CN114401494A - Short message issuing abnormity detection method and device, computer equipment and storage medium - Google Patents
Short message issuing abnormity detection method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN114401494A CN114401494A CN202210043421.7A CN202210043421A CN114401494A CN 114401494 A CN114401494 A CN 114401494A CN 202210043421 A CN202210043421 A CN 202210043421A CN 114401494 A CN114401494 A CN 114401494A
- Authority
- CN
- China
- Prior art keywords
- data
- short message
- model
- preset
- issuing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/12—Messaging; Mailboxes; Announcements
- H04W4/14—Short messaging services, e.g. short message services [SMS] or unstructured supplementary service data [USSD]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Debugging And Monitoring (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of operation process optimization, and discloses a short message issuing abnormity detection method, a device, computer equipment and a storage medium, wherein the method comprises the following steps: receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators; processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor; acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is outside the preset normal interval; and if the expected successful data is outside the predicted normal interval, sending an abnormal prompt. The invention can quickly detect whether the fluctuation of the feedback data is in the prediction normal interval, thereby timely judging whether the current short message pushing task is abnormal.
Description
Technical Field
The invention relates to the field of operation process optimization, in particular to a short message issuing abnormity detection method and device, computer equipment and a storage medium.
Background
Short message service is widely applied, and plays a very important role in the fields of OTP (One Time Password) verification, enterprise message notification and the like. The short message service has the advantages of low cost, high access efficiency, high conversion rate and the like.
However, based on the carrier of the mobile phone number of the operator, the short message service is easily affected by factors such as regions and operators, and the success rate of sending short messages fluctuates. Such fluctuations have a severe negative impact on some service events that are time-sensitive, which severely affects the conversion rate of the service event. Moreover, when a fluctuation event occurs, the problem is not easy to find and eliminate in time.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a method, an apparatus, a computer device and a storage medium for detecting an abnormal issue of a short message, and to detect whether the short message is issued abnormally in real time.
A short message issuing abnormity detection method comprises the following steps:
receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is outside the preset normal interval;
and if the expected successful data is outside the predicted normal interval, sending an abnormal prompt.
A short message issuing abnormity detection device comprises:
the feedback data receiving module is used for receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
the prediction expectation module is used for processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
the matching normal interval module is used for acquiring a preset normal interval matched with the short message pushing task and judging whether the expected successful data is outside the preset normal interval or not;
and the abnormal reminding module is used for sending an abnormal reminding if the expected successful data is outside the prediction normal interval.
A computer device comprises a memory, a processor and computer readable instructions stored in the memory and capable of running on the processor, wherein the processor executes the computer readable instructions to realize the short message issuing abnormity detection method.
One or more readable storage media storing computer-readable instructions which, when executed by one or more processors, cause the one or more processors to perform a short message delivery anomaly detection method as described above.
According to the short message issuing abnormity detection method, the short message issuing abnormity detection device, the computer equipment and the storage medium, the feedback data of the short message pushing task is received, and the feedback data comprises the short message issuing success data of a plurality of operators, so that the issuing condition of the short message is monitored in real time. Processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor, so that real-time variable feedback data is converted into expected success data with higher predictability, and the identification accuracy of abnormal conditions is improved. And acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is out of the preset normal interval so as to evaluate whether the expected successful data is abnormal or not through the preset normal interval. And if the expected successful data is outside the predicted normal interval, sending an abnormal prompt so that the staff can obtain the corresponding abnormal prompt in time. The invention can quickly detect whether the fluctuation of the feedback data is in the prediction normal interval, thereby timely judging whether the current short message pushing task is abnormal.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive labor.
Fig. 1 is a schematic view of an application environment of a short message issuance anomaly detection method according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a short message issuance anomaly detection method according to an embodiment of the present invention;
FIG. 3 is a plurality of prediction curves in an embodiment of the present invention;
FIG. 4 is a schematic diagram of a point distribution of a first image according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a point distribution of a second image according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of a point distribution of a third image according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of the point distribution of the abnormal point detection image according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of an apparatus for detecting an abnormality in short message delivery according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of a computer device according to an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The method for detecting an exception issued by a short message according to this embodiment can be applied to an application environment shown in fig. 1, in which a client communicates with a server. The client includes, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices. The server can be implemented by an independent server or a server cluster composed of a plurality of servers.
In an embodiment, as shown in fig. 2, a method for detecting an exception issued by a short message is provided, which is described by taking the method applied to the server in fig. 1 as an example, and includes the following steps S60-S90.
And S60, receiving feedback data of the short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators.
Understandably, the short message push task refers to a task of transmitting a short message containing specified contents to a target group of people. Corresponding specified content can be written based on specific service requirements, and then a corresponding short message push task is generated. The business requirements may be marketing requirements, campaign requirements, and the like. For example, a company needs to push certain activity information to one million customers on a certain day.
And submitting the short message pushing task to an operator. Then, during the task execution, feedback data of the short message push task is received from different operators in real time. Here, the feedback data includes short message delivery success data of several operators. The short message delivery success data may include the success number of short message delivery per minute. If the short message pushing task relates to multiple regions, the short message successful delivery data may include region identifiers for distinguishing successful delivery quantities of different regions.
S70, processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and dimension factors, wherein one dimension parameter corresponds to one dimension factor.
Understandably, the multivariate linear prediction model can be a multivariate linear prediction model obtained by training based on historical data, has a prediction function, and can predict the expected number of short messages to be sent. The multivariate linear prediction model comprises a constant term and a dimensionality factor. In one example, the multivariate linear prediction model can be represented as:
Y＝β0+β1X1+β2X2+β3X3+…+βnXn
wherein Y is a predicted value;
β0is a constant term;
X1represents the 1 st dimension parameter, and so on, XnRepresenting the nth dimension parameter, wherein n is the total number of the dimension parameters;
β1～βnare respectively and X1～XnThe corresponding dimensional factor.
And extracting the values of all the dimensional parameters from the feedback data, and substituting the values into the multivariate linear prediction model to obtain a predicted value. The expected success data includes predicted values for various points in time. For example, the push success number of the first minute is extracted from the feedback data to obtain the parameter value of each dimension parameter, and then the parameter value is substituted into the multi-element linear prediction model to obtain the predicted value Y of the first minute1(ii) a Extracting the pushing success quantity of the second minute from the feedback data to obtain the parameter value of each dimension parameter, substituting the parameter value into the multi-element linear prediction model to obtain the predicted value Y of the second minute2(ii) a And so on.
In the example of fig. 3, the expected success data may be represented using a predictor curve 02.
S80, acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected success data is outside the preset normal interval.
Understandably, the push population of each short message push task generally has a certain difference, and therefore, a plurality of historical push tasks which are most similar (even the same) to the push population of the current short message push task need to be searched, and no abnormal event occurs in the historical push tasks. Feedback data of the historical pushing tasks are processed through the multivariate linear prediction model, multiple historical expected success data can be obtained, and corresponding preset normal intervals can be drawn according to the historical expected success data. In the example of fig. 3, a highest success curve 01, a lowest success curve 04, and a fitted success curve 03 are included. The preset normal interval may refer to an interval between the highest success curve 01 and the fitted success curve 03.
And S90, if the expected success data is out of the prediction normal interval, sending an abnormal prompt.
Understandably, if the expected success data calculated currently is within the prediction normal interval, it indicates that the current short message is sent normally. If the expected success data calculated currently is outside the prediction normal interval, it is indicated that there may be a problem in the current short message transmission, and therefore, a corresponding abnormal prompt needs to be sent out. Here, exception reminders include, but are not limited to, pop-up window reminders, mail reminders, and short message reminders.
In one example, at a certain time point in a certain day marketing activity, the amount of the marketing short message success corresponding to a certain operator becomes low, as shown in table 1 below.
Table 1 example feedback data and expected success data
| Time | X1 | X2 | X3 | X4 | Y |
| 09:01:00 | 755.5 | 613.5 | 423.5 | 2213 | 2051.93015 |
| 09:02:00 | 754.5 | 615.5 | 424.5 | 2204 | 2047.40373 |
| 09:03:00 | 553.0 | 617.0 | 425.0 | 2195 | 1871.60428 |
| 09:04:00 | 557.5 | 613.0 | 426.5 | 2195 | 1872.32626 |
Wherein, X1、X2、X3Respectively a first operator, a second operator and a third operator, X4And Y is a predicted value. The first operator has a work load of 755.5 in a 09:01 short message.
As can be seen from Table 1, times 09:03 and 09:04, X1The number of short messages of the first operator decreases, which results in a smaller final predicted value, i.e., the value of the Y column. The difference between the predicted success amount and the expected success amount in the normal sample is too large, and the comparison with the normal value of the same region can reflect that the link of the first operator in the region has problems at present and needs to be adjusted.
In the steps S60-S90, feedback data of the short message push task is received, where the feedback data includes successful short message sending data of a plurality of operators, so as to monitor the sending status of the short message in real time. Processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor, so that real-time variable feedback data is converted into expected success data with higher predictability, and the identification accuracy of abnormal conditions is improved. And acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is out of the preset normal interval so as to evaluate whether the expected successful data is abnormal or not through the preset normal interval. And if the expected successful data is outside the predicted normal interval, sending an abnormal prompt so that the staff can obtain the corresponding abnormal prompt in time. The embodiment can quickly detect whether the fluctuation of the feedback data is in the prediction normal interval, and further judge whether the current short message pushing task is abnormal or not in time.
Optionally, before step S70, that is, before the feedback data is processed by the multivariate linear prediction model and the expected success data is obtained, the method further includes:
s10, acquiring the short message issuing historical data of the appointed time span;
s20, processing the short message issuing historical data into a multi-dimensional array set according to a preset data extraction rule, wherein the multi-dimensional array set comprises a plurality of multi-dimensional arrays, one multi-dimensional array corresponds to a unit time interval, the designated time span is divided into a plurality of unit time intervals, and the multi-dimensional array comprises the number of successful short message issuing of a plurality of dimensions in the unit time interval and the expected number;
s30, constructing a multiple linear regression model according to the multidimensional array set, and solving model parameters of the multiple linear regression model, wherein the model parameters comprise constant terms, dimension factors and residual error terms;
s40, verifying and correcting the multiple linear regression model by using a preset significance verification method to generate a multiple linear correction model;
and S50, verifying the multivariate linear correction model by using a preset model diagnosis method, and determining the verified multivariate linear correction model as the multivariate linear prediction model.
Understandably, the designated time span can be set according to actual needs, such as 1-24 hours. The short message issuing historical data comprises target issuing data and actual issuing success data of the historical task.
The preset data extraction rule can be set according to actual needs, for example, the successful number of short messages issued by a certain operator in a certain area in a unit time interval can be counted. In an example, the preset data extraction rule may extract a plurality of multidimensional arrays from the short message sending history data to form a multidimensional data set. The unit time interval can be set according to actual needs, such as one minute. The dimension of the multidimensional array can be set according to actual conditions. For example, the corresponding dimension parameters can be set according to different operators and different regions.
In one example, the multidimensional array set is shown in Table 2.
TABLE 2 multidimensional data set in an example
Wherein each row is a multidimensional array.
Multiple linear regression models can be constructed, such as:
Y＝β0+β1X1+β2X2+β3X3+…+βnXn+ε
wherein beta is0Is a constant term, β1、β2、β3、βnIs a dimensional factor, X1、X2、X3、XnDimension parameters (independent variables), Y predicted values (dependent variables), epsilon residual terms, and residual terms, wherein the residual terms are the sum of all uncertain factors, and the values of the residual terms cannot be observed. Can be falseLet epsilon follow a normal distribution: n (0, sigma)2) The value of which needs to be determined by a specific residual analysis means. In an example, a multiple linear regression model is constructed by using four dimensional factors of a region, a first operator, a second operator, and a third operator, respectively, and then the model can be expressed as:
Y＝β0+β1X1+β2X2+β3X3+β4X4+ε
wherein Y is a predicted value;
β0is a constant term;
X1、X2、X3respectively a first operator, a second operator and a third operator;
X4representing a region identifier;
ε is the residual term.
And substituting the multidimensional array set into a multiple linear regression model, and calculating to obtain model parameters. The model parameters include constant terms, individual dimensional factors, and residual terms (following a normal distribution, within a data range).
After solving the model parameters of the multiple linear regression model, the multiple linear regression model may be verified and corrected using a preset significance verification method. The preset significance checking method can be selected according to actual needs, such as a T-testing method, an F-testing method and an R-square testing method. One or more significance checking methods may be used. And obtaining a multivariate linear correction model through the check sum correction of a preset significance check method. The accuracy of the model parameters of the multivariate linear correction model is higher.
And then, verifying the multi-element linear correction model by using a preset model diagnosis method, and analyzing a residual error item and an abnormal point of the multi-element linear correction model to ensure the accuracy of the multi-element linear correction model. If the verification is passed, the multivariate linear correction model can be determined as the multivariate linear prediction model. And if the verification fails, adjusting the model parameters of the multivariate linear model or presetting a data extraction rule. For example, when the preset data extraction rule is modified, the unit time interval may be modified to change the original one minute to two minutes or thirty seconds.
Optionally, the preset significance checking method includes a T-test method, an F-test method, and an R-square test method;
step S40, the verifying and correcting the multiple linear regression model by using the preset significance verification method to generate a multiple linear correction model, including:
s401, verifying the multiple linear regression model according to the T test method to generate a first verification result;
s402, verifying the multiple linear regression model according to the F test method to generate a second verification result;
s403, verifying the multiple linear regression model according to the R square test method to generate a third verification result;
s404, if the first check result, the second check result and the third check result are all significant results, correcting the multiple linear regression model according to the first check result, the second check result and the third check result to obtain the multiple linear correction model.
Understandably, the preset significance checking method includes a T-test method, an F-test method, and an R-square test method. And verifying the multiple linear regression model by using each verification method to obtain a corresponding verification result. The verification result comprises two types, wherein one type represents that the multiple linear regression model has significance, namely the significant result, and the other type represents that the multiple linear regression model has no significance, namely the non-significant result.
Here, the T test method (student's T test) can test the independent variable X in the multiple linear regression modeliSignificance of Y (i ═ 1 to n) is generally judged by P-value, and when P-value is less than 0.01 or less, the independent variable X is describediThe correlation with Y is significant.
F test method for testing all independent variables { XiLinear significance for Y as a whole, also judged by P-valueLess than 0.01 or less, the argument { X ] as a whole is specifiediThe correlation between Y and Y is significant.
R2(R-Square) test method for determining the degree of fit of a regression equation, R2The value of (2) is (0, 1), and the closer to 1, the better the fitting degree.
In one example, the verification results of the three predetermined significance verification methods are as follows:
a first verification result of the T verification method: all independent variables were very significant;
and the second check result of the F check method is as follows: the verification result is remarkable, and p-value:<2.2e-16；
R2the third verification result of the verification method: the correlation was very strong at 0.972.
Before verification, the multiple linear regression model is as follows:
Y＝212.8780+0.8542*X1+0.6672*X2-0.6674*X3+0.4821*X4
after correction, the obtained multivariate linear correction model is:
Y＝212.87996+0.85423*X1+0.66724*X2-0.66741*X3+0.48214*X4
after the verification is completed, the multivariate linear correction model has more accurate relation coefficient compared with the multivariate linear regression model.
Optionally, the preset model diagnosis method includes a residual error analysis method and an abnormal point detection method;
step S50, namely, the verifying the multivariate linear correction model by using a preset model diagnosis method, and determining the verified multivariate linear correction model as the multivariate linear prediction model, including:
s501, verifying the multi-element linear correction model according to the residual error analysis method to generate a first verification result;
s502, verifying the multivariate linear correction model according to the abnormal point detection method to generate a second verification result;
s503, if the first verification result and the second verification result both meet preset verification requirements, judging that the multivariate linear correction model passes verification.
Understandably, in order to make the multivariate linear correction model as accurate as possible, the model can also be checked for correctness using a residual analysis method and an abnormal point detection method. When the residual error analysis method and the abnormal point detection method are used for checking, a graph for model diagnosis can be generated by means of a drawing tool, and visual analysis is carried out to determine whether the multivariate linear correction model passes the verification. The preset verification requirements can be set according to actual needs. In some examples, the preset validation requirements include: the first verification result is a pass verification and the second verification result is a pass verification.
Optionally, in step S501, the verifying the multivariate linear correction model according to the residual error analysis method to generate a first verification result, including:
s5011, obtaining a residual error and a fitting value of the multivariate linear correction model, filling the residual error and the fitting value into a first image, analyzing a first point distribution form of the residual error and the fitting value in the first image, and generating first analysis data;
s5012, acquiring a standardized residual error and a theoretical quantile of the multi-element linear correction model, filling the standardized residual error and the theoretical quantile into a second image, analyzing a second point distribution form of the standardized residual error and the theoretical quantile in the second image, and generating second analysis data;
s5013, acquiring a normalized residual square root of the multiple linear correction model, filling the normalized residual square root and the fitting value into a third image, analyzing the distribution form of a third point of the normalized residual square root and the fitting value in the third image, and generating third analysis data;
s5014, generating the first verification result according to the first analysis data, the second analysis data and the third analysis data.
Understandably, the residual error and the fitting value of the multivariate linear correction model can be obtained, the residual error and the fitting value are filled into the first image, and the distribution form of the residual error and the fitting value at the first point in the first image is analyzed to generate first analysis data. In an example, as shown in fig. 4, fig. 4 is a first image comparing a residual (Residuals) and a fitting value (fitedvalues or abbreviated as Fitted). By identifying (here, manually observable and then manually entered, or automatically identified by a graphical analysis tool) a first point distribution profile of the residuals and fit values in the first image, the following first analysis data can be obtained: data points between the residual error and the fitting value are uniformly distributed on two sides of y which is 0, random distribution is presented, and a trend line is a stable curve and has no obvious shape characteristic.
The normalized residual error and the theoretical quantile of the multi-element linear correction model can be obtained, the normalized residual error and the theoretical quantile are filled into the second image, the second point distribution form of the normalized residual error and the theoretical quantile in the second image is analyzed, and second analysis data are generated. In one example, as shown in FIG. 5, FIG. 5 is a second image (i.e., a normalQ-Q, standard qq plot) in which normalized residuals (normalized residuals) are compared to theoretical quantiles (theoretical quantiles). By identifying a second point distribution shape of the normalized residual and the theoretical quantile in the second image, the following second analysis data can be obtained: the data points in the second image are arranged in a diagonal straight line, approach to a straight line and are directly penetrated by the diagonal line, so that the data points intuitively conform to normal distribution.
The normalized residual square root and the fitting value of the multi-element linear correction model can be obtained, the normalized residual square root and the fitting value are filled into the third image, the distribution form of a third point of the normalized residual square root and the fitting value in the third image is analyzed, and third analysis data are generated. In an example, as shown in fig. 6, fig. 6 is a third image (Scale-Location) in which the normalized residual square root (squareStandardizadResiduals) and the fitting value (Fittedvalues) are compared. By identifying a third point distribution profile of the normalized residual square root and the fitted values in the third image, the following third analysis data can be obtained: the data points in the third image are uniformly distributed on two sides of y, which is 0, and the data points present random distribution, and the curve in the third image is smooth and has no obvious shape characteristics.
And if the first analysis data, the second analysis data and the third analysis data are not abnormal, the first verification result is that the multivariate linear correction model passes the verification.
In other examples, normalized residuals and lever values of the multivariate linear correction model may be obtained, the normalized residuals and the lever values may be filled into the abnormal point detection image, and distribution patterns of the normalized residuals and the lever values in the abnormal point detection image may be analyzed to generate abnormal point detection analysis data. In one example, as shown in fig. 7, fig. 7 is an anomaly detection image in which normalized residuals (normalized residuals) and Leverage values (lever) are compared. By identifying the distribution form of the normalized residual and the lever values in the abnormal point detection image, the following analysis data can be obtained: if no contour line appears, it indicates that there is no abnormal point in the data that particularly affects the regression result.
Optionally, in step S80, that is, the acquiring the preset normal interval matched with the short message pushing task includes:
s801, extracting a pushing keyword of the short message pushing task;
s802, searching corresponding normal task data in a task database according to the pushing keywords, wherein the normal task data comprises task data of a plurality of historical normal pushing tasks;
and S803, calculating the issuing upper limit and the issuing lower limit of each time point when the short message pushing task is executed according to the normal task data to form the preset normal interval.
Understandably, the push keyword includes one or more keywords. The service characteristics or the regional attributes of the short message pushing task can be selected as keywords, and the keywords can also be set according to the crowd characteristics corresponding to the pushing crowd. And the extraction rule for extracting the push keyword can be set according to actual needs.
The task database records task data of the historical short message pushing task, wherein the task data comprises feedback data and expected success data (calculated by the same multivariate linear prediction model). The historical normal pushing task refers to a pushing task which is verified at a later stage and is determined not to have an abnormal condition. The normal task data includes task data of a plurality of historical normal push tasks.
The task data of each historical normal push task includes predicted values for a plurality of points in time. At the same time point (which may be fixed time, such as 9: 00; or relative time, such as time fed back by an operator for the first time is taken as a starting point), the predicted values of different historical normal push tasks are different, the maximum predicted value of the different historical normal push tasks can be set as an upper issuing limit, and the minimum predicted value of the different historical normal push tasks can be set as an upper issuing limit. In some cases, if the predicted values at the same time point follow a normal distribution, the upper delivery limit may be set to μ +2 σ, and the lower delivery limit may be set to μ -2 σ. Here, μ is the average of the predicted values, and σ is the variance. The preset normal interval comprises an upper issuing limit and a lower issuing limit of each time point.
Optionally, in step S80, that is, the determining whether the expected success data is outside the preset normal interval includes:
s804, judging whether each predicted value in the expected success data is larger than the corresponding issuing upper limit or smaller than the issuing lower limit;
s805, when a predicted value in the expected success data is larger than the corresponding issuing upper limit, calculating a first difference between the predicted value and the issuing upper limit, and adding one to the overproof accumulated value; or, when the predicted value is smaller than the corresponding lower limit, calculating a second difference between the lower limit and the predicted value, and adding one to the overproof accumulated value; the initial value of the overproof accumulated value is zero;
s306, when the overproof accumulated value is larger than a preset accumulated threshold, or the first difference is larger than a first comparison threshold, or the second difference is larger than a second comparison threshold, judging that the expected successful data is out of a preset normal interval.
Understandably, a certain predicted value in expected success data can be compared with an upper issuing limit or a lower issuing limit at the same time point, if the predicted value is greater than the corresponding upper issuing limit, a first difference between the predicted value and the upper issuing limit is calculated, and meanwhile, an overproof accumulated value is added by one; and if the sum is less than the corresponding lower issuing limit, calculating a second difference between the lower issuing limit and the predicted value, and adding one to the overproof accumulated value. Here, the initial value of the superscalar cumulative value is zero.
And when the overproof accumulated value is larger than a preset accumulated threshold, or the first difference is larger than a first comparison threshold, or the second difference is larger than a second comparison threshold, judging that the expected successful data is out of a preset normal interval. And when the overproof accumulated value is smaller than or equal to the preset accumulated threshold, the first difference is smaller than or equal to the first comparison threshold, and the second difference is smaller than or equal to the second comparison threshold, judging that the expected successful data is in the preset normal interval. The overproof accumulated value can be set according to actual needs. The first comparison threshold and the second comparison threshold may be set according to actual needs, and may be fixed values or percentage values (for example, calculated by taking an average value as a base number).
In this place, set up the accumulation value and two comparison thresholds that exceed standard at the same time, can detect the short message and issue the anomaly well, reduce the situation that the misinformation takes place.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
In an embodiment, a short message delivery anomaly detection device is provided, and the short message delivery anomaly detection device corresponds to the short message delivery anomaly detection method in the above embodiments one to one. As shown in fig. 8, the short message delivery anomaly detection apparatus includes a feedback data receiving module 60, an expectation predicting module 70, a matching normal interval module 80, and an anomaly reminding module 90. The functional modules are explained in detail as follows:
a feedback data receiving module 60, configured to receive feedback data of a short message pushing task, where the feedback data includes data of successful short message delivery of multiple operators;
a prediction expectation module 70, configured to process the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
a matching normal interval module 80, configured to obtain a preset normal interval that matches the short message push task, and determine whether the expected success data is outside the preset normal interval;
and an exception reminding module 90, configured to send an exception reminder if the expected successful data is outside the predicted normal interval.
Optionally, the apparatus for detecting an exception issued by a short message further includes:
the acquisition historical data module is used for acquiring the short message issuing historical data of a specified time span;
the multidimensional array module is used for processing the short message issuing historical data into a multidimensional array set according to a preset data extraction rule, the multidimensional array set comprises a plurality of multidimensional arrays, one multidimensional array corresponds to a unit time interval, the specified time span is divided into a plurality of unit time intervals, and the multidimensional arrays comprise the number of successful orders of the short messages of a plurality of dimensions in the unit time interval and the expected number;
the system comprises a multi-dimensional array set establishing module, a regression model establishing module and a model parameter calculating module, wherein the multi-dimensional array set establishing module is used for establishing a multi-element linear regression model according to the multi-dimensional array set and solving the model parameters of the multi-element linear regression model, and the model parameters comprise constant terms, dimension factors and residual error terms;
the generating and correcting model module is used for verifying and correcting the multiple linear regression model by using a preset significance verifying method to generate a multiple linear correcting model;
and the generation prediction model module is used for verifying the multivariate linear correction model by using a preset model diagnosis method and determining the verified multivariate linear correction model as the multivariate linear prediction model.
Optionally, the preset significance checking method includes a T-test method, an F-test method, and an R-square test method;
the generating a modification model module comprises:
the first verification unit is used for verifying the multiple linear regression model according to the T test method to generate a first verification result;
the second verification unit is used for verifying the multivariate linear regression model according to the F test method to generate a second verification result;
the third verification unit is used for verifying the multiple linear regression model according to the R square verification method to generate a third verification result;
and the correction model obtaining unit is used for correcting the multiple linear regression model according to the first check result, the second check result and the third check result to obtain the multiple linear correction model if the first check result, the second check result and the third check result are all significant.
Optionally, the preset model diagnosis method includes a residual error analysis method and an abnormal point detection method;
the generating a modification model module comprises:
the residual error analysis unit is used for verifying the multi-element linear correction model according to the residual error analysis method to generate a first verification result;
the abnormal point detection unit is used for verifying the multivariate linear correction model according to the abnormal point detection method to generate a second verification result;
and the pass verification unit is used for judging that the multivariate linear correction model passes the verification if the first verification result and the second verification result both meet the preset verification requirement.
Optionally, the residual analysis unit includes:
the first image analysis unit is used for acquiring residual errors and fitting values of the multi-element linear correction model, filling the residual errors and the fitting values into a first image, analyzing the distribution form of the residual errors and the fitting values at first points in the first image and generating first analysis data;
the second image analysis unit is used for acquiring a standardized residual error and a theoretical quantile of the multi-element linear correction model, filling the standardized residual error and the theoretical quantile into a second image, and analyzing a second point distribution form of the standardized residual error and the theoretical quantile in the second image to generate second analysis data;
a third image analysis unit, configured to obtain a normalized residual square root of the multiple linear correction model, fill the normalized residual square root and the fitting value into a third image, and analyze a third point distribution form of the normalized residual square root and the fitting value in the third image to generate third analysis data;
a first verification result generation unit configured to generate the first verification result according to the first analysis data, the second analysis data, and the third analysis data.
Optionally, the matching normal interval module 80 includes:
the extraction keyword unit is used for extracting the push keywords of the short message push task;
the matching task data unit is used for searching corresponding normal task data in a task database according to the pushing keywords, and the normal task data comprises task data of a plurality of historical normal pushing tasks;
and a normal interval forming unit, configured to calculate, according to the normal task data, an upper issue limit and a lower issue limit at each time point when the short message push task is executed, so as to form the preset normal interval.
Optionally, the matching normal interval module 80 includes:
a predicted value judgment unit, configured to judge whether each predicted value in the expected success data is greater than the corresponding delivery upper limit or less than the delivery lower limit;
a difference calculating unit, configured to calculate a first difference between the predicted value and the upper issuing limit when the predicted value in the expected success data is greater than the corresponding upper issuing limit, and add one to the superstandard cumulative value; or, when the predicted value is smaller than the corresponding lower limit, calculating a second difference between the lower limit and the predicted value, and adding one to the overproof accumulated value; the initial value of the overproof accumulated value is zero;
and the abnormity judging unit is used for judging that the expected successful data is out of a preset normal interval when the overproof accumulated value is larger than a preset accumulated threshold, or the first difference is larger than a first comparison threshold, or the second difference is larger than a second comparison threshold.
For the specific limitation of the short message issuing abnormality detection apparatus, reference may be made to the above limitation on the short message issuing abnormality detection method, which is not described herein again. All modules in the short message issuing abnormity detection device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 9. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a readable storage medium and an internal memory. The readable storage medium stores an operating system, computer readable instructions, and a database. The internal memory provides an environment for the operating system and execution of computer-readable instructions in the readable storage medium. The database of the computer equipment is used for storing data related to the short message issuing abnormity detection method. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer readable instructions are executed by a processor to realize a short message issuing abnormity detection method. The readable storage media provided by the present embodiment include nonvolatile readable storage media and volatile readable storage media.
In one embodiment, a computer device is provided, comprising a memory, a processor, and computer readable instructions stored on the memory and executable on the processor, the processor when executing the computer readable instructions implementing the steps of:
receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is outside the preset normal interval;
and if the expected successful data is outside the predicted normal interval, sending an abnormal prompt.
In one embodiment, one or more computer-readable storage media storing computer-readable instructions are provided, the readable storage media provided by the embodiments including non-volatile readable storage media and volatile readable storage media. The readable storage medium has stored thereon computer readable instructions which, when executed by one or more processors, perform the steps of:
receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is outside the preset normal interval;
and if the expected successful data is outside the predicted normal interval, sending an abnormal prompt.
It will be understood by those of ordinary skill in the art that all or part of the processes of the methods of the above embodiments may be implemented by hardware related to computer readable instructions, which may be stored in a non-volatile readable storage medium or a volatile readable storage medium, and when executed, the computer readable instructions may include processes of the above embodiments of the methods. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-mentioned functions.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present invention, and are intended to be included within the scope of the present invention.
Claims (10)
1. A short message issuing abnormity detection method is characterized by comprising the following steps:
receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
acquiring a preset normal interval matched with the short message pushing task, and judging whether the expected successful data is outside the preset normal interval;
and if the expected successful data is outside the predicted normal interval, sending an abnormal prompt.
2. The method for detecting short message issuance anomaly according to claim 1, wherein said processing the feedback data through the multivariate linear prediction model further comprises, before obtaining the expected success data:
acquiring short message issuing historical data of a specified time span;
processing the short message issuing historical data into a multi-dimensional array set according to a preset data extraction rule, wherein the multi-dimensional array set comprises a plurality of multi-dimensional arrays, one multi-dimensional array corresponds to a unit time interval, the designated time span is divided into a plurality of unit time intervals, and the multi-dimensional arrays comprise the number of successful orders of the short messages of a plurality of dimensions in the unit time interval and the expected number;
constructing a multiple linear regression model according to the multi-dimensional array set, and solving model parameters of the multiple linear regression model, wherein the model parameters comprise a constant term, a dimension factor and a residual error term;
verifying and correcting the multiple linear regression model by using a preset significance verification method to generate a multiple linear correction model;
and verifying the multivariate linear correction model by using a preset model diagnosis method, and determining the verified multivariate linear correction model as the multivariate linear prediction model.
3. The short message issuance abnormality detection method according to claim 2, wherein said preset significance verification method includes a T-test method, an F-test method, and an R-square test method;
the verifying and correcting the multiple linear regression model by using a preset significance verifying method to generate a multiple linear correction model, comprising the following steps of:
verifying the multiple linear regression model according to the T test method to generate a first verification result;
verifying the multiple linear regression model according to the F test method to generate a second verification result;
verifying the multiple linear regression model according to the R square test method to generate a third verification result;
and if the first check result, the second check result and the third check result are all significant results, correcting the multiple linear regression model according to the first check result, the second check result and the third check result to obtain the multiple linear correction model.
4. The short message issuance abnormality detection method according to claim 2, wherein said preset model diagnosis method includes a residual error analysis method and an abnormality point detection method;
the verifying the multivariate linear correction model by using a preset model diagnosis method, and determining the verified multivariate linear correction model as the multivariate linear prediction model comprises the following steps:
verifying the multi-element linear correction model according to the residual error analysis method to generate a first verification result;
verifying the multivariate linear correction model according to the abnormal point detection method to generate a second verification result;
and if the first verification result and the second verification result both meet the preset verification requirement, judging that the multivariate linear correction model passes the verification.
5. The method according to claim 4, wherein the verifying the multivariate linear correction model according to the residual error analysis method to generate a first verification result comprises:
obtaining a residual error and a fitting value of the multi-element linear correction model, filling the residual error and the fitting value into a first image, analyzing a first point distribution form of the residual error and the fitting value in the first image, and generating first analysis data;
acquiring a standardized residual error and a theoretical quantile of the multivariate linear correction model, filling the standardized residual error and the theoretical quantile into a second image, analyzing a second point distribution form of the standardized residual error and the theoretical quantile in the second image, and generating second analysis data;
acquiring a normalized residual square root of the multi-element linear correction model, filling the normalized residual square root and the fitting value into a third image, analyzing the distribution form of a third point of the normalized residual square root and the fitting value in the third image, and generating third analysis data;
and generating the first verification result according to the first analysis data, the second analysis data and the third analysis data.
6. The short message delivery abnormality detection method according to claim 1, wherein said obtaining a preset normal interval matched with said short message push task includes:
extracting a pushing keyword of the short message pushing task;
searching corresponding normal task data in a task database according to the pushing keywords, wherein the normal task data comprises task data of a plurality of historical normal pushing tasks;
and calculating the issuing upper limit and the issuing lower limit of each time point when the short message pushing task is executed according to the normal task data to form the preset normal interval.
7. The short message delivery abnormality detecting method according to claim 6, wherein said judging whether the expected success data is outside a preset normal interval includes:
judging whether each predicted value in the expected success data is larger than the corresponding issuing upper limit or smaller than the issuing lower limit;
when the predicted value in the expected success data is larger than the corresponding upper issuing limit, calculating a first difference between the predicted value and the upper issuing limit, and adding one to the overproof accumulated value; or, when the predicted value is smaller than the corresponding lower limit, calculating a second difference between the lower limit and the predicted value, and adding one to the overproof accumulated value; the initial value of the overproof accumulated value is zero;
and when the overproof accumulated value is larger than a preset accumulated threshold, or the first difference is larger than a first comparison threshold, or the second difference is larger than a second comparison threshold, judging that the expected successful data is out of a preset normal interval.
8. A short message issuing abnormity detection device is characterized by comprising:
the feedback data receiving module is used for receiving feedback data of a short message pushing task, wherein the feedback data comprises short message successful issuing data of a plurality of operators;
the prediction expectation module is used for processing the feedback data through a multivariate linear prediction model to obtain expected success data; the multivariate linear prediction model comprises a constant term and a dimension factor, wherein one dimension parameter corresponds to one dimension factor;
the matching normal interval module is used for acquiring a preset normal interval matched with the short message pushing task and judging whether the expected successful data is outside the preset normal interval or not;
and the abnormal reminding module is used for sending an abnormal reminding if the expected successful data is outside the prediction normal interval.
9. A computer device comprising a memory, a processor and computer readable instructions stored in the memory and executable on the processor, wherein the processor when executing the computer readable instructions implements the short message delivery anomaly detection method according to any one of claims 1 to 7.
10. One or more readable storage media storing computer-readable instructions which, when executed by one or more processors, cause the one or more processors to perform the short message delivery anomaly detection method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210043421.7A CN114401494B (en) | 2022-01-14 | 2022-01-14 | Short message issuing abnormality detection method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210043421.7A CN114401494B (en) | 2022-01-14 | 2022-01-14 | Short message issuing abnormality detection method, device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114401494A true CN114401494A (en) | 2022-04-26 |
| CN114401494B CN114401494B (en) | 2023-05-26 |
Family
ID=81230770
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210043421.7A Active CN114401494B (en) | 2022-01-14 | 2022-01-14 | Short message issuing abnormality detection method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114401494B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008113290A1 (en) * | 2007-03-16 | 2008-09-25 | Huawei Technologies Co., Ltd. | Method and device for pushing information |
| CN105323763A (en) * | 2014-06-27 | 2016-02-10 | 中国移动通信集团湖南有限公司 | Method and apparatus for identifying spam messages |
| CN110888788A (en) * | 2019-10-16 | 2020-03-17 | 平安科技（深圳）有限公司 | Anomaly detection method and device, computer equipment and storage medium |
| CN111460294A (en) * | 2020-03-31 | 2020-07-28 | 汉海信息技术(上海)有限公司 | Message pushing method and device, computer equipment and storage medium |
| CN112231181A (en) * | 2020-12-08 | 2021-01-15 | 平安科技（深圳）有限公司 | Data abnormal update detection method and device, computer equipment and storage medium |
-
2022
- 2022-01-14 CN CN202210043421.7A patent/CN114401494B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008113290A1 (en) * | 2007-03-16 | 2008-09-25 | Huawei Technologies Co., Ltd. | Method and device for pushing information |
| CN105323763A (en) * | 2014-06-27 | 2016-02-10 | 中国移动通信集团湖南有限公司 | Method and apparatus for identifying spam messages |
| CN110888788A (en) * | 2019-10-16 | 2020-03-17 | 平安科技（深圳）有限公司 | Anomaly detection method and device, computer equipment and storage medium |
| CN111460294A (en) * | 2020-03-31 | 2020-07-28 | 汉海信息技术(上海)有限公司 | Message pushing method and device, computer equipment and storage medium |
| CN112231181A (en) * | 2020-12-08 | 2021-01-15 | 平安科技（深圳）有限公司 | Data abnormal update detection method and device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114401494B (en) | 2023-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110086649B (en) | Abnormal flow detection method, device, computer equipment and storage medium | |
| CN110493190B (en) | Data information processing method and device, computer equipment and storage medium | |
| US8019734B2 (en) | Statistical determination of operator error | |
| US7844641B1 (en) | Quality management in a data-processing environment | |
| CN111475370A (en) | Operation and maintenance monitoring method, device and equipment based on data center and storage medium | |
| CN111193609B (en) | Application abnormity feedback method and device and application abnormity monitoring system | |
| CN110888911A (en) | Sample data processing method and device, computer equipment and storage medium | |
| CN111709794A (en) | Abnormal order processing method and device, computer equipment and storage medium | |
| CN110472114B (en) | Abnormal data early warning method and device, computer equipment and storage medium | |
| CN111259791B (en) | Nuclear power station material acceptance method and device, computer equipment and storage medium | |
| CN110069364B (en) | Spare part data error correction method and device, computer equipment and storage medium | |
| CN113312258B (en) | Interface testing method, device, equipment and storage medium | |
| CN110764902A (en) | Virtual resource allocation method and device based on AI (Artificial Intelligence), computer equipment and storage medium | |
| CN112101584A (en) | Nuclear power station spare part systematic maintenance method, device, equipment and storage medium | |
| CN110928859A (en) | Model monitoring method and device, computer equipment and storage medium | |
| CN115879748A (en) | Enterprise informatization management integration platform based on big data | |
| CN111091276A (en) | Enterprise risk scoring method and device, computer equipment and storage medium | |
| CN113535449B (en) | Abnormal event restoration processing method and device, computer equipment and storage medium | |
| CN112379913B (en) | Software optimization method, device, equipment and storage medium based on risk identification | |
| CN114202256A (en) | Architecture upgrading early warning method and device, intelligent terminal and readable storage medium | |
| CN113821933A (en) | APP flow prediction method and device, computer equipment and storage medium | |
| CN114401494B (en) | Short message issuing abnormality detection method, device, computer equipment and storage medium | |
| CN112463783A (en) | Index data monitoring method and device, computer equipment and storage medium | |
| CN114896955A (en) | Data report processing method and device, computer equipment and storage medium | |
| CN115237796A (en) | Link abnormity identification method and device, storage medium and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |