embodiment;

The collection module is used to collect network-side big data of users in areas with high risk of network problems;

An extraction module is used to extract voice performance indicators and Internet quality indicators from the big data on the network side;

The determination module is used to determine the target user based on the voice performance index and the Internet quality index, combined with the user's traffic usage, the user's average income and the B-domain economic data.

The network quality detection device in the embodiment of the present application can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. Exemplarily, the mobile electronic device can be a mobile phone, a tablet computer, a laptop computer, a PDA, an in-vehicle electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc. The non-mobile electronic device can be a server, a network attached storage (NAS), a personal computer (PC), a television (TV), a teller machine or a self-service machine, etc., which is not specifically limited in the embodiment of the present application.

The network quality detection device in the embodiment of the present application may be a device having an operating system. The operating system may be an Android operating system, an iOS operating system, or other possible operating systems, which are not specifically limited in the embodiment of the present application.

The network quality detection device provided in the embodiment of the present application can implement each process implemented by the method embodiments of Figures 1 to 10, and will not be described again here to avoid repetition.

Optionally, as shown in Figure 11, an embodiment of the present application also provides an electronic device 1100, including a processor 1101, a memory 1102, and a program or instruction stored in the memory 1102 and executable on the processor 1101. When the program or instruction is executed by the processor 1101, each process of the above-mentioned network quality detection method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

FIG. 12 is a schematic diagram of the hardware structure of an electronic device implementing an embodiment of the present application.

The hardware 1200 of the electronic device includes but is not limited to: a radio frequency unit 1201, a network module 1202, an audio output unit 1203, an input unit 1204, a sensor 1205, a display unit 1206, a user input unit 1207, an interface unit 1208, a memory 1209, and a processor 1210 and other components.

Those skilled in the art can understand that the hardware 1200 of the electronic device can also include a power supply (such as a battery) for supplying power to each component, and the power supply can be logically connected to the processor 1210 through a power management system, so that the power management system can manage charging, discharging, and power consumption management. The electronic device structure shown in FIG12 does not constitute a limitation on the electronic device, and the electronic device can include more or fewer components than shown in the figure, or combine certain components, or arrange components differently, which will not be described in detail here.

Among them, processor 1210 is used to respond to the user's network quality complaint request, obtain the user's target data information, and the target data information is used to locate the user; input the data information into the network status prediction model to obtain the user's network problem. The network status prediction model is a model that predicts the user's network status based on time factors, regional scenarios and network factors that affect the network status.

The hardware of the electronic device in the above-mentioned embodiment obtains the user's target data information, and the target data information is used to locate the user; the data information is input into the network status prediction model to obtain the user's network problem. The network status prediction model used by this method comprehensively considers the time factors, regional scenarios and network factors that affect the network status, and the factors that affect the network problem are considered more comprehensively. Therefore, the model predicts network problems more accurately, and then predicts the complaining users according to the network status, which can solve network complaint problems in a more targeted manner.

It should be understood that in the embodiment of the present application, the input unit 1204 may include a graphics processor (Graphics Processing Unit, GPU) 12041 and a microphone 12042, and the graphics processor 12041 processes the image data of the static picture or video obtained by the image capture device (such as a camera) in the video capture mode or the image capture mode. The display unit 1206 may include a display panel 12061, and the display panel 12061 may be configured in the form of a liquid crystal display, an organic light emitting diode, etc. The user input unit 1207 includes a touch panel 12071 and other input devices 12072. The touch panel 12071 is also called a touch screen. The touch panel 12071 may include two parts: a touch detection device and a touch controller. Other input devices 12072 may include, but are not limited to, a physical keyboard, a function key (such as a volume control button, a switch button, etc.), a trackball, a mouse, and a joystick, which will not be repeated here. The memory 1209 can be used to store software programs and various data, including but not limited to applications and operating systems. The processor 1210 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the modem processor may not be integrated into the processor 1210.

An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of the above-mentioned network quality detection method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

The processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned network quality detection method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises one..." does not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be noted that the scope of the method and device in the embodiment of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM/RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims (12)

1. A network quality detection method, characterized by comprising: In response to a network quality complaint request from a user, obtaining target data information of the user, wherein the target data information is used to locate the user; Inputting the target data information into a network status prediction model to obtain the network problem of the user, wherein the network status prediction model is a model for predicting the network status of the user based on time factors, regional scenarios and network factors that affect the network status; The time factors include long-term trend factors, seasonal variation factors, periodic cycle factors and irregular variation factors; The network factors include KPI, KQI, network environment, abnormal events, user voice quality indicators, Internet perception indicators, traffic usage, Internet time, ARPU and DOU; The regional scenario includes the scenario type, scenario area, number of scenario cells, amount of service-affecting alarms, and duration of alarms. 2. The method according to claim 1 is characterized in that the target data information is user location information and/or user telephone number information. 3. The method according to claim 1, characterized in that before obtaining the user's target data information in response to the user's network quality complaint request, the method further comprises: Obtain network complaint data at daily granularity; Extracting quantifiable scenario complaint evaluation index data from the network complaint data to obtain a complaint evaluation index set; Delete the holiday data in the complaint evaluation index set, and mark the complaint evaluation index set according to the time series characteristics to obtain sample data; The sample data is used to train the machine learning model to obtain the network status prediction model. 4. The method according to claim 3 is characterized in that extracting quantifiable scenario complaint evaluation index data from the network complaint data to obtain a complaint evaluation index set includes: Normalizing the abnormal data in the network complaint data to obtain regular indicator data; The quantifiable scenario complaint evaluation index data is extracted from the conventional index data to obtain a complaint evaluation index set. 5. The method according to claim 3 is characterized in that the quantifiable scene complaint evaluation index data includes: KQI, KPI, AOI scene, environmental index, abnormal index and soft collection index. 6. The method according to claim 3, characterized in that the using the sample data to train and test the machine learning model to obtain the network status prediction model comprises: Performing feature enhancement processing on the sample data to obtain enhanced sample data; The enhanced sample data is used to train and test the machine learning model to obtain the network status prediction model. 7. The method according to any one of claims 1 to 6, characterized in that after inputting the data information into a network status prediction model to obtain the network problem of the user, the method further comprises: generating an explanation of the appeal based on the network problem; The request explanation is sent to the user. 8. A method for determining a target user, comprising: Using the network status prediction model described in any one of claims 1 to 6 to predict an area where the probability of network problems exceeding a preset value; Collecting network-side big data of users in the high-risk area of network problems; Extracting voice performance indicators and Internet quality indicators from the network side big data; The target user is determined according to the voice performance index and the Internet access quality index, and in combination with the user's traffic usage, the user's average income and the B-domain economic data. 9. A network quality detection device, comprising: An acquisition device, used to acquire target data information of a user in response to a network quality complaint request of the user, wherein the target data information is used to locate the user; A prediction module, used for inputting the data information into a network status prediction model to obtain the network problem of the user, wherein the network status prediction model is a model for predicting the network status of the user based on time factors, regional scenarios and network factors that affect the network status; The time factors include long-term trend factors, seasonal variation factors, periodic cycle factors and irregular variation factors; The network factors include KPI, KQI, network environment, abnormal events, user voice quality indicators, Internet perception indicators, traffic usage, Internet time, ARPU and DOU; The regional scenario includes the scenario type, scenario area, number of scenario cells, amount of service-affecting alarms, and duration of alarms. 10. A target user determination device, comprising: A regional prediction module, used to use the network status prediction model according to any one of claims 1 to 6 to predict the area where the probability of network problems exceeding a preset value; A collection module, used to collect network-side big data of users in the area with high risk of network problems; An extraction module, used to extract voice performance indicators and Internet quality indicators from the network side big data; The determination module is used to determine the target user according to the voice performance index and the Internet access quality index, and in combination with the user's traffic usage, the user's average income and the B-domain economic data. 11. An electronic device, characterized in that it comprises: a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed by the processor, the steps of the network quality detection method according to any one of claims 1 to 7 or the steps of determining a target user according to claim 8 are implemented. 12. A readable storage medium, characterized in that the readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the network quality detection method according to any one of claims 1 to 7 or the steps of target user determination according to claim 8 are implemented.