Logistics delivery influencing factor prediction method, device, equipment and storage medium

Abstract

The present invention relates to the field of logistics technology, and discloses a method, device, equipment and storage medium for predicting factors affecting logistics delivery, which are used to improve the accuracy of predicting factors affecting logistics delivery. The method for predicting factors affecting logistics delivery includes: obtaining an initial logistics indicator data set from a preset logistics network point, the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; performing descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set; performing pairwise indicator data correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence; performing analysis and prediction processing on the initial logistics indicator data set based on a preset influencing factor prediction model, a statistical data set and a correlation coefficient sequence to obtain a target risk influencing factor, the target risk influencing factor being a factor affecting order delivery that originates from a preset logistics business level or a preset logistics algorithm level.

Description

Technical Field

[0001] The present invention relates to the field of logistics technology, and in particular to a method, device, equipment and storage medium for predicting factors affecting logistics delivery. Background Art

[0002] With the rapid development of the express delivery industry in recent years, the volume of express delivery orders has also increased dramatically. Express delivery companies face delivery problems such as poor timeliness and accuracy in the process of express delivery orders. Express delivery companies can conduct data mining and indicator configuration based on delivery problems to form relevant logistics indicator data. The relevant logistics indicator data include recognition rate indicators for each category and accuracy indicators for each category.

[0003] The recognition rate and accuracy indicators of various categories in the express delivery industry are affected by both the address data maintenance at the business level and different natural language processing (NLP) algorithms. The existing recognition rate and accuracy indicators of various categories come from many sources and cover a wide range, resulting in low efficiency and accuracy in the analysis of factors affecting logistics delivery. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for predicting factors affecting logistics delivery, which are used to improve the accuracy of predicting factors affecting logistics delivery.

[0005] To achieve the above-mentioned purpose, the first aspect of the present invention provides a method for predicting factors affecting logistics delivery, comprising: obtaining an initial logistics indicator data set from a preset logistics outlet, the initial logistics indicator data set being used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; performing descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, the statistical data set being used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data; performing pairwise correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence; performing analysis and prediction processing on the initial logistics indicator data set based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain a target risk influencing factor, the target risk influencing factor being a factor affecting order delivery that originates from a preset logistics business level or a preset logistics algorithm level.

[0006] In a feasible implementation manner, the initial logistics indicator data set is obtained from a preset logistics network, and the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors, including: receiving a logistics delivery influencing factor analysis request, and performing parameter parsing on the logistics delivery influencing factor analysis request to obtain a logistics network identifier; setting the logistics network identifier as a search index, and querying a preset logistics network database table according to the search index to obtain an initial logistics indicator data set, and the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; and storing the initial logistics indicator data set in a logistics indicator collection data table of a preset distributed cluster database.

[0007] In a feasible implementation, the initial logistics indicator data set is subjected to descriptive statistical analysis to obtain a statistical data set, and the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data, including: classifying the initial logistics indicator data set according to a preset clustering algorithm to obtain multiple groups of classified logistics indicator data; performing descriptive statistical analysis on the multiple groups of classified data according to a preset statistical function to obtain statistical data corresponding to each group of classified data; merging the statistical data corresponding to each group of classified data into a statistical data set, and storing the statistical data set in a logistics indicator statistical data table of a preset distributed cluster database, and the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0008] In a feasible implementation, the initial logistics indicator data set is subjected to pairwise indicator data correlation analysis to obtain a correlation coefficient sequence, including: performing pairwise indicator data correlation analysis on the initial logistics indicator data set through a preset correlation analysis function to obtain multiple pairwise correlation coefficients; combining the multiple pairwise correlation coefficients into a correlation coefficient sequence in order from the earliest to the latest analysis time, and storing the correlation coefficient sequence in a logistics indicator correlation data table of a preset distributed cluster database.

[0009] In a feasible implementation manner, the initial logistics indicator data set is analyzed and predicted based on the preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors, and the target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, including: hypothesis testing and indicator screening are performed on the initial logistics indicator data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics indicator data set; multiple linear regression is performed on the candidate logistics indicator data set through the preset influencing factor prediction model to obtain fitting results corresponding to each candidate logistics indicator data, and the fitting results corresponding to each candidate logistics indicator data are stored in a logistics indicator fitting data table of a preset distributed cluster database; the fitting results corresponding to each candidate logistics indicator data are comprehensively analyzed and processed according to preset data analysis rules to obtain target risk influencing factors, and the target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level.

[0010] In a feasible implementation, the initial logistics indicator data set is subjected to hypothesis testing and indicator screening based on the statistical data set and the correlation coefficient sequence to obtain a candidate logistics indicator data set, including: performing a distribution test on the statistical data set to obtain statistical data that pass the test, and performing a significance test on the correlation coefficient sequence to obtain correlation coefficients that pass the test; and screening the initial logistics indicator data set according to the statistical data that pass the test and the correlation coefficients that pass the test to obtain a candidate logistics indicator data set.

[0011] In a feasible implementation, before obtaining an initial logistics indicator data set from a preset logistics outlet, wherein the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors, the logistics delivery influencing factor prediction method also includes: reading an initial logistics sample data set, and performing descriptive statistical analysis and correlation analysis on the initial logistics sample data set to obtain a target logistics sample data set; performing model iterative training based on a preset training function and the target logistics sample data set to obtain a preset influencing factor prediction model.

[0012] The second aspect of the present invention provides a device for predicting factors affecting logistics delivery, including: an acquisition module, used to acquire an initial logistics indicator data set from preset logistics outlets, the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; a statistical module, used to perform descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data; an analysis module, used to perform pairwise indicator data correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence; a prediction module, used to perform analysis and prediction processing on the initial logistics indicator data set based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain a target risk influencing factor, the target risk influencing factor is a factor affecting order delivery that originates from a preset logistics business level or a preset logistics algorithm level.

[0013] In a feasible implementation, the acquisition module is specifically used to: receive a request for analyzing factors affecting logistics delivery, and perform parameter analysis on the request for analyzing factors affecting logistics delivery to obtain a logistics point identifier; set the logistics point identifier as a search index, and query a preset logistics point database table according to the search index to obtain an initial logistics indicator data set, wherein the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; and store the initial logistics indicator data set in a logistics indicator collection data table of a preset distributed cluster database.

[0014] In a feasible implementation, the statistical module is specifically used to: classify the initial logistics indicator data set according to a preset clustering algorithm to obtain multiple groups of classified logistics indicator data; perform descriptive statistical analysis on the multiple groups of classified data according to a preset statistical function to obtain statistical data corresponding to each group of classified data; merge the statistical data corresponding to each group of classified data into a statistical data set, and store the statistical data set in a logistics indicator statistical data table of a preset distributed cluster database, wherein the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0015] In a feasible implementation, the analysis module is specifically used to: perform pairwise correlation analysis on the initial logistics indicator data set through a preset correlation analysis function to obtain multiple pairwise correlation coefficients; combine the multiple pairwise correlation coefficients into a correlation coefficient sequence in order from the first to the last analysis time, and store the correlation coefficient sequence in a logistics indicator correlation data table of a preset distributed cluster database.

[0016] In a feasible implementation, the prediction module includes: a screening unit, used to perform hypothesis testing and indicator screening on the initial logistics indicator data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics indicator data set; a processing unit, used to perform multivariate linear regression processing on the candidate logistics indicator data set through a preset influencing factor prediction model to obtain a fitting result corresponding to each candidate logistics indicator data, and store the fitting result corresponding to each candidate logistics indicator data in a logistics indicator fitting data table of a preset distributed cluster database; an analysis unit, used to perform comprehensive analysis and processing on the fitting results corresponding to each candidate logistics indicator data according to preset data analysis rules to obtain target risk influencing factors, wherein the target risk influencing factors are factors that affect order delivery and are derived from a preset logistics business level or a preset logistics algorithm level.

[0017] In a feasible implementation, the screening unit is specifically used to: perform a distribution test on the statistical data set to obtain statistical data that pass the test, and perform a significance test on the correlation coefficient sequence to obtain correlation coefficients that pass the test; and screen the initial logistics indicator data set according to the statistical data that pass the test and the correlation coefficients that pass the test to obtain a candidate logistics indicator data set.

[0018] In a feasible implementation, the logistics delivery influencing factor prediction device also includes: a reading module, used to read the initial logistics sample data set, and perform descriptive statistical analysis and correlation analysis on the initial logistics sample data set to obtain a target logistics sample data set; a training module, used to perform model iterative training based on a preset training function and the target logistics sample data set to obtain a preset influencing factor prediction model.

[0019] The third aspect of the present invention provides a logistics delivery influencing factor prediction device, comprising: a memory and at least one processor, the memory storing instructions, the memory and the at least one processor being interconnected via a line; the at least one processor calling the instructions in the memory so that the logistics delivery influencing factor prediction device executes the above-mentioned logistics delivery influencing factor prediction method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the computer-readable storage medium is run on a computer, the computer executes the above-mentioned method for predicting factors affecting logistics delivery.

[0021] In the technical solution provided by the present invention, an initial logistics indicator data set is obtained from a preset logistics network point, and the initial logistics indicator data set is used to indicate multiple categories of accuracy indicators and multiple categories of recognition rate indicators related to logistics delivery influencing factors; a descriptive statistical analysis is performed on the initial logistics indicator data set to obtain a statistical data set, and the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data; a pairwise indicator data correlation analysis is performed on the initial logistics indicator data set to obtain a correlation coefficient sequence; based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence, the initial logistics indicator data set is analyzed and predicted to obtain target risk influencing factors, and the target risk influencing factors are factors that affect order delivery and are derived from a preset logistics business level or a preset logistics algorithm level. In an embodiment of the present invention, by performing descriptive statistical analysis and pairwise correlation analysis on the initial logistics indicator data set, a statistical data set and a correlation coefficient sequence are obtained, and the initial logistics indicator data set is analyzed and predicted based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors. The target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, thereby improving the accuracy and analysis efficiency of predicting logistics delivery influencing factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of an embodiment of a method for predicting factors affecting logistics delivery in an embodiment of the present invention;

[0023] Figure 2 It is a schematic diagram of another embodiment of the method for predicting factors affecting logistics delivery in an embodiment of the present invention;

[0024] Figure 3 A schematic diagram of an embodiment of a device for predicting factors affecting logistics delivery according to an embodiment of the present invention;

[0025] Figure 4 It is a schematic diagram of another embodiment of the device for predicting factors affecting logistics delivery in an embodiment of the present invention;

[0026] Figure 5 Schematic diagram of an embodiment of a device for predicting factors affecting logistics delivery in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present invention provide a method, device, equipment and storage medium for predicting factors affecting logistics delivery, which are used to analyze and predict the initial logistics indicator data set through a preset influencing factor prediction model, a statistical data set and the correlation coefficient sequence, to obtain factors affecting order delivery from a preset logistics business level or a preset logistics algorithm level, and to improve the accuracy of predicting factors affecting logistics delivery.

[0028] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 An embodiment of the method for predicting factors affecting logistics delivery in the embodiment of the present invention includes:

[0030] 101. Obtain an initial logistics indicator data set from a preset logistics network point, where the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to factors affecting logistics delivery.

[0031] Among them, the logistics delivery influencing factors can come from the preset logistics business level or the preset logistics algorithm level. The initial logistics indicator data set (multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors) can include predicted network point (two-segment) recognition rate, predicted network point (two-segment) accuracy, predicted network point (three-segment) recognition rate, predicted network point (three-segment) accuracy, delivery code (three-segment) recognition rate, delivery code (three-segment) accuracy, algorithm-network recognition rate, algorithm-network accuracy, algorithm-delivery code recognition rate, algorithm-delivery code accuracy, three-segment code historical library predicted network point, three-segment code historical library predicted network point accuracy, three-segment code historical library delivery code recognition rate, three-segment code historical library delivery code accuracy, Cainiao-destination network point Recognition rate, Cainiao-destination network point accuracy, Cainiao-delivery code recognition rate, Cainiao-delivery code accuracy, delivery code electronic fence predicted network point recognition rate, delivery code electronic fence predicted network point accuracy, delivery code electronic fence delivery code recognition rate, delivery code electronic fence delivery code accuracy, three-segment code history library-road number recognition rate, three-segment code history library-road number accuracy, three-segment code history library-point of interest POI recognition rate, three-segment code history library-POI accuracy, three-segment code manual designated delivery code recognition rate, three-segment code manual designated delivery code accuracy, three-segment code keyword manual designated delivery code recognition rate, three-segment code keyword manual designated delivery code accuracy, three-segment code key dictionary PSFW predicted delivery code recognition rate, three-segment code key dictionary PSFW predicted delivery code accuracy. The initial logistics indicator data set may also include network point name, distribution name, parent network company name, city name and receipt volume, which are not limited here. Each initial logistics indicator data in the initial logistics indicator data set has a corresponding indicator name and at least one indicator value.

[0032] Specifically, the server periodically queries the preset distributed file storage system corresponding to the preset logistics outlets through the preset indicator mining task to obtain an initial logistics indicator data set, and the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors. The preset distributed file storage system is used to store multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors, and the multiple category accuracy indicators and multiple category recognition rate indicators can be updated and modified in real time.

[0033] It is understandable that the execution subject of the present invention may be a logistics delivery influencing factor prediction device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0034] 102. Perform descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, where the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0035] Specifically, the server counts the number of positive numbers, the number of negative numbers, the mean, the mode, the median, the frequency, the relative number, the range, the variance, the standard deviation, the standard error, the full range, the interquartile range, the quartiles, the percentiles, the coefficient of variation, the skewness, and the kurtosis based on the initial logistics indicator data set to obtain a statistical data set, which is used to indicate the accuracy distribution corresponding to each initial logistics indicator data set and the recognition rate distribution corresponding to each initial logistics indicator data set, wherein the mean includes the truncated mean, the geometric mean, the harmonic mean, etc., and the quartiles include the upper quartile and the lower quartile. The mean, median, and mode are used to indicate the central tendency of the initial logistics indicator data set. The range, variance, and standard deviation are used to indicate the degree of dispersion of the initial logistics indicator data set. Skewness and kurtosis are used to indicate the distribution shape of the initial logistics indicator data set.

[0036] 103. Perform pairwise correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence.

[0037] Among them, the correlation coefficient sequence is used to indicate the indicator data related to the factors affecting logistics delivery in the initial logistics indicator data set. Specifically, the server performs a correlation analysis of the indicator data of each pair on the initial logistics indicator data set according to a preset correlation analysis algorithm to obtain multiple pairwise correlation coefficients, and sorts and combines the multiple pairwise correlation coefficients from large to small to obtain a correlation coefficient sequence. The multiple pairwise correlation coefficients include positive correlation coefficients (the changes of the pairwise indicator data are the same) and negative correlation coefficients (the changes of the pairwise indicator data are opposite). Further, the server can also perform absolute value operations and / or same value deduplication processing on the positive correlation coefficients and negative correlation coefficients, respectively, to obtain multiple processed correlation coefficients, and sort and combine the multiple processed correlation coefficients from large to small to obtain a correlation coefficient sequence. For example, the pairwise correlation coefficient A can be used to indicate the relationship between each category recognition rate indicator, each category accuracy indicator and the preset algorithm, and can also be used to indicate the relationship between each category recognition rate indicator, each category accuracy indicator and the address quality.

[0038] It should be noted that the preset correlation analysis algorithm is a partial correlation analysis algorithm or a distance analysis algorithm. The partial correlation analysis algorithm is used to calculate the linear correlation between any two initial logistics indicator data in the initial logistics indicator data set, and the distance analysis algorithm is used to indicate a measure of the degree of similarity or dissimilarity between any two initial logistics indicator data in the initial logistics indicator data set.

[0039] 104. Based on the preset influencing factor prediction model, statistical data set and correlation coefficient sequence, the initial logistics indicator data set is analyzed and predicted to obtain the target risk influencing factors. The target risk influencing factors are factors that affect order delivery from the preset logistics business level or the preset logistics algorithm level.

[0040] Specifically, the server screens the initial logistics indicator data set according to the statistical data set and the correlation coefficient sequence to obtain the candidate logistics indicator data set. The candidate logistics indicator data set is used to indicate the indicator data that is most related to the factors affecting logistics delivery. The fitting results corresponding to each candidate logistics indicator data include P value, R square and standard deviation; the server analyzes and predicts the candidate indicator data set based on the preset influencing factor prediction model to obtain the target risk influencing factor. The target risk influencing factor is the factor that affects order delivery from the preset logistics business level or the preset logistics algorithm level. For example, the recognition rate index of each category can characterize the address quality, and the accuracy index of each category can better reflect the accuracy of the algorithm. If the recognition rate and accuracy of each preset logistics algorithm at the logistics outlet are low, it means that the address quality is poor. If the recognition rate and accuracy of each preset logistics algorithm at the outlet are high, it means that the address quality is good.

[0041] It should be noted that the server generates multiple types of statistical chart data according to the initial logistics indicator data set, the statistical data set and the correlation coefficient sequence, and sends the multiple types of statistical chart data to the target terminal, so that the target terminal draws and displays the multiple types of statistical chart data to realize real-time analysis and monitoring of each initial logistics indicator data.

[0042] In an embodiment of the present invention, by performing descriptive statistical analysis and pairwise correlation analysis on the initial logistics indicator data set, a statistical data set and a correlation coefficient sequence are obtained, and the initial logistics indicator data set is analyzed and predicted based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors. The target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, thereby improving the accuracy and analysis efficiency of predicting logistics delivery influencing factors.

[0043] See also Figure 2 Another embodiment of the method for predicting factors affecting logistics delivery in the embodiment of the present invention includes:

[0044] 201. Obtain an initial logistics indicator data set from a preset logistics network point, where the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to factors affecting logistics delivery.

[0045] Among them, the initial logistics indicator data set comes from multiple preset logistics outlets. Optionally, the server receives a request for analyzing factors affecting logistics delivery, and performs parameter analysis on the request for analyzing factors affecting logistics delivery to obtain a logistics outlet identifier. Specifically, the server receives a request for analyzing factors affecting logistics delivery, and updates the request for analyzing factors affecting logistics delivery to a preset message queue, and performs resource locator decoding processing on the request for analyzing factors affecting logistics delivery in the order of writing through the preset message queue to obtain a decoded analysis request. The server performs parameter analysis and parameter verification on the decoded analysis request to obtain a logistics outlet identifier. For example, the logistics outlet identifier can be yd_001, or it can be "(yd_001, yd_002)", that is, the logistics outlet identifier can be One, or two, or more, which are not limited here; the server sets the logistics network point identifier as the search index, and queries the preset logistics network point database table according to the search index to obtain the initial logistics indicator data set, which is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors, wherein the preset logistics network point database table can be a memory database table (for example, a remote service dictionary redis or a mapping table map, which are not limited here), or a relational database table, which are not limited here; the server stores the initial logistics indicator data set in the logistics indicator collection data table of the preset distributed cluster database. The preset distributed cluster database can be a big data engine greenplum, or other types of databases, which are not limited here.

[0046] Furthermore, before step 201, the server reads the initial logistics sample data set, and performs descriptive statistical analysis and correlation analysis on the initial logistics sample data set to obtain the target logistics sample data set; the server performs model iteration training based on the preset training function and the target logistics sample data set (for example, the server performs multivariate linear regression processing through the preset training function madlib.linregr_train and the target logistics sample data set) to obtain the preset influencing factor prediction model. The initial logistics sample data set includes sample data of accuracy index of each category, sample data of recognition rate index of each category, and sample data of logistics influencing factors.

[0047] 202. Perform descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, where the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0048] Among them, descriptive statistical analysis is used to describe each initial logistics indicator data using statistical indicators. Optionally, the server classifies the initial logistics indicator data set according to a preset clustering algorithm to obtain multiple groups of classified logistics indicator data, wherein the preset clustering algorithm may include k-means clustering and density-based clustering algorithms, and may also include Gaussian mixture clustering algorithms, which are not specifically limited here; the server performs descriptive statistical analysis on multiple groups of classified data according to a preset statistical function (for example, the machine learning library madlib.summary) to obtain statistical data corresponding to each group of classified data; the server merges the statistical data corresponding to each group of classified data into a statistical data set, and stores the statistical data set in the logistics indicator statistical data table of the preset distributed cluster database, and the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0049] 203. Perform pairwise correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence.

[0050] It should be noted that the correlation coefficient sequence is a string separated and combined according to preset characters, and the preset characters can be commas or underscores, which are not limited here. Optionally, the server performs a pairwise correlation analysis on the initial logistics indicator data set through a preset correlation analysis function to obtain multiple pairwise correlation coefficients, wherein the preset correlation analysis function can be madlib.correlation or other functions, which are not limited here; the server combines multiple pairwise correlation coefficients into a correlation coefficient sequence in the order of analysis time from first to last, and stores the correlation coefficient sequence in the logistics indicator correlation data table of the preset distributed cluster database.

[0051] Furthermore, the server determines whether any two initial logistics indicator data of the initial logistics indicator data set conform to the normal distribution according to the statistical data set; the server calculates the Pearson correlation coefficient for any two initial logistics indicator data that conform to the normal distribution in the initial logistics indicator data set, and obtains multiple pairwise correlation coefficients: the server calculates the Spearman rank correlation coefficient for any two initial logistics indicator data that do not conform to the normal distribution in the initial logistics indicator data set, and obtains multiple pairwise correlation coefficients. The value range of each pairwise correlation coefficient is greater than or equal to -1 and less than or equal to 1. Each pairwise correlation coefficient is used to indicate the degree of linear relationship between any two initial logistics indicator data.

[0052] 204. Hypothesis testing and indicator screening are performed on the initial logistics indicator data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics indicator data set.

[0053] Among them, the hypothesis test includes chi-square distribution test, t distribution test, F distribution test and significance test. Optionally, the server performs a distribution test on the statistical data set (that is, chi-square distribution test, t distribution test and / or F distribution test) to obtain statistical data that pass the test, and performs a significance test on the correlation coefficient sequence to obtain the correlation coefficient that passes the test. For example, the server compares the adjacent correlation coefficients of the correlation coefficient sequence in turn and performs redundancy removal operations to obtain the correlation coefficient that passes the test; the server screens the initial logistics indicator data set according to the statistical data that pass the test and the correlation coefficient that pass the test to obtain the candidate logistics indicator data set. For example, when the target pairwise correlation coefficient (belonging to the initial logistics indicator data set) is 0 or the absolute value corresponding to the target pairwise correlation coefficient is less than 0.1, the server determines that the two initial logistics indicator data corresponding to the target pairwise correlation coefficient are not related. When the absolute value corresponding to the target pairwise correlation coefficient is greater than 0.9, the server determines that the two initial logistics indicator data corresponding to the target pairwise correlation coefficient are significantly correlated. When the absolute value corresponding to the target pairwise correlation coefficient is greater than or equal to 0.1 and less than or equal to 0.9, the server determines that the two initial logistics indicator data corresponding to the target pairwise correlation coefficient are moderately correlated. The server deletes the irrelevant coefficients to obtain the correlation coefficient that passes the test, and merges the statistical data that passes the test and the initial logistics indicator data corresponding to the correlation coefficient that passes the test into a candidate logistics indicator data set (that is, input variables).

[0054] 205. Perform multivariate linear regression processing on the candidate logistics indicator data set through the preset influencing factor prediction model to obtain the fitting results corresponding to each candidate logistics indicator data, and store the fitting results corresponding to each candidate logistics indicator data in the logistics indicator fitting data table of the preset distributed cluster database.

[0055] Specifically, the server performs parameter encapsulation based on the candidate logistics index data set to obtain input parameters. The server performs multivariate linear regression processing on the input parameters through a preset influencing factor prediction model to obtain fitting results corresponding to each candidate logistics index data. The fitting results corresponding to each candidate logistics index data include P value, R square and standard deviation; and the fitting results corresponding to each candidate logistics index data are stored in the logistics index fitting data table of the preset distributed cluster database. Further, the server fits the predicted network point (two-segment) recognition rate, predicted network point (two-segment) accuracy rate, predicted network point (three-segment) recognition rate, predicted network point (three-segment) accuracy rate, delivery code (three-segment) recognition rate and delivery code (three-segment) accuracy rate with each category recognition rate index and each category accuracy index to process the correlation, causal relationship and / or nonlinear relationship, and obtain a preset influencing factor prediction model.

[0056] For example, the server fits the prediction network point (two-segment) accuracy rate with each category recognition rate index and each category accuracy index, and obtains an R-square of about 0.5, with a general fitting effect. The server fits the prediction network point (three-segment) recognition rate with each category recognition rate index and each category accuracy index, and obtains an R-square of 0.94, with a good fitting effect and a significant relationship. The server fits the prediction network point (three-segment) accuracy rate with each category recognition rate index and each category accuracy index, and obtains an R-square of 0.98, with a good fitting effect and a significant relationship. The server fits the delivery code (three-segment) recognition rate with each category recognition rate index and each category accuracy index, and obtains an R-square of 0.94, with a good fitting effect and a significant relationship. The server fits the delivery code (three-segment) accuracy rate with each category recognition rate index and each category accuracy index, and obtains an R-square of 0.99, with a good fitting effect and a significant relationship. Based on the above fitting process, when the P value is less than 0.05 and the R-square is about 0.99, the fitting effect of the preset influencing factor prediction model is good and the model is significant. That is, the preset influencing factor prediction model is used to locate the delivery code recognition rate and delivery code accuracy rate, which comes from the preset logistics business level or the preset logistics algorithm level.

[0057] 206. According to the preset data analysis rules, the fitting results corresponding to the candidate logistics indicator data are comprehensively analyzed and processed to obtain the target risk influencing factors. The target risk influencing factors are factors that affect order delivery from the preset logistics business level or the preset logistics algorithm level.

[0058] Among them, the preset logistics business level may include logistics ordering business, logistics sorting and loading business, logistics transportation business and logistics distribution business. The preset logistics business level may also include logistics code scanning business, which is not limited here; the preset logistics algorithm level is used to indicate multiple natural language processing algorithms involved in the logistics business. The preset data analysis rules are used to indicate that the P value is less than 0.05 and the R square is greater than or equal to 0.99. The server screens the fitting results corresponding to each candidate logistics indicator data according to the preset data analysis rules, obtains the candidate fitting results, and obtains multiple target logistics indicator data corresponding to the candidate fitting results; the server obtains the regression coefficient, and determines the target risk influencing factor based on the regression coefficient and the multiple target logistics indicator data. The target risk influencing factor is the factor that affects the order delivery from the preset logistics business level or the preset logistics algorithm level. The regression coefficient is the relationship between the dependent variable and the respective variable. When the regression coefficient is greater than 0, the dependent variable increases as the respective variable increases. When the regression coefficient is less than 0, the dependent variable decreases as the respective variable increases. The greater the absolute value of the regression coefficient, the greater the influence of the respective variable on the dependent variable. It can be understood that the p value is a decreasing indicator of the credibility of the result. The closer the R square is to 1, the better the fitting regression effect.

[0059] In an embodiment of the present invention, by performing descriptive statistical analysis and pairwise correlation analysis on the initial logistics indicator data set, a statistical data set and a correlation coefficient sequence are obtained, and the initial logistics indicator data set is analyzed and predicted based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors. The target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, thereby improving the accuracy and analysis efficiency of predicting logistics delivery influencing factors.

[0060] The above describes the method for predicting factors affecting logistics delivery in the embodiment of the present invention. The following describes the device for predicting factors affecting logistics delivery in the embodiment of the present invention. Figure 3 In one embodiment of the present invention, a device for predicting factors affecting logistics delivery includes:

[0061] An acquisition module 301 is used to acquire an initial logistics index data set from a preset logistics network point, where the initial logistics index data set is used to indicate multiple category accuracy indexes and multiple category recognition rate indexes related to logistics delivery influencing factors;

[0062] A statistical module 302 is used to perform descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, where the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data;

[0063] The analysis module 303 is used to perform correlation analysis on the initial logistics indicator data set between two indicators to obtain a correlation coefficient sequence;

[0064] The prediction module 304 is used to analyze and predict the initial logistics indicator data set based on the preset influencing factor prediction model, statistical data set and correlation coefficient sequence to obtain the target risk influencing factors. The target risk influencing factors are factors that affect order delivery from the preset logistics business level or the preset logistics algorithm level.

[0065] In an embodiment of the present invention, by performing descriptive statistical analysis and pairwise correlation analysis on the initial logistics indicator data set, a statistical data set and a correlation coefficient sequence are obtained, and the initial logistics indicator data set is analyzed and predicted based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors. The target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, thereby improving the accuracy and analysis efficiency of predicting logistics delivery influencing factors.

[0066] See also Figure 4Another embodiment of the device for predicting factors affecting logistics delivery in the embodiment of the present invention includes:

[0067] An acquisition module 301 is used to acquire an initial logistics index data set from a preset logistics network point, where the initial logistics index data set is used to indicate multiple category accuracy indexes and multiple category recognition rate indexes related to logistics delivery influencing factors;

[0068] A statistical module 302 is used to perform descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, where the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data;

[0069] The analysis module 303 is used to perform correlation analysis on the initial logistics indicator data set between two indicators to obtain a correlation coefficient sequence;

[0070] The prediction module 304 is used to analyze and predict the initial logistics indicator data set based on the preset influencing factor prediction model, statistical data set and correlation coefficient sequence to obtain the target risk influencing factors. The target risk influencing factors are factors that affect order delivery from the preset logistics business level or the preset logistics algorithm level.

[0071] Optionally, the acquisition module 301 may also be specifically used for:

[0072] Receive a request for analyzing factors affecting logistics delivery, and perform parameter analysis on the request for analyzing factors affecting logistics delivery to obtain a logistics point identifier;

[0073] The logistics network point identifier is set as a search index, and a preset logistics network point database table is queried according to the search index to obtain an initial logistics indicator data set, where the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to factors affecting logistics delivery;

[0074] The initial logistics indicator data set is stored in the logistics indicator collection data table of the preset distributed cluster database.

[0075] Optionally, the statistics module 302 may also be specifically used for:

[0076] The initial logistics indicator data set is classified and processed according to a preset clustering algorithm to obtain multiple groups of classified logistics indicator data;

[0077] Perform descriptive statistical analysis on multiple groups of classified data according to preset statistical functions to obtain statistical data corresponding to each group of classified data;

[0078] The statistical data corresponding to each group of classified data are merged into a statistical data set, and the statistical data set is stored in the logistics indicator statistical data table of the preset distributed cluster database. The statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

[0079] Optionally, the analysis module 303 may also be specifically used for:

[0080] Performing pairwise correlation analysis on the initial logistics indicator data set through a preset correlation analysis function to obtain multiple pairwise correlation coefficients;

[0081] Multiple pairwise correlation coefficients are combined into a correlation coefficient sequence in the order of analysis time from earliest to latest, and the correlation coefficient sequence is stored in a logistics index correlation data table of a preset distributed cluster database.

[0082] Optionally, the prediction module 304 may further include:

[0083] A screening unit 3041 is used to perform hypothesis testing and index screening on the initial logistics index data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics index data set;

[0084] The processing unit 3042 is used to perform multivariate linear regression processing on the candidate logistics index data set through a preset influencing factor prediction model to obtain a fitting result corresponding to each candidate logistics index data, and store the fitting result corresponding to each candidate logistics index data in a logistics index fitting data table of a preset distributed cluster database;

[0085] The analysis unit 3043 is used to comprehensively analyze and process the fitting results corresponding to each candidate logistics indicator data according to the preset data analysis rules to obtain the target risk influencing factors, which are factors affecting order delivery from the preset logistics business level or the preset logistics algorithm level.

[0086] Optionally, the screening unit 3041 may also be specifically used for:

[0087] Perform distribution test on statistical data set to obtain statistical data that pass the test, and perform significance test on correlation coefficient sequence to obtain correlation coefficient that pass the test;

[0088] The initial logistics indicator data set is screened according to the statistical data and the correlation coefficient that have passed the test to obtain the candidate logistics indicator data set.

[0089] Optionally, the logistics delivery influencing factor prediction device also includes:

[0090] The reading module 305 is used to read the initial logistics sample data set, and perform descriptive statistical analysis and correlation analysis on the initial logistics sample data set to obtain the target logistics sample data set;

[0091] The training module 306 is used to perform model iterative training based on a preset training function and a target logistics sample data set to obtain a preset influencing factor prediction model.

[0092] In an embodiment of the present invention, by performing descriptive statistical analysis and pairwise correlation analysis on the initial logistics indicator data set, a statistical data set and a correlation coefficient sequence are obtained, and the initial logistics indicator data set is analyzed and predicted based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors. The target risk influencing factors are factors that affect order delivery from a preset logistics business level or a preset logistics algorithm level, thereby improving the accuracy and analysis efficiency of predicting logistics delivery influencing factors.

[0093] above Figure 3 and Figure 4 The device for predicting factors affecting logistics delivery in the embodiment of the present invention is described in detail from a modular perspective. The device for predicting factors affecting logistics delivery in the embodiment of the present invention is described in detail from a hardware processing perspective.

[0094] Figure 5 : is a structural diagram of a logistics dispatch influencing factor prediction device provided by an embodiment of the present invention. The logistics dispatch influencing factor prediction device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more massive storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 can be short-term storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the logistics dispatch influencing factor prediction device 500. Furthermore, the processor 510 can be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the logistics dispatch influencing factor prediction device 500.

[0095] The logistics delivery influencing factor prediction device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that Figure 5 The structure of the logistics delivery influencing factor prediction device shown does not constitute a limitation of the logistics delivery influencing factor prediction device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0096] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the method for predicting factors affecting logistics delivery.

[0097] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0098] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0099] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting factors affecting logistics delivery, characterized in that: The method for predicting factors affecting logistics delivery includes: Acquire an initial logistics index data set from a preset logistics network point, wherein the initial logistics index data set is used to indicate multiple category accuracy indexes and multiple category recognition rate indexes related to logistics delivery influencing factors; Performing descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, where the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data; Performing pairwise correlation analysis on the initial logistics indicator data set to obtain a correlation coefficient sequence; Based on the preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence, the initial logistics indicator data set is analyzed and predicted to obtain the target risk influencing factor, wherein the target risk influencing factor is a factor affecting order delivery from the preset logistics business level or the preset logistics algorithm level, and the influencing factor prediction model is a multivariate linear regression model; The initial logistics indicator data set consists of multiple categories of accuracy indicators and multiple categories of recognition rate indicators related to factors affecting logistics delivery, including the two-segment recognition rate of predicted outlets, the two-segment accuracy of predicted outlets, the three-segment recognition rate of predicted outlets, the three-segment accuracy of predicted outlets, the three-segment recognition rate of delivery codes, the three-segment accuracy of delivery codes, the algorithm-outlet recognition rate, the algorithm-outlet accuracy rate, the algorithm-delivery code recognition rate, the algorithm-delivery code accuracy rate, the three-segment code historical library predicted outlets, the three-segment code historical library predicted outlets accuracy rate, the three-segment code historical library delivery code recognition rate, the three-segment code historical library delivery code accuracy rate, the Cainiao-destination outlet recognition rate, the Cainiao-destination outlet accuracy rate, the Cainiao-delivery code recognition rate, and the Cainiao -Delivery code accuracy, delivery code electronic fence predicted network point recognition rate, delivery code electronic fence predicted network point accuracy, delivery code electronic fence delivery code recognition rate, delivery code electronic fence delivery code accuracy, three-segment code history library-road number recognition rate, three-segment code history library-road number accuracy, three-segment code history library-point of interest POI recognition rate, three-segment code history library-POI accuracy, three-segment code manually specified delivery code recognition rate, three-segment code manually specified delivery code accuracy, three-segment code keyword manually specified delivery code recognition rate, three-segment code keyword manually specified delivery code accuracy, three-segment code key dictionary PSFW predicted delivery code recognition rate, three-segment code key dictionary PSFW predicted delivery code accuracy.

2. The method for predicting factors affecting logistics delivery according to claim 1, characterized in that: The initial logistics index data set is obtained from the preset logistics network points, and the initial logistics index data set is used to indicate multiple category accuracy indexes and multiple category recognition rate indexes related to logistics delivery influencing factors, including: Receiving a request for analyzing factors affecting logistics delivery, and performing parameter analysis on the request for analyzing factors affecting logistics delivery to obtain a logistics point identifier; The logistics network point identifier is set as a search index, and a preset logistics network point database table is queried according to the search index to obtain an initial logistics indicator data set, where the initial logistics indicator data set is used to indicate multiple category accuracy indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; The initial logistics index data set is stored in a logistics index collection data table of a preset distributed cluster database.

3. The method for predicting factors affecting logistics delivery according to claim 1, characterized in that: The descriptive statistical analysis is performed on the initial logistics indicator data set to obtain a statistical data set, which is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data, including: Classifying the initial logistics indicator data set according to a preset clustering algorithm to obtain multiple groups of classified logistics indicator data; Performing descriptive statistical analysis on the multiple groups of classified data respectively according to a preset statistical function to obtain statistical data corresponding to each group of classified data; The statistical data corresponding to each group of classified data are merged into a statistical data set, and the statistical data set is stored in a logistics indicator statistical data table of a preset distributed cluster database. The statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data.

4. The method for predicting factors affecting logistics delivery according to claim 1, characterized in that: The pairwise correlation analysis of the initial logistics indicator data set is performed to obtain a correlation coefficient sequence, including: Performing pairwise correlation analysis on the initial logistics indicator data set using a preset correlation analysis function to obtain a plurality of pairwise correlation coefficients; The plurality of pairwise correlation coefficients are combined into a correlation coefficient sequence in the order of analysis time from earliest to latest, and the correlation coefficient sequence is stored in a logistics index correlation data table of a preset distributed cluster database.

5. The method for predicting factors affecting logistics delivery according to claim 1, characterized in that: The initial logistics indicator data set is analyzed and predicted based on the preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain target risk influencing factors, which are factors affecting order delivery from a preset logistics business level or a preset logistics algorithm level, including: Perform hypothesis testing and index screening on the initial logistics index data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics index data set; Performing multivariate linear regression processing on the candidate logistics index data set through a preset influencing factor prediction model to obtain fitting results corresponding to each candidate logistics index data, and storing the fitting results corresponding to each candidate logistics index data in a logistics index fitting data table of a preset distributed cluster database; According to the preset data analysis rules, the fitting results corresponding to each candidate logistics indicator data are comprehensively analyzed and processed to obtain the target risk influencing factors, which are factors affecting order delivery from the preset logistics business level or the preset logistics algorithm level.

6. The method for predicting factors affecting logistics delivery according to claim 5, characterized in that: The performing hypothesis testing and index screening on the initial logistics index data set according to the statistical data set and the correlation coefficient sequence to obtain a candidate logistics index data set includes: Performing a distribution test on the statistical data set to obtain statistical data that pass the test, and performing a significance test on the correlation coefficient sequence to obtain correlation coefficients that pass the test; The initial logistics indicator data set is screened according to the statistical data that have passed the inspection and the correlation coefficient that has passed the inspection to obtain a candidate logistics indicator data set.

7. The method for predicting factors affecting logistics delivery according to any one of claims 1 to 6, characterized in that: Before obtaining an initial logistics index data set from a preset logistics network point, wherein the initial logistics index data set is used to indicate multiple category accuracy indexes and multiple category recognition rate indexes related to logistics delivery influencing factors, the logistics delivery influencing factor prediction method further includes: Reading an initial logistics sample data set, and performing descriptive statistical analysis and correlation analysis on the initial logistics sample data set to obtain a target logistics sample data set; Model iterative training is performed based on a preset training function and the target logistics sample data set to obtain a preset influencing factor prediction model.

8. A device for predicting factors affecting logistics delivery, characterized in that: The logistics delivery influencing factor prediction device comprises: An acquisition module, used to acquire an initial logistics index data set from a preset logistics network point, wherein the initial logistics index data set is used to indicate multiple category accuracy rate indicators and multiple category recognition rate indicators related to logistics delivery influencing factors; A statistical module, used for performing descriptive statistical analysis on the initial logistics indicator data set to obtain a statistical data set, wherein the statistical data set is used to indicate the accuracy distribution corresponding to each initial logistics indicator data and the recognition rate distribution corresponding to each initial logistics indicator data; An analysis module, used to perform correlation analysis of two indicator data on the initial logistics indicator data set to obtain a correlation coefficient sequence; A prediction module is used to analyze and predict the initial logistics indicator data set based on a preset influencing factor prediction model, the statistical data set and the correlation coefficient sequence to obtain a target risk influencing factor, wherein the target risk influencing factor is a factor affecting order delivery from a preset logistics business level or a preset logistics algorithm level, and the influencing factor prediction model is a multiple linear regression model; The initial logistics indicator data set consists of multiple categories of accuracy indicators and multiple categories of recognition rate indicators related to factors affecting logistics delivery, including the two-segment recognition rate of predicted outlets, the two-segment accuracy of predicted outlets, the three-segment recognition rate of predicted outlets, the three-segment accuracy of predicted outlets, the three-segment recognition rate of delivery codes, the three-segment accuracy of delivery codes, the algorithm-outlet recognition rate, the algorithm-outlet accuracy rate, the algorithm-delivery code recognition rate, the algorithm-delivery code accuracy rate, the three-segment code historical library predicted outlets, the three-segment code historical library predicted outlets accuracy rate, the three-segment code historical library delivery code recognition rate, the three-segment code historical library delivery code accuracy rate, the Cainiao-destination outlet recognition rate, the Cainiao-destination outlet accuracy rate, the Cainiao-delivery code recognition rate, and the Cainiao -Delivery code accuracy, delivery code electronic fence predicted network point recognition rate, delivery code electronic fence predicted network point accuracy, delivery code electronic fence delivery code recognition rate, delivery code electronic fence delivery code accuracy, three-segment code history library-road number recognition rate, three-segment code history library-road number accuracy, three-segment code history library-point of interest POI recognition rate, three-segment code history library-POI accuracy, three-segment code manually specified delivery code recognition rate, three-segment code manually specified delivery code accuracy, three-segment code keyword manually specified delivery code recognition rate, three-segment code keyword manually specified delivery code accuracy, three-segment code key dictionary PSFW predicted delivery code recognition rate, three-segment code key dictionary PSFW predicted delivery code accuracy.

9. A device for predicting factors affecting logistics delivery, characterized in that: The logistics delivery influencing factor prediction device comprises: a memory and at least one processor, the memory stores instructions, and the memory and the at least one processor are interconnected via a line; The at least one processor calls the instructions in the memory so that the logistics delivery influencing factor prediction device executes the logistics delivery influencing factor prediction method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting factors affecting logistics delivery as described in any one of claims 1 to 7 is implemented.

Images