Data analysis method, device, equipment and storage medium

Abstract

The present application discloses a method, device, equipment and storage medium for data analysis. Specifically, it includes: obtaining the KQI of the target service and the wireless side characteristic information of the target cell; using the first algorithm to calculate the correlation between the wireless side characteristic information and the quality difference characteristic information of the KQI, and sorting to obtain the first sorting of the wireless side characteristic information; using the second algorithm and the third algorithm to respectively calculate the importance of the wireless side characteristic information and the quality difference characteristic information of the KQI, and sorting them respectively to obtain the second sorting and third sorting of the wireless side characteristic information; combining the first sorting, the second sorting and the third sorting to obtain the wireless side characteristic information that is strongly correlated with the KQI quality difference. According to the embodiment of the present application, the main cause of the wireless side when the KQI indicator is of poor quality can be identified more accurately and efficiently.

Description

Technical Field

[0001] The present application belongs to the field of computer technology, and in particular, relates to a data analysis method, device, equipment and computer storage medium. Background Art

[0002] Generally, in wireless communication networks, the user experience of services running on applications is often expressed in terms of key quality indicators (KQI). For telecom operators, when the KQI of some services in the communication service network deteriorates, they hope to quickly and accurately locate the main wireless factors affecting the KQI indicators in order to optimize the network accordingly.

[0003] In the related art, the method of determining the main cause characteristics of the wireless side is usually to analyze and determine the main cause characteristics of the wireless side that affect the KQI index based on the long-term accumulated wireless network optimization experience. The method of selecting and analyzing the main cause characteristics of the wireless side in the related art has certain limitations, and the accuracy of the analysis results is not high. Summary of the invention

[0004] The embodiments of the present application provide a data analysis method, apparatus, device and computer storage medium, which can analyze the wireless side feature information that affects KQI by combining multiple algorithms, thereby more accurately and efficiently identifying the main wireless side cause of poor KQI indicator quality.

[0005] In a first aspect, an embodiment of the present application provides a method for data analysis, the method comprising:

[0006] Obtain the key quality indicator KQI of the target service and the wireless side characteristic information of the target cell;

[0007] Using a first algorithm, calculating the correlation between the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a first sorting of the wireless side feature information;

[0008] Using the second algorithm and the third algorithm, respectively calculating the importance of the wireless side feature information and the quality difference feature information of the KQI, and sorting them respectively to obtain a second sorting and a third sorting of the wireless side feature information;

[0009] The first sorting, the second sorting and the third sorting are combined and sorted to obtain wireless side feature information that is strongly correlated with the KQI quality difference.

[0010] Optionally, using a first algorithm to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a first sorting of the wireless side feature information, including:

[0011] According to the type of the target cell, the correlation between the wireless side characteristic information and the quality difference characteristic information of the KQI is calculated;

[0012] When the correlation satisfies a preset feature importance condition, the wireless side feature information corresponding to the correlation is sorted to obtain a first sort.

[0013] Optionally, a second algorithm is used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sort them to obtain a second sorting of the wireless side feature information, including:

[0014] Calculating, according to the quality difference feature information of the KQI, the first importance of the wireless side feature information corresponding to the quality difference feature information of the KQI;

[0015] The wireless side feature information is sorted according to the first importance to obtain a second sort.

[0016] Optionally, a third algorithm is used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sort them to obtain a third sorting of the wireless side feature information, including:

[0017] Calculate the evidence weight of the wireless side feature information corresponding to the KQI quality difference feature information according to the KQI quality difference feature information;

[0018] Calculate the second importance based on the weight of evidence;

[0019] When the second importance meets the preset threshold condition, the wireless side feature information is sorted according to the second importance to obtain a third sort.

[0020] Optionally, the first sorting, the second sorting, and the third sorting are combined and sorted to obtain wireless side feature information that is strongly related to the KQI quality difference, including:

[0021] Convert the rankings corresponding to the first ranking, the second ranking, and the third ranking into scores respectively;

[0022] Calculate the average values ​​of the scores corresponding to the first ranking, the second ranking, and the third ranking respectively;

[0023] According to the preset weight value of each average value, the first sorting, the second sorting and the third sorting are combined and sorted to obtain a combined sorting result;

[0024] According to the sorting results, wireless side feature information strongly related to the KQI quality difference is obtained.

[0025] Optionally, obtaining a key quality indicator KQI of a target service and radio side characteristic information of a target cell includes:

[0026] The KQI and wireless side feature information are preprocessed using a preset business logic algorithm to obtain filtered KQI and wireless side feature information.

[0027] Optionally, the second algorithm includes a CatBoost algorithm, and the third algorithm includes an optimal iterative binning algorithm.

[0028] In a second aspect, an embodiment of the present application provides a data analysis device, the device comprising:

[0029] An acquisition module, used to acquire the key quality indicator KQI of the target service and the wireless side characteristic information of the target cell;

[0030] A first calculation module, configured to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI by using a first algorithm, and sort the wireless side feature information to obtain a first sorting of the wireless side feature information;

[0031] A second calculation module is used to use a second algorithm and a third algorithm to respectively calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and to sort them respectively to obtain a second sorting and a third sorting of the wireless side feature information;

[0032] The sorting module is used to combine and sort the first sorting, the second sorting and the third sorting to obtain wireless side feature information that is strongly related to the KQI quality difference.

[0033] In a third aspect, an embodiment of the present application provides a data analysis device, the device comprising:

[0034] a processor and a memory storing computer program instructions;

[0035] When the processor executes the computer program instructions, the method for data analysis as described in the first aspect and the optional method of the first aspect is implemented.

[0036] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon, and when the computer program instructions are executed by a processor, a method for data analysis as in the first aspect and as an optional aspect of the first aspect is implemented.

[0037] The data analysis method, device, equipment and computer storage medium of the embodiment of the present application can use the first algorithm to calculate the correlation between the acquired wireless side feature information and the quality difference feature information of the KQI, and determine the first ranking. The second algorithm and the third algorithm are used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, respectively, and determine the second ranking and the third ranking, respectively. Then, the three rankings are combined to determine the wireless side feature information that is strongly correlated with the KQI quality difference. Therefore, based on a variety of different algorithms, the correlation between the wireless side feature information and the quality difference feature information of the KQI is analyzed, and the wireless side feature information that may cause the KQI quality difference can be predicted more accurately and efficiently, that is, the accuracy and efficiency of the main cause of the wireless side when determining the KQI indicator quality difference is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0039] Figure 1 is a flowchart of a data analysis method provided by some embodiments of the present application;

[0040] Figure 2 is a flowchart of a data analysis method provided by other embodiments of the present application;

[0041] Figure 3 is a schematic diagram of the structure of a data analysis device provided in some other embodiments of the present application;

[0042] Figure 4 This is a schematic diagram of the hardware structure of a data analysis device provided in some embodiments of the present application. DETAILED DESCRIPTION

[0043] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present application and are not configured to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0044] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" 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, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0045] In wireless communication networks, KQI usually includes the percentage of poor WeChat quality, the percentage of video freezes, the percentage of game freezes, the delay when opening a web page, etc. The main factors affecting KQI on the wireless side include, for example, failure, interference, utilization of physical resource blocks, number of users, measurement report (MR) coverage, and the percentage of channel quality indication (CQI).

[0046] The existing selection of main wireless-side factors is done manually based on long-term experience in wireless network optimization, listing the main wireless-side factors that affect KQI. Incomplete consideration of manual selection may lead to misselection, underselection, or failure to select features. Inexperienced analysts may even miss important features. Therefore, the main wireless-side factors that cause poor KQI indicators cannot be accurately identified.

[0047] In order to solve the problems of the prior art, the embodiments of the present application provide a data analysis method, device, equipment and computer storage medium, which can analyze the correlation between the wireless side feature information and the quality difference feature information of the KQI based on a variety of different algorithms, and can more accurately and efficiently predict the wireless side feature information that may cause the KQI quality difference, thereby improving the accuracy and efficiency of determining the main cause of the wireless side when determining the KQI indicator quality difference.

[0048] The following describes the data analysis method, device, equipment and computer storage medium provided according to the embodiments of the present application in conjunction with the accompanying drawings. It should be noted that these embodiments are not intended to limit the scope of the present application.

[0049] First, the data analysis method provided in the embodiments of the present application is introduced.

[0050] Figure 1 FIG. 1 is a flow chart of a data analysis method provided by an embodiment of the present application. Figure 1 As shown, in the application embodiment, the data analysis method may include the following steps:

[0051] S101: Acquire the KQI of the target service and the radio side characteristic information of the target cell.

[0052] Here, multiple KQIs of the target service are obtained. The wireless side feature information may include the key performance indicators (KPI) of the wireless side of the communication equipment manufacturer, MR statistical indicators, and the communication data collected within a predetermined time period of the target cell.

[0053] For example, for the selected data volume, the data of 120,000 cells in the entire network for one week can be collected, and the frequency of data collection is one record per cell per hour. The total amount of all data is 120,000*10*24=28,800,000.

[0054] S102: Using a first algorithm, calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI, and sort them to obtain a first sorting of the wireless side feature information.

[0055] In some embodiments of the present application, the first algorithm may include a correlation algorithm. Through the correlation algorithm, the correlation between the wireless side feature information and the quality difference feature information of the KQI is calculated to determine the wireless side feature information related to the KQI quality difference.

[0056] The poor quality feature information of KQI may include poor quality and non-poor quality. Poor quality of KQI means that the proportion of poor quality of KQI is greater than 5%.

[0057] The first sorting may be a sorted list obtained by sorting the relevance from high to low.

[0058] S103: using the second algorithm and the third algorithm to respectively calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sorting them respectively to obtain the second sorting and the third sorting of the wireless side feature information.

[0059] In some embodiments of the present application, the second algorithm may include a supervised model algorithm such as a CatBoost algorithm, and the third algorithm may include a supervised model algorithm such as an optimal iterative binning algorithm.

[0060] The second sorting and the third sorting can be sorted lists obtained by sorting the importance from high to low.

[0061] S104: Combining and sorting the first sorting, the second sorting, and the third sorting to obtain wireless side feature information that is strongly correlated with the KQI quality difference.

[0062] Here, the first sorting, the second sorting and the third sorting may be a sorting list including the quality difference characteristic information of the KQI and the corresponding wireless side characteristic information.

[0063] In summary, the data analysis method of the embodiment of the present application can use the first algorithm to calculate the correlation between the acquired wireless side feature information and the quality difference feature information of the KQI, and determine the first ranking. The second algorithm and the third algorithm are used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, respectively, and determine the second ranking and the third ranking, respectively. Then, the three rankings are combined to determine the wireless side feature information that is strongly correlated with the KQI quality difference. Therefore, based on a variety of different algorithms, the correlation between the wireless side feature information and the quality difference feature information of the KQI is analyzed, and the wireless side feature information that may cause the KQI quality difference can be predicted more accurately and efficiently, that is, the accuracy and efficiency of the main cause of the wireless side when determining the KQI indicator quality difference is improved.

[0064] In order to better illustrate the data analysis method of the embodiment of the present application, the implementation of the method is described in detail below. Figure 2 As shown, Figure 2 : is a flow chart of a data analysis method provided by another embodiment of the present application. The method can be specifically implemented as follows:

[0065] S201: Acquire the KQI of the target service and the radio side characteristic information of the target cell.

[0066] In some embodiments of the present application, various collected raw data need to be preprocessed, which may include data cleaning, removal and replacement of outliers, and interpolation of missing values.

[0067] In some embodiments of the present application, data cleaning may include removing or replacing "dirty data" from various collected raw data. For example, for special characters such as "#N / A", "#VALUE!", "NIL", " / 0" and so on in numerical data, these characters can be set to empty. For the problem of inconsistent dimensions of proportion indicators, in order to facilitate calculation, the dimensions of these indicators can be unified between 0 and 1 and normalized.

[0068] In some embodiments of the present application, an outlier, also known as an outlier, refers to an individual value in a sample whose value is significantly deviated from the majority of observed values ​​of the sample to which it belongs. Outlier processing can be done by combining physical discrimination with model recognition. Exemplarily, based on exploratory analysis, check whether there is data with an obviously abnormal value range, such as a connection rate of -3.5%, and remove and replace values ​​based on physical discrimination. The mean three times standard deviation algorithm, the box plot method, and the skewness-based algorithm can also be used to remove and replace outliers.

[0069] In some embodiments of the present application, missing value processing may include using mean interpolation, Hot Dec interpolation, regression interpolation and multiple interpolation to process missing values. In addition, based on missing rate analysis and actual needs, some high missing rate data may be considered for deletion.

[0070] It is understandable that the method of preprocessing the corresponding data may not be limited here. In practical applications, a suitable data preprocessing method may be selected according to specific needs.

[0071] S202: Preprocess the KQI and wireless side feature information using a preset business logic algorithm to obtain filtered KQI and wireless side feature information.

[0072] In some embodiments of the present application, the key quality indicator KQI and the wireless side characteristic information may be pre-screened using a preset business logic algorithm to obtain the screened key quality indicator KQI and the wireless side characteristic information.

[0073] Here, the preset service logic algorithm may be a service logic analysis algorithm based on expert experience, and wireless side feature information that may have an impact on the KQI is found according to the service logic.

[0074] For example, for the target service KQI quality difference as WeChat quality difference, 20 wireless side feature information are analyzed based on the business logic of expert experience, including 6 features of capacity, 5 features of quality, 4 features of delay, and 5 features of coverage.

[0075] S203: Calculate the correlation between the wireless side characteristic information and the quality difference characteristic information of the KQI according to the type of the target cell.

[0076] S204: When the relevance satisfies a preset feature importance condition, the wireless side feature information is sorted to obtain a first sort.

[0077] Here, the preset feature importance condition may be a representation value of the feature importance determined according to the type of the target cell and the corresponding relevance of the target cell. The wireless side feature information is sorted according to the preset feature importance condition.

[0078] In some embodiments of the present application, the Pearson correlation coefficient is used to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI and the autocorrelation between each wireless side feature information. The Pearson correlation coefficient calculation formula (1) is as follows:

[0079]

[0080] Among them, r represents the Pearson correlation coefficient, X and Y represent the quality difference feature information of the wireless side and KQI respectively. It is the average value of the wireless side characteristic information and the average value of the quality difference characteristic information of KQI.

[0081] In some embodiments of the present application, since the wireless side feature information X with high correlation with the quality difference feature information Y of the KQI may be different in different cells, if it is directly calculated, there will be a problem that the same X in different cells cancels each other out. In order to improve the accuracy of the calculation and facilitate the feature correlation ranking, the algorithm process for calculating the correlation between the wireless side feature information and the quality difference feature information of the KQI and sorting them is as follows:

[0082] First, the target cells are grouped and the correlation between the characteristic information of each radio side of each cell and the quality difference characteristic information of KQI is calculated respectively;

[0083] Secondly, the proportion of cells with an absolute value of correlation greater than 0.3 is calculated as a measure of the “importance” between the wireless side feature information and the KQI quality difference feature information. The calculation formula (2) is as follows:

[0084] IMC(i)=count(ABS(r)>0.3) / count(NOTNULL(r)) (2)

[0085] IMC(i) represents the proportion of cells whose absolute value of the i-th feature correlation is greater than 0.3, count represents the count, ABS is the absolute value, r is the correlation coefficient, and NOTNULL represents not null.

[0086] The wireless side feature information is ranked according to IMC. In addition, before sorting, features with IMC values ​​less than 0.01 can be eliminated according to the conservative elimination method.

[0087] In some embodiments of the present application, multicollinearity refers to the fact that the model estimation is distorted or difficult to estimate accurately due to the existence of a high correlation between the explanatory variables (X), and should be eliminated. Therefore, in order to reduce the adverse effects of the high correlation on the prediction accuracy of the algorithm, the autocorrelation between the feature information of each wireless side and the autocorrelation between the quality difference feature information of KQI are calculated to form a correlation coefficient matrix, mark the strongly correlated features, and eliminate the feature information with strong correlation.

[0088] S205: Calculate the first importance of the wireless side feature information corresponding to the quality difference feature information of the KQI according to the quality difference feature information of the KQI.

[0089] S206: Sort the wireless side feature information according to the first importance to obtain a second sort.

[0090] In some embodiments of the present application, the second algorithm is used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and the execution process of sorting the wireless side feature information to obtain the second sorting may include S205 to S206.

[0091] In some embodiments of the present application, the second algorithm may be a CatBoost algorithm. The CatBoost algorithm is used to calculate the first importance of the wireless side feature information and the quality difference feature information of the KQI, and sort them to determine the second sort of the wireless side feature information.

[0092] S207: Calculate the evidence weight of the wireless side feature information corresponding to the quality difference feature information of the KQI according to the quality difference feature information of the KQI.

[0093] S208: Calculate the second importance according to the weight of evidence.

[0094] S209: When the second importance meets a preset threshold condition, the wireless side feature information is sorted according to the second importance to obtain a third sort.

[0095] In some embodiments of the present application, a third algorithm is used to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and the execution process of sorting the wireless side feature information to obtain the third sorting may include S207 to S209.

[0096] In some embodiments of the present application, the third algorithm may be an optimal iterative binning algorithm. The implementation of the optimal iterative binning algorithm includes the following process:

[0097] First, the bins are initialized. Initialization is performed based on the minimum number of samples per class (min_per), and the maximum number of bins (max_bins) for initialization is obtained. The calculation formula (3) is as follows:

[0098] max_bins=trunc(1 / min_per) (3)

[0099] Among them, trunc means rounding.

[0100] Secondly, determine the initial split point. The initial split point of each feature is calculated using the maximum binning sequence quotient as the quantile. The calculation formula (4) is as follows:

[0101] cutp=quantile[df,(0:max_bins) / max_bins] (4)

[0102] Among them, quantile is the quantile function, df is the feature, and 0:max_bins represents an integer sequence from 0 to max_bins.

[0103] Again, merging of sparse target classes. If the percentage of a target class column in a bin is below this limit, for example below 0.1%, the corresponding bin is joined with other bins, and in the case of numerical variables, adjacent predictor classes are merged.

[0104] Finally, iterative binning and merging are performed. Iterative binning and merging include initial merging and final merging.

[0105] Initial merge. Merge boxes with similar weight of evidence (WOE) and gradually calculate the corresponding WOE and information value (IV). Here, information value IV represents the second importance. WOE represents the difference between "the proportion of poor quality samples in the current group to all poor quality samples" and "the proportion of non-poor quality samples in the current group to all non-poor quality samples". The larger the WOE, the greater the difference.

[0106] The calculation formula (5) of WOE is as follows:

[0107]

[0108] Among them, py i is the proportion of the response (poor quality) samples in this group to all the response samples, pn i is the proportion of all non-responding samples in this group of non-responding (non-poor quality) samples, #y i is the number of responses in this group, #n i is the number of non-responses in this group, #y T is the number of all responses in the sample, #n T is the number of all non-responses in the sample.

[0109] The calculation formula (6) of IV is as follows:

[0110]

[0111] Based on the IV of each group of a feature variable, the IV of the entire feature variable can be calculated using the summation formula (7):

[0112]

[0113] Finally merge. The nearby boxes with the most similar evidence weights are gradually connected until the IV decreases by more than the preset percentage value threshold or the specified value, then the merge is stopped. The hyperparameters of the pre-set algorithm model are iterated cyclically until the IV of each feature reaches the maximum when the above stop merging conditions are met. The information value IV is used as the second importance, that is, the feature importance. According to the second importance, the feature information is sorted to obtain the third sort.

[0114] In addition, in some embodiments of the present application, the optimal iterative binning algorithm also includes the processing of extreme values.

[0115] For example, if the crosstab with the target class bin contains a frequency of 0, then adjust to be able to calculate WOE and IV, add the offset ρ to each column percentage and then recalculate. If extreme WOE occurs, merge it with bins with sparse target classes.

[0116] Here, the IV calculated by the optimal iterative binning is used as the feature importance, which can not only sort according to the size of IV, but also divide the importance level of IV. This algorithm can bin continuous variables and type variables at the same time, which is powerful and more interpretable.

[0117] S210: Convert the rankings corresponding to the first ranking, the second ranking and the third ranking into scores respectively.

[0118] S211: Calculate the average values ​​of the scores corresponding to the first ranking, the second ranking, and the third ranking respectively.

[0119] S212: Sorting the first sorting, the second sorting and the third sorting in combination according to the preset weight value of each average value to obtain a combined sorting result.

[0120] S213: According to the sorting result, wireless side feature information strongly related to the KQI quality difference is obtained.

[0121] In some embodiments of the present application, after obtaining the first ranking, the second ranking and the third ranking of the wireless side feature information, a combined ranking algorithm can be used to perform a comprehensive ranking to select the wireless side feature information that is strongly correlated with the KQI quality difference.

[0122] In some embodiments of the present application, the first ranking, the second ranking and the third ranking are determined by different algorithmic processes, which can be regarded as three rankings obtained by three different ranking methods. First, the ranking ranking obtained by each ranking method can be converted into a score using a ranking scoring method, and the calculation formula (8) is as follows:

[0123] R k =nr k +1 (8)

[0124] R k Indicates the rating, r k It represents the ranking under the k-th sorting method, i = 1, 2, ... n, k = 1, 2, ... p, that is, the first place gets n points, ..., the nth place gets 1 point, and the kth place gets n-k+1 points. If there are identical rankings, the score can be taken as the average of the scores of these positions.

[0125] Then, the average score of each ranking method is calculated, and the calculation formula (9) is as follows:

[0126]

[0127] The average of the scores from each ranking method was re-ranked by average.

[0128] In some embodiments of the present application, since the importance of different sorting methods is different, the weight values ​​corresponding to different sorting methods can be predetermined. According to the order of the preset weight values ​​from large to small, the average values ​​of the scores of each sorting method are sorted. Then, according to the sorting result of the average value, the first sorting, the second sorting, and the third sorting are combined and sorted to obtain the combined sorting result. For example, the preset weight value corresponding to the first sorting is

[0129] In some embodiments of the present application, the weight values ​​corresponding to different sorting methods can be predetermined by using the Analytic Hierarchy Process (AHP).

[0130] In some embodiments of the present application, after determining the sorting result, the sorting result can also be corrected in combination with the feature missing rate and the business logic, including considering the prediction ability of the wireless side feature information, and whether the wireless side feature information has network optimization value, etc., and finally determining the wireless side feature information that is strongly related to the KQI quality difference. In this way, the accuracy of determining the wireless side main cause corresponding to the KQI quality difference can be improved.

[0131] To sum up, the data analysis method provided in the embodiment of the present application can analyze the correlation between the wireless side feature information and the quality difference feature information of the KQI based on a variety of different algorithms, and determine the feature importance relationship between the two features, so as to more accurately and efficiently predict the wireless side feature information that may cause the KQI quality difference, and improve the accuracy and efficiency of determining the main cause of the wireless side when determining the KQI indicator quality difference.

[0132] Based on the data analysis method provided in the above embodiment, the present application also provides a specific implementation of a data analysis device. Please refer to the following embodiment.

[0133] like Figure 3 As shown, Figure 3 : is a schematic diagram of the structure of a data analysis device provided by another embodiment of the present application. In the embodiment of the present application, the data analysis device may include:

[0134] The acquisition module 301 is used to acquire the key quality indicator KQI of the target service and the radio side characteristic information of the target cell.

[0135] The first calculation module 302 is used to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI by using a first algorithm, and sort them to obtain a first sorting of the wireless side feature information.

[0136] The second calculation module 303 is used to use the second algorithm and the third algorithm to respectively calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sort them respectively to obtain the second sorting and the third sorting of the wireless side feature information.

[0137] The sorting module 304 is used to combine and sort the first sorting, the second sorting and the third sorting to obtain wireless side feature information that is strongly related to the KQI quality difference.

[0138] Figure 3 Each module / unit in the device shown has the function of realizing Figure 1 and 2 The functions of each step in the process can achieve the corresponding technical effects, which will not be described in detail here for the sake of brevity.

[0139] To summarize, the data analysis device provided in the embodiments of the present application can be used to execute the data analysis method described in the above embodiments. The method can analyze the correlation between the wireless side feature information and the quality difference feature information of the KQI based on a variety of different algorithms, and determine the feature importance relationship between the two features, so as to more accurately and efficiently predict the wireless side feature information that may cause the KQI quality difference, and improve the accuracy and efficiency of determining the main cause of the wireless side when determining the KQI indicator quality difference.

[0140] Based on the data analysis method provided in the above embodiment, the present application also provides a specific implementation of a data analysis device. Please refer to the following embodiment.

[0141] Figure 4 This is a schematic diagram of the hardware structure of a data analysis device provided in some embodiments of the present application.

[0142] The data analysis device may include a processor 401 and a memory 402 storing computer program instructions.

[0143] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0144] The memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 402 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 402 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 402 is a non-volatile solid-state memory. In a specific embodiment, the memory 402 includes a read-only memory (ROM). In appropriate cases, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM) or a flash memory or a combination of two or more of these.

[0145] The processor 401 implements any one of the data analysis methods in the above embodiments by reading and executing computer program instructions stored in the memory 402 .

[0146] In one example, the data analysis device may further include a communication interface 403 and a bus 410. Figure 4 As shown, the processor 401, the memory 402, and the communication interface 403 are connected via a bus 410 and communicate with each other.

[0147] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0148] Bus 410 includes hardware, software or both, and the parts of the equipment of data analysis are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0149] The data analysis device can execute the data analysis method in the embodiment of the present application, thereby realizing the combination Figure 1 and Figure 2 Describe the methods of data analysis.

[0150] In addition, in combination with the data analysis method in the above embodiment, the embodiment of the present application can provide a computer storage medium to implement. The computer storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the data analysis methods in the above embodiment is implemented.

[0151] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0152] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in 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, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0153] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. 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 embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0154] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and computer program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram 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 data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. 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 can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0155] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A method for data analysis, characterized in that: include: Obtain the key quality indicator KQI of the target service and the wireless side characteristic information of the target cell; Using a first algorithm, calculating the correlation between the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a first sorting of the wireless side feature information; Using the second algorithm and the third algorithm, respectively calculating the importance of the wireless side feature information and the quality difference feature information of the KQI, and sorting them respectively to obtain a second sorting and a third sorting of the wireless side feature information; The first sorting, the second sorting and the third sorting are combined and sorted to obtain wireless side feature information that is strongly correlated with the KQI quality difference.

2. The method according to claim 1, characterized in that The adopting of the first algorithm to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a first sorting of the wireless side feature information, includes: Calculating the correlation between the wireless side characteristic information and the quality difference characteristic information of the KQI according to the type of the target cell; When the correlation satisfies a preset feature importance condition, the wireless side feature information corresponding to the correlation is sorted to obtain a first sort.

3. The method according to claim 1, characterized in that The adopting of the second algorithm to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a second sorting of the wireless side feature information, includes: Calculating, according to the quality difference feature information of the KQI, a first importance of the wireless side feature information corresponding to the quality difference feature information of the KQI; The wireless side feature information is sorted according to the first importance to obtain a second sort.

4. The method according to claim 1, characterized in that: The adopting of the third algorithm to calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sorting them to obtain a third sorting of the wireless side feature information includes: Calculating, according to the quality difference feature information of the KQI, the evidence weight of the wireless side feature information corresponding to the quality difference feature information of the KQI; Calculating a second importance according to the weight of evidence; When the second importance satisfies a preset threshold condition, the wireless side feature information is sorted according to the second importance to obtain a third sort.

5. The method according to claim 1, characterized in that The combining and sorting the first sorting, the second sorting, and the third sorting to obtain wireless side feature information strongly related to the KQI quality difference includes: Convert the rankings corresponding to the first ranking, the second ranking and the third ranking into scores respectively; Calculate the average values ​​of the scores corresponding to the first ranking, the second ranking, and the third ranking respectively; According to a preset weight value of each of the average values, the first sorting, the second sorting and the third sorting are combined and sorted to obtain a combined sorting result; According to the sorting result, wireless side feature information strongly related to the KQI quality difference is obtained.

6. The method according to any one of claims 1 to 5, characterized in that: The obtaining of the key quality indicator KQI of the target service and the radio side characteristic information of the target cell includes: The KQI and wireless side feature information are preprocessed using a preset business logic algorithm to obtain filtered KQI and wireless side feature information.

7. The method according to any one of claims 1 to 5, characterized in that: The second algorithm includes a CatBoost algorithm, and the third algorithm includes an optimal iterative binning algorithm.

8. A data analysis device, characterized in that: The device comprises: An acquisition module, used to acquire the key quality indicator KQI of the target service and the wireless side characteristic information of the target cell; A first calculation module, configured to calculate the correlation between the wireless side feature information and the quality difference feature information of the KQI by using a first algorithm, and sort the wireless side feature information to obtain a first sorting of the wireless side feature information; A second calculation module is used to use a second algorithm and a third algorithm to respectively calculate the importance of the wireless side feature information and the quality difference feature information of the KQI, and sort them respectively to obtain a second sorting and a third sorting of the wireless side feature information; The sorting module is used to combine and sort the first sorting, the second sorting and the third sorting to obtain wireless side feature information that is strongly related to the KQI quality difference.

9. A data analysis device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the method for data analysis according to any one of claims 1 to 7 is implemented.

10. A computer storage medium, characterized in that: The computer storage medium stores computer program instructions, which, when executed by a processor, implement the data analysis method according to any one of claims 1 to 7.

Images