Regional business environment evaluation analysis system and method based on big data analysis

Abstract

The invention discloses a big data analysis-based regional business environment evaluation and analysis system and method. The method comprises the steps of collecting multi-source customer service data; identifying abnormal growth points in the data by calculating the approximation degree between adjacent subdivision time nodes; calculating a local outlier factor (LOF) for sampling data, and evaluating the abnormal degree of a data point; meanwhile, in the scheme, based on the LOF and the abnormal growth evaluation value, the comprehensive abnormal evaluation score of each datum is calculated, the abnormal degrees of different data points are compared more visually, and the analysis operability is enhanced; and comparing the comprehensive anomaly evaluation score with a preset threshold value, identifying a result and generating a report to provide decision support, thereby facilitating optimization of the overall operation efficiency in the region.

Description

Technical Field

[0001] The present invention relates to the field of information data analysis, and in particular to a system and method for evaluating and analyzing a regional business environment based on big data analysis. Background Art

[0002] For large-scale manufacturing enterprises, the efficiency of logistics in the region and even the entire surrounding area is an important factor affecting the main profits and business environment of the enterprise. Only through an efficient and smooth logistics system can the goods produced by these enterprises be quickly transferred from the production line to the hands of consumers and realize their value. The level of logistics efficiency directly determines the time cost of products from production to sales. An efficient logistics system can shorten the product turnover cycle, speed up inventory turnover, and reduce the risk of inventory backlog, thereby saving a lot of operating costs for enterprises. At the same time, the rapid logistics response capability can also improve the terminal satisfaction of enterprises and enhance the market competitiveness of large-scale manufacturing enterprises. Obviously, logistics efficiency can be said to be an important component of the enterprise supply chain.

[0003] In regional economic development, the improvement of logistics efficiency and regional logistics management evaluation can drive the development of related industries, such as warehousing, transportation, information processing, etc., forming an industrial agglomeration effect. It is the key to improving the regional economy and one of the important means to optimize the business environment and promote regional economic development.

[0004] The study found that logistics companies (logistics terminals) in the region generate a large amount of multi-source customer service data in their operations. These data include order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal cost and efficiency indicator data, etc. Performing logistics anomaly detection on these multi-source customer service data provides important information for the logistics industry's operation optimization, customer management, and efficiency improvement;

[0005] However, traditional anomaly detection methods mainly rely on overall statistical analysis or rule setting, but lack sensitivity to dynamic abnormal changes at subdivided time nodes. In practical applications, due to the large volatility of logistics terminal data, traditional methods often cannot accurately identify abnormal changes in details, resulting in the inability to timely warn of potential operational problems;

[0006] Moreover, traditional anomaly detection methods often rely only on global analysis when dealing with high-dimensional data, while ignoring the importance of local anomalies in the data. The local outlier factor (LOF) method can detect anomalies in local data in some cases, but traditional methods have not effectively integrated it into the analysis framework of multi-source data, nor have they combined other factors, such as abnormal growth assessment values, to comprehensively analyze the occurrence of anomalies.

[0007] At the same time, traditional anomaly detection often only targets pre-sales data such as multi-source customer service data, and does not conduct sufficient analysis of post-sales logistics data, making it difficult to make targeted adjustments to logistics efficiency. Summary of the invention

[0008] The purpose of the present invention is to provide a regional business environment evaluation and analysis system and method based on big data analysis, which solves the above-mentioned technical problems pointed out in the prior art.

[0009] The present invention provides a regional business environment evaluation and analysis system based on big data analysis, comprising: an information collection platform, an identification module, a sampling calculation module, an evaluation calculation module and an output processing module;

[0010] The information collection platform is used to collect logistics multi-source customer service data at subdivided time nodes within a preset time period; the logistics multi-source customer service data includes order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal fees and logistics terminal efficiency index data;

[0011] The identification module is used to obtain the approximation of the logistics multi-source customer service data at each segmented time node based on the logistics multi-source customer service data at two adjacent segmented time nodes, and obtain the abnormal growth evaluation value G(p) of each data p' in the logistics multi-source customer service data at each segmented time node based on the approximation of the logistics multi-source customer service data at each segmented time node;

[0012] The sampling calculation module is used to obtain the sampling data p of each data p' in the logistics multi-source customer service data of each subdivided time node, and then calculate the local outlier factor based on the neighborhood k obtained by cross-validation of the sampling data p, and output the local outlier factor evaluation value LOF(p) of the sampling data p;

[0013] The evaluation and calculation module is used to calculate the comprehensive abnormal evaluation score E(p) of each data in the logistics multi-source customer service data at each segmented time node based on the local outlier factor evaluation value LOF(p) of the sampled data p and the abnormal growth evaluation value G(p) of each data;

[0014] The output processing module is used to perform logistics benefit evaluation processing operations according to the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node.

[0015] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0016] From the analysis of the regional business environment evaluation and analysis system and method based on big data analysis provided by the present invention, it can be seen that in specific applications, multi-source customer service data is first collected, including orders, logistics terminal information, customer feedback, etc., to provide comprehensive basic data support, and through data preprocessing, the high quality and consistency of the data are ensured to lay a solid foundation for subsequent analysis; the above high-quality basic data can improve the accuracy of subsequent analysis and provide a reliable basis for anomaly detection; further, by calculating the approximation between adjacent segmented time nodes and identifying abnormal growth points in the data, potential problems and risks can be quickly discovered, providing technical data support and clues for further analysis and decision-making; by calculating the local outlier factor (LOF) of the sampled data ), evaluate the degree of abnormality of data points. The unsupervised learning method effectively captures local anomalies in the data set, improving the flexibility and accuracy of anomaly detection; at the same time, in the above scheme, based on LOF and abnormal growth evaluation value, the comprehensive anomaly evaluation score of each data is calculated. The comprehensive evaluation score provides a quantitative indicator, which can more intuitively compare the degree of abnormality of different data points and enhance the operability of the analysis; compare the comprehensive anomaly evaluation score with the preset threshold, identify abnormal customers or logistics links, and generate reports to provide decision support, help optimize logistics processes and customer service. By identifying anomalies and providing targeted suggestions, it can significantly improve logistics efficiency and enterprise terminal satisfaction, and help optimize the overall operational efficiency of the region. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the main process of a regional business environment evaluation and analysis method based on big data analysis;

[0018] Figure 2 It is a schematic diagram of the simulation effect of calculating the domain local density ρ(i) of each sampling data p based on the optimized neighborhood k in a regional business environment evaluation and analysis method based on big data analysis;

[0019] Figure 3 A schematic diagram of the operation steps of introducing logistics after-sales variable information and combining comprehensive abnormal evaluation scores to perform logistics benefit evaluation processing operations in a regional business environment evaluation analysis method based on big data analysis;

[0020] Figure 4 A schematic diagram of the operation steps for calculating the Spearman rank correlation coefficient ρ in a regional business environment evaluation analysis method based on big data analysis;

[0021] Figure 5 It is a schematic diagram of the simulation of variables in a regional business environment evaluation and analysis method based on big data analysis;

[0022] Figure 6A schematic diagram of the operation steps for obtaining the transfer efficiency evaluation coefficient of a logistics transfer station in a regional business environment evaluation and analysis method based on big data analysis;

[0023] Figure 7 This is a schematic diagram of the simulation effect of calculating the Pearson correlation coefficient r for a combination of sub-factors in a regional business environment evaluation analysis method based on big data analysis. DETAILED DESCRIPTION

[0024] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The present invention is further described in detail below through specific embodiments in conjunction with the accompanying drawings.

[0026] The present invention provides a regional business environment evaluation and analysis system based on big data analysis, comprising: an information collection platform, an identification module, a sampling calculation module, an evaluation calculation module and an output processing module;

[0027] The information collection platform is used to collect logistics multi-source customer service data at subdivided time nodes within a preset time period; the logistics multi-source customer service data includes order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal fees and logistics terminal efficiency index data;

[0028] The identification module is used to obtain the approximation of the logistics multi-source customer service data at each segmented time node based on the logistics multi-source customer service data at two adjacent segmented time nodes, and obtain the abnormal growth evaluation value G(p) of each data p' in the logistics multi-source customer service data at each segmented time node based on the approximation of the logistics multi-source customer service data at each segmented time node;

[0029] The sampling calculation module is used to obtain the sampling data p of each data p' in the logistics multi-source customer service data of each subdivided time node, and then calculate the local outlier factor based on the neighborhood k obtained by cross-validation of the sampling data p, and output the local outlier factor evaluation value LOF(p) of the sampling data p;

[0030] The evaluation and calculation module is used to calculate the comprehensive abnormal evaluation score E(p) of each data in the logistics multi-source customer service data at each segmented time node based on the local outlier factor evaluation value LOF(p) of the sampled data p and the abnormal growth evaluation value G(p) of each data;

[0031] The output processing module is used to perform logistics benefit evaluation processing operations according to the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node.

[0032] Preferably, the evaluation calculation module includes an initial processing module, a selection processing module, a domain calculation module, a correlation calculation module and an evaluation value calculation module:

[0033] The initial processing module is used to select a neighborhood k of an initial size; and evaluate the accuracy of the local outlier factor evaluation value based on the neighborhood k through cross-validation based on the neighborhood k;

[0034] The selection processing module is used to select the optimized neighborhood k based on the accuracy rate;

[0035] The domain calculation module is used to calculate the domain local density ρ(i) of each of the sampled data p based on the optimized neighborhood k;

[0036] The association calculation module is used to calculate the association sensitivity factor m of each of the sampled data based on the local density ρ(i) of the field i ;

[0037] The evaluation value calculation module is used to calculate the value of the neighborhood k and the associated sensitivity factor m based on the optimized neighborhood k and the associated sensitivity factor m. i Calculate and obtain the local outlier factor evaluation value.

[0038] Preferably, the calculation method of the local density ρ(i9) of the field is:

[0039]

[0040] Where N k (i) is the neighborhood of the i-th sample data; d(ij) is the distance function, which represents the distance between sample data i and sample data j;

[0041] The associated sensitivity factor m i The calculation method is:

[0042]

[0043] The local outlier factor evaluation value is calculated as follows:

[0044]

[0045] In the formula, m j is the correlation sensitivity factor of the i-th sampling data; m i is the correlation sensitivity factor of the i-th sampling data.

[0046] Preferably, the output processing module includes an acquisition module, a Spearman rank calculation module and an evaluation output processing module:

[0047] The acquisition module is used to collect and acquire logistics after-sales variable information; the logistics after-sales variable information includes logistics satisfaction, return rate and logistics complaint rate, transfer efficiency evaluation coefficient of logistics transfer station, and delivery delay time;

[0048] The Spearman rank calculation module is used to calculate the Spearman rank correlation coefficient ρ based on the combination of the logistics after-sales variable information and the comprehensive abnormality assessment score;

[0049] The evaluation output processing module is used to determine whether the Spearman rank correlation coefficient ρ is greater than 0 or less than 0; if so, a logistics benefit evaluation processing operation is performed based on the logistics after-sales variable information corresponding to the Spearman rank correlation coefficient ρ and the comprehensive abnormality evaluation score; if not, a logistics benefit evaluation processing operation is performed based on the comprehensive abnormality evaluation score.

[0050] Preferably, the Spearman rank calculation module specifically includes: a traversal processing module, a first calculation module and a second calculation module;

[0051] The traversal processing module is used to traverse each of the logistics after-sales variable information, and obtain a variable pair based on the combination of the logistics after-sales variable information and the comprehensive abnormality evaluation score corresponding to the logistics after-sales variable information;

[0052] The first calculation module is used to obtain a first variable sequence set by sorting the values ​​of all the logistics after-sales variable information from high to low; and obtain a second variable sequence set by sorting the values ​​of the comprehensive abnormality evaluation scores from high to low;

[0053] The second calculation module is used to calculate the ranking difference of the variable pair based on the first variable sequence set and the second variable sequence set; and calculate the Spearman rank correlation coefficient ρ based on the ranking difference;

[0054] The Spearman rank correlation coefficient ρ is calculated as:

[0055]

[0056] Where, d i is the ranking difference; n is the total value of the logistics after-sales variable information and the comprehensive abnormality evaluation score.

[0057] Preferably, the acquisition module includes a transfer data acquisition module, a matrix construction module and an evaluation coefficient calculation module:

[0058] The transit data acquisition module is used to collect and obtain logistics transit data; the logistics transit data includes transit station data M ij , the throughput P of each transfer station in the transfer station data jh , regional total throughput R iy and the transfer station coverage area P ij ;

[0059] The matrix construction module is used to construct a logistics transfer matrix based on the logistics transfer data; the logistics transfer matrix includes a transfer station throughput matrix P, a regional total throughput matrix vector R and a coverage area matrix A;

[0060] Each element P in the transfer station throughput matrix P jh represents the unit time throughput of the hth transfer station in the jth transfer station; each element R in the regional total throughput matrix vector R iy represents the total item throughput per unit time in the i-th area; each element A in the coverage area matrix A ij represents the coverage area of ​​the jth transfer station in the i-th area;

[0061] The evaluation coefficient calculation module is used to calculate the transfer efficiency evaluation coefficient E of the logistics transfer station based on the transfer station throughput matrix P, the regional total throughput matrix vector R and the coverage area matrix A. ij .

[0062] Preferably, the transfer efficiency evaluation coefficient E of the logistics transfer station is ij The calculation method is:

[0063]

[0064] In the formula, ∑ h P jh R represents the sum of the throughput per unit time of all transfer stations in the jth transfer station; iy is the total item throughput of the i-th area; A ij is the coverage area of ​​the j-th transfer station in the i-th area.

[0065] Preferably, it also includes an optimization processing module; the optimization processing module is used to iteratively optimize the logistics multi-source customer service data based on the transfer efficiency evaluation coefficient of the logistics transfer station to obtain updated logistics multi-source customer service data; and re-perform logistics benefit evaluation processing operations based on the updated logistics multi-source customer service data.

[0066] Preferably, the optimization processing module specifically includes a sub-factor module, a Pearson correlation coefficient module, a target sub-factor combination processing module and an update processing module;

[0067] The sub-factor module is used to obtain multiple sub-factor combinations by combining sub-factors in pairs based on the transfer efficiency evaluation coefficient of the logistics transfer station and the logistics multi-source customer service data;

[0068] The Pearson correlation coefficient module is used to calculate the Pearson correlation coefficient r based on the value of the logistics transfer station's transfer efficiency evaluation coefficient sub-factor and the value of the logistics multi-source customer service data sub-factor in the sub-factor combination;

[0069] The Pearson correlation coefficient r is calculated as follows:

[0070]

[0071] Where X is the value of the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station; Y is the value of the sub-factor in the logistics multi-source customer service data; is the average and sum of the sub-factors of the transfer efficiency evaluation coefficient of the logistics transfer station is the average of the values ​​of the sub-factors in the logistics multi-source customer service data;

[0072] The target sub-factor combination processing module is used to screen and obtain the sub-factor combination corresponding to the Pearson correlation coefficient r being greater than or equal to a preset Pearson correlation coefficient threshold as the target sub-factor combination;

[0073] The update processing module is used to add the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station in the target sub-factor combination to the logistics multi-source customer service data to obtain updated logistics multi-source customer service data.

[0074] Embodiment 2

[0075] like Figure 1 As shown, the second embodiment of the present invention provides a method for evaluating and analyzing the regional business environment based on big data analysis, including the following steps:

[0076] Step S10: Collecting logistics multi-source customer service data at subdivided time nodes within a preset time period;

[0077] The logistics multi-source customer service data includes order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal (i.e. logistics company) transportation path data, customer feedback order response data (the customer here refers to the production enterprise in the current area, and its feedback on the logistics order in the area is the order response data), logistics terminal (i.e. logistics company) fees and logistics terminal efficiency index data; that is, the logistics terminal specifically refers to the logistics company in the area;

[0078] It should be noted that in the above-mentioned embodiment of the present application, within a preset time period, the system collects logistics multi-source customer service data at segmented time nodes through multiple channels (such as logistics management system, customer relationship management system, social media, etc.), and it is also necessary to pre-process the logistics multi-source customer service data to make the data have high quality and consistency, thereby ensuring the accuracy of subsequent analysis.

[0079] Step S20: obtaining the approximation of the logistics multi-source customer service data at each segmented time node according to the logistics multi-source customer service data at two adjacent segmented time nodes, and obtaining the abnormal growth assessment value G)p) of each data p' in the logistics multi-source customer service data at each segmented time node according to the approximation of the logistics multi-source customer service data at each segmented time node;

[0080] It should be noted that the above-mentioned embodiment of the present application may use distance metrics such as Euclidean distance or cosine similarity to evaluate the similarity between data points (i.e., the above-mentioned approximation), and identify the abnormal growth evaluation value of each node in combination with the approximation (i.e., when the approximation is small, it proves that the growth of logistics multi-source customer service data at two adjacent segmented time nodes is abnormal, thereby analyzing and identifying to obtain the abnormal growth evaluation value), thereby identifying the data points that fluctuate significantly in the time series according to the abnormal growth evaluation value; specifically, the calculation method of the approximation is:

[0081] Where D i With D i+1 is the logistics multi-source customer service data of two adjacent segmented time nodes, d(D i , D i+1 ) is the distance function, which represents the logistics multi-source customer service data D at the segmented time nodes i Logistics multi-source customer service data D with segmented time nodes i+1 The distance between.

[0082] Step S30: obtaining the sampled data p of each data p' in the logistics multi-source customer service data of each segmented time node, and then calculating the local outlier factor based on the neighborhood k obtained by cross-validation of the sampled data p, and outputting the local outlier factor evaluation value LOF(p) of the sampled data p;

[0083] It should be noted that the above-mentioned sampled data p and each data p' in the logistics multi-source customer service data at each subdivided time node are substantially the same data; the difference is that the sampled data p is sub-data obtained by sampling from each data p' in the logistics multi-source customer service data at each subdivided time node;

[0084] The above-mentioned local outlier factor (LOF) is an unsupervised learning algorithm for detecting local outliers in a data set. It evaluates the degree of abnormality of each point by comparing the density of the data point with its neighboring points. When calculating the local outlier factor evaluation value LOF of the sampled data, it is calculated by comparing the density of a point with the density of other points in its neighborhood. If the LOF value of a point is significantly greater than 1, it is usually regarded as an abnormal point. Among them, the neighborhood means that when calculating the local outlier factor evaluation value LOF of the sampled data, a neighborhood k is defined for each sampled data by setting a fixed radius or selecting a fixed number of nearest neighbors, and the local density of each point is calculated, that is, the number of data points in its neighborhood. Generally, points with high density are considered normal, and points with low density may be abnormal.

[0085] Step S40: Based on the local outlier factor evaluation value LOF(p) of the sampled data p and the abnormal growth evaluation value G(p) of each data, the comprehensive abnormal evaluation score E(p) of each data in the logistics multi-source customer service data at each segmented time node is calculated;

[0086] The calculation method of the comprehensive abnormality evaluation score E(p) is:

[0087] E(p)=w 1 LOF(p)+w 2 G(p);

[0088] In the formula, w 1 、w 2 is the weight coefficient, reflecting the importance of each indicator;

[0089] Step S50: performing a logistics benefit evaluation processing operation according to the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node;

[0090] Specifically, the above-mentioned embodiment of the present application performs logistics benefit evaluation processing operations based on the comprehensive abnormality evaluation score, which specifically means: comparing the comprehensive evaluation score with a preset threshold, identifying customers or logistics links with abnormal performance, and then generating reports based on the abnormal customers or logistics links, providing decision support, and helping to optimize logistics processes and customer service.

[0091] It should be noted that the above-mentioned embodiment of the present application first collects multi-source customer service data, including orders, logistics terminal information, customer feedback (or feedback from enterprises in the region on logistics orders), etc., to provide comprehensive basic data support; and ensures high data quality and consistency through data preprocessing, laying a solid foundation for subsequent analysis; the above-mentioned high-quality basic data can improve the accuracy of subsequent analysis and provide a reliable basis for anomaly detection; further, by calculating the proximity between adjacent segmented time nodes and identifying abnormal growth points in the data, potential problems and risks can be quickly discovered, providing technical data support and clues for further analysis and decision-making; by calculating the local outlier factor (LOF) of the sampled data, the data can be evaluated. The unsupervised learning method effectively captures the local anomalies in the data set and improves the flexibility and accuracy of anomaly detection. At the same time, in the above scheme, the comprehensive anomaly evaluation score of each data is calculated based on LOF and abnormal growth evaluation value. The comprehensive evaluation score provides a quantitative indicator, which can more intuitively compare the degree of abnormality of different data points and enhance the operability of the analysis. The comprehensive anomaly evaluation score is compared with the preset threshold to identify abnormal customers or logistics links, and generate reports to provide decision support, help optimize logistics processes and customer service. By identifying anomalies and providing targeted suggestions, it can significantly improve logistics efficiency and enterprise terminal satisfaction, and help optimize the overall operational efficiency of the region.

[0092] Specifically, Figure 2 As shown, in step S30, the local outlier factor is calculated based on the neighborhood k obtained by cross-validation of the sampled data p, and the local outlier factor evaluation value LOF(p0) of the sampled data p is output, including the following steps:

[0093] Step S31: selecting a neighborhood k of an initial size; evaluating the accuracy of the local outlier factor evaluation value based on the neighborhood k through cross-validation based on the neighborhood k;

[0094] It should be noted that the above-mentioned embodiment of the present application is to collect sample sampling data with a preset local outlier factor evaluation value, and then divide the sample sampling data into a training set and a validation set with a fixed ratio (when the sample sampling data division is completed, the preset local outlier factor evaluation value is also divided into a training local outlier factor evaluation value and a validation local outlier factor evaluation value); further, based on the training set, through the k-fold cross-validation method, the influence of the neighborhood k on the preset local outlier factor evaluation value is evaluated (that is, the training local outlier factor evaluation value on each fold of the k folds is calculated), and then the validation local outlier factor evaluation value of the validation set is used to calculate the accuracy of the preset local outlier factor evaluation value of the training set;

[0095] Step S32: selecting an optimized neighborhood k based on the accuracy rate;

[0096] It should be noted that the above-mentioned embodiment of the present application selects the corresponding neighborhood k with the highest accuracy as the optimized neighborhood k.

[0097] Step S33: Calculating the local density ρ(i) of each sampled data p based on the optimized neighborhood k;

[0098] The calculation method of the local density ρ(i) of the field is:

[0099]

[0100] Where N k (i) is the neighborhood of the i-th sample data; d(ij) is the distance function, which represents the distance between sample data i and sample data j;

[0101] Step S34: Calculate the correlation sensitivity factor m of each of the sampled data based on the local density ρ(i) of the field i ;

[0102] The associated sensitivity factor m i The calculation method is:

[0103]

[0104] It should be noted that the above embodiment of the present application introduces the correlation sensitivity factor m i To represent the relative density of sampled data i in its neighborhood, quantify the influence of sampled data i on its neighborhood, and thus reflect the importance of sampled data i in the neighborhood.

[0105] Step S35: Based on the optimized neighborhood k and the associated sensitivity factor m i Calculate and obtain the local outlier factor evaluation value;

[0106] The local outlier factor evaluation value is calculated as follows:

[0107]

[0108] In the formula, m j is the correlation sensitivity factor of the i-th sampling data; m i is the correlation sensitivity factor of the i-th sampling data;

[0109] It should be noted that the above-mentioned embodiment of the present application optimizes the neighborhood size k by introducing cross-validation and enhances the accuracy of local outlier factors by calculating associated sensitive factors, which can effectively improve the stability and accuracy of anomaly detection; it not only improves the performance of the model, but also lays the foundation for subsequent comprehensive anomaly evaluation and logistics benefit analysis.

[0110] During the specific implementation of the above embodiment of the present application, the technicians found that the above comprehensive abnormality evaluation score was obtained by analyzing multi-source customer service data. Therefore, when the technical solution adopted by the above embodiment of the present application is directly used to perform logistics benefit evaluation processing operations, only the pre-sales data is evaluated and processed, but it is basically pre-sales analysis data, and it does not participate in the use of post-sales data;

[0111] After a period of use, a variable that affects the comprehensive abnormality evaluation score will occur (this variable is the logistics after-sales variable data information), and it is impossible to perform targeted processing operations on this variable; therefore, when performing logistics benefit evaluation processing operations based on the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node, it is necessary not only to consider the comprehensive abnormality evaluation score, but also to consider the impact of the logistics after-sales variable data information on the comprehensive abnormality evaluation score, so as to perform targeted logistics benefit evaluation processing operations based on the comprehensive abnormality evaluation score.

[0112] Specifically, Figure 3 As shown, in step S50, a logistics benefit evaluation processing operation is performed according to the comprehensive abnormal evaluation score of each data in the logistics multi-source customer service data at each segmented time node, including the following operation steps:

[0113] Step S51: Collect and obtain logistics after-sales variable information; the logistics after-sales variable information includes logistics satisfaction, return rate and logistics complaint rate, transfer efficiency evaluation coefficient of logistics transfer station, and delivery delay time;

[0114] It should be noted that in the above-mentioned embodiment of the present application, enterprise terminal satisfaction refers to the satisfaction score (such as 1 to 5 points) of customers (i.e., production enterprises in the region) for the service collected through questionnaires or online scoring systems; return rate refers to the proportion of return logistics orders to total logistics orders calculated within a certain period of time; logistics complaint rate refers to the ratio of the number of customer complaints received within a certain period of time to the total number of orders processed; the transit number evaluation coefficient of the logistics transfer station refers to recording the number of transit times of each transfer station and evaluating its impact on the overall logistics efficiency; delivery delay time refers to calculating the difference between the actual delivery time and the promised delivery time, and recording the number of delay days;

[0115] It should also be noted that after obtaining the logistics after-sales variable information, the logistics after-sales variable information needs to be preprocessed to make the data high-quality and consistent, thereby ensuring the accuracy of subsequent analysis.

[0116] Step S52: Calculating the Spearman rank correlation coefficient ρ based on the combination of the logistics after-sales variable information and the comprehensive abnormality assessment score;

[0117] It should be noted that the above-mentioned Spearman rank correlation coefficient ρ refers to the correlation coefficient of each element in the logistics after-sales variable information on the comprehensive abnormality evaluation score, such as the correlation coefficient of logistics satisfaction on the comprehensive abnormality evaluation score.

[0118] Step S53: Determine whether the Spearman rank correlation coefficient ρ is greater than 0 or less than 0; if so, perform a logistics benefit evaluation processing operation based on the logistics after-sales variable information corresponding to the Spearman rank correlation coefficient ρ and the comprehensive abnormality evaluation score; if not, perform a logistics benefit evaluation processing operation based on the comprehensive abnormality evaluation score.

[0119] It should be noted that, in the above-mentioned embodiment of the present application, the Spearman rank correlation coefficient ρ is less than 0, indicating that the comprehensive abnormality evaluation score is negatively correlated with a certain logistics after-sales variable (such as delivery delay time), and delayed delivery may be the main cause of the abnormality. Therefore, when performing the logistics benefit evaluation processing operation, the comprehensive abnormality evaluation score and the logistics after-sales variable information are comprehensively considered to perform the logistics benefit evaluation processing operation. For example, if the comprehensive abnormality evaluation score is negatively correlated with the delivery delay time, it is recommended to optimize the logistics process and reduce delivery delays;

[0120] When the Spearman rank correlation coefficient ρ is greater than 0, it means that the comprehensive abnormality evaluation score is positively correlated with a certain logistics after-sales variable (such as enterprise terminal satisfaction). Low satisfaction may lead to abnormalities. Therefore, when performing logistics benefit evaluation processing operations, the comprehensive abnormality evaluation score and the logistics after-sales variable information are comprehensively considered to perform logistics benefit evaluation processing operations. For example, the comprehensive abnormality evaluation score is positively correlated with enterprise terminal satisfaction. It is recommended to improve customer service quality and feedback mechanism;

[0121] Furthermore, when the Spearman rank correlation coefficient ρ is equal to 0, it proves that the logistics post-sales variable information has no effect on the comprehensive abnormality evaluation score, and proves that the comprehensive abnormality evaluation score is obtained only because of the pre-sales data in the aforementioned process. Therefore, when performing the logistics benefit evaluation processing operation, the logistics benefit evaluation processing operation is only performed on the comprehensive abnormality evaluation score (that is, there is no need to refer to this logistics post-sales variable information).

[0122] Specifically, Figure 4As shown, in step S52, the Spearman rank correlation coefficient ρ is calculated based on the combination of the logistics after-sales variable information and the comprehensive abnormality evaluation score, including the following operation steps:

[0123] Step S521: traverse each of the logistics after-sales variable information, and obtain a variable pair based on combining the logistics after-sales variable information and the comprehensive abnormality evaluation score corresponding to the logistics after-sales variable information;

[0124] It should be noted that the above embodiment of the present application is to obtain a variable pair by combining each element in a plurality of logistics after-sales variable information and the comprehensive abnormality evaluation score corresponding to each logistics after-sales variable information respectively;

[0125] For example, there are currently three logistics after-sales variable information, which are respectively represented as: after-sales 1 (logistics satisfaction 1, return rate and logistics complaint rate 1, logistics transfer station transfer efficiency evaluation coefficient 1, delivery delay time 1), after-sales 2 (logistics satisfaction 2, return rate and logistics complaint rate 2, logistics transfer station transfer efficiency evaluation coefficient 2, delivery delay time 2), after-sales 3 (logistics satisfaction 3, return rate and logistics complaint rate 3, logistics transfer station transfer efficiency evaluation coefficient 3, delivery delay time 3); the comprehensive abnormal evaluation scores corresponding to the three logistics after-sales variable information are comprehensive score 1, comprehensive score 2, and comprehensive score 3 respectively;

[0126] Furthermore, each element in each logistics after-sales variable information is combined with the comprehensive abnormality evaluation score to obtain a variable pair, such as Figure 5 As shown, it is expressed as: logistics satisfaction 1-comprehensive score 1, return rate and logistics complaint rate 1-comprehensive score 1, transportation efficiency evaluation coefficient of logistics transfer station 1-comprehensive score 1, delivery delay time 1-comprehensive score 1, logistics satisfaction 2-comprehensive score 2, return rate and logistics complaint rate 2-comprehensive score 2, transportation efficiency evaluation coefficient of logistics transfer station 2-comprehensive score 2, delivery delay time 2-comprehensive score 2, logistics satisfaction 3-comprehensive score 3…;

[0127] Step S522: sorting the values ​​of all the logistics after-sales variable information (the values ​​of the logistics after-sales variable information are score values ​​or other forms of descriptive quantitative values) from high to low to obtain a first variable sequence set; sorting the values ​​of the comprehensive abnormality evaluation scores from high to low to obtain a second variable sequence set;

[0128] It should be noted that the above embodiment of the present application is to sort the values ​​of each element in each logistics after-sales variable information respectively to obtain a first variable sequence set;

[0129] Continuing with the above example: the values ​​of the elements in after-sales 1 are logistics satisfaction 1=2, return rate and logistics complaint rate 1=3, logistics transfer station transfer efficiency evaluation coefficient 1=5, and delivery delay time 1=1; the values ​​of the elements in after-sales 2 are logistics satisfaction 2=3, return rate and logistics complaint rate 2=4, logistics transfer station transfer efficiency evaluation coefficient 2=6, and delivery delay time 2=2; the values ​​of the elements in after-sales 3 are logistics satisfaction 3=7, return rate and logistics complaint rate 3=6, logistics transfer station transfer efficiency evaluation coefficient 3=2, and delivery delay time 3=7; the values ​​of the three comprehensive abnormality evaluation scores are comprehensive score 1=8, comprehensive score 2=6, and comprehensive score 3=4;

[0130] According to the technical solution adopted in the above-mentioned embodiment of the present application, the values ​​of each element are sorted respectively to obtain a first variable sequence set and a second variable sequence set;

[0131] Sorting by the value of logistics satisfaction, the first variable sequence set is obtained, which is expressed as: first variable sequence set 1 = {logistics satisfaction 3 is 7, logistics satisfaction 2 is 3, logistics satisfaction 1 is 2}, and the first first variable sequence set 1 corresponding to logistics satisfaction is obtained;

[0132] The first variable sequence set is obtained by sorting the values ​​of return rate and logistics complaint rate, which is expressed as: first variable sequence set 2 = {return rate and logistics complaint rate 3 is 6, return rate and logistics complaint rate 2 is 4, return rate and logistics complaint rate 1 is 3}, and the second first variable sequence set 2 corresponding to the return rate and logistics complaint rate is obtained;

[0133] The first variable sequence set is obtained by sorting the values ​​of the transfer efficiency evaluation coefficients of the logistics transfer stations, which is expressed as: the first variable sequence set 3 = {the transfer efficiency evaluation coefficient 2 of the logistics transfer station is 6, the transfer efficiency evaluation coefficient 1 of the logistics transfer station is 5, and the transfer efficiency evaluation coefficient 3 of the logistics transfer station is 2}, and the third first variable sequence set 3 corresponding to the transfer efficiency evaluation coefficient of the logistics transfer station is obtained;

[0134] Sorting by the value of the delivery delay time, a first variable sequence set is obtained, which is expressed as: first variable sequence set 4 = {delivery delay time 3 is 7, delivery delay time 2 is 2, delivery delay time 1 is 1}, and a fourth first variable sequence set 4 corresponding to the delivery delay time is obtained;

[0135] Sorting by the value of the comprehensive abnormality assessment score, a second variable sequence set is obtained, which is expressed as: second variable sequence set = {comprehensive score 1 is 8, comprehensive score 2 is 6, comprehensive score 3 is 4};

[0136] Step S523: Calculating the ranking difference of the variable pair based on the first variable sequence set and the second variable sequence set;

[0137] It should be noted that the above embodiment of the present application is to obtain the ranking difference of the variable pair by subtracting the ranking of the comprehensive abnormality evaluation score corresponding to the value of the logistics after-sales variable information in the second variable sequence set from the ranking of the value of the logistics after-sales variable information in the first variable sequence set;

[0138] Continuing with the above example, according to the technical solution adopted in the above embodiment of the present application, the ranking difference corresponding to each first variable sequence set and the second variable sequence set is calculated, that is, the ranking difference of the variable pairs is calculated using the first first variable sequence set 1 and the second variable sequence set corresponding to the logistics satisfaction, wherein the variable pairs exist logistics satisfaction 1-comprehensive score 1, logistics satisfaction 2-comprehensive score 2, logistics satisfaction 3-comprehensive score 3; in the first variable sequence set 1, logistics satisfaction 1 ranks third, logistics satisfaction 2 ranks second, and logistics satisfaction 3 ranks third; in the second variable sequence set, comprehensive score 1 ranks first, comprehensive score 2 ranks second, and comprehensive score 3 ranks third, and the calculated ranking difference is: the ranking difference of the first variable to logistics satisfaction 1-comprehensive score 1 is |3-1|=2; the ranking difference of the second variable to logistics satisfaction 2-comprehensive score 2 is |2-2|=0; the ranking difference of the third variable to logistics satisfaction 3-comprehensive score 3 is |3-1|=2...;

[0139] Step S524: Calculate the Spearman rank correlation coefficient ρ based on the ranking difference;

[0140] The Spearman rank correlation coefficient ρ is calculated as:

[0141]

[0142] Where, d i is the ranking difference; n is the total value of the logistics after-sales variable information and the comprehensive abnormality evaluation score.

[0143] It should be noted that the above-mentioned embodiment of the present application first combines each logistics after-sales variable information (such as logistics satisfaction, return rate, etc.) with its corresponding comprehensive abnormality evaluation score to form a "variable pair", thereby laying the foundation for subsequent sorting and correlation analysis; further, the variables are sorted according to their values ​​and comprehensive abnormality evaluation scores, and then the relative deviation of each variable is quantified by comparing the ranking difference of each logistics after-sales variable in the two sequences; further, the Spearman rank correlation coefficient is used to quantify the correlation between the variable pairs, and the Spearman rank correlation coefficient is used to deeply understand the relationship between the comprehensive abnormality evaluation score and other key variables, so as to more effectively identify the causes of abnormal performance and formulate corresponding improvement measures.

[0144] Specifically, Figure 6 As shown, in step S51, obtaining the transfer efficiency evaluation coefficient of the logistics transfer station includes the following steps:

[0145] Step S511: Collect and obtain logistics transfer data; the logistics transfer data includes transfer station data M ij , the throughput P of each transfer station in the transfer station data jh , regional total throughput R iy and the transfer station coverage area P ij ;

[0146] It should be noted that, in the above-mentioned embodiment of the present application, the transfer station data M ij refers to the jth transfer station in the i-th area; the transfer station throughput P jh It refers to the throughput of items per unit time at the h-th transfer station in the j-th transfer station; the total regional throughput R iy Refers to the total throughput of items per unit time in the i-th area; the transfer station covers an area P ij It refers to the coverage area of ​​the jth transfer station in the i-th area;

[0147] Step S512: constructing a logistics transit matrix based on the logistics transit data;

[0148] The logistics transfer matrix includes the transfer station throughput matrix P, the regional total throughput matrix vector R and the coverage area matrix A;

[0149] Each element P in the transfer station throughput matrix P jh represents the unit time throughput of the hth transfer station in the jth transfer station; each element R in the regional total throughput matrix vector R iy represents the total item throughput per unit time in the i-th area; each element A in the coverage area matrix A ij represents the coverage area of ​​the jth transfer station in the i-th area;

[0150] Step S513: Calculate the transfer efficiency evaluation coefficient E of the logistics transfer station based on the transfer station throughput matrix P, the regional total throughput matrix vector R and the coverage area matrix A ij ;

[0151] The transfer efficiency evaluation coefficient E of the logistics transfer station ij The calculation method is:

[0152]

[0153] In the formula, ∑ h P jh R represents the sum of the throughput per unit time of all transfer stations in the jth transfer station;iy is the total item throughput of the i-th area; A ij is the coverage area of ​​the j-th transfer station in the i-th area.

[0154] It should be noted that the above-mentioned embodiment of the present application first collects data in steps S411 and S412 and converts it into a matrix form, which can ensure the comprehensiveness and operability of the data. The constructed matrix clearly shows the performance, regional requirements and coverage capabilities of each transfer station, so that subsequent analysis can be more detailed and accurate; further, through the calculation of step S413, the transfer efficiency of each transfer station is comprehensively evaluated.

[0155] Furthermore, during the specific implementation of the above-mentioned embodiment of the present application, the technicians also found that the acquisition of logistics multi-source customer service data will have more or less subjective factors due to the source of the data, such as logistics management system, customer relationship management system, social media, etc., and the transfer efficiency evaluation coefficient E of the above-mentioned logistics transfer station ij In essence, it can feedback and adjust the logistics multi-source customer service data, thereby improving logistics efficiency. For example, when the throughput of the transfer station is high, the transportation distance is shorter, and the customer feedback order response time is faster, thereby improving logistics efficiency and enterprise terminal satisfaction, which helps to optimize the overall operational efficiency of the region; therefore, when obtaining logistics multi-source customer service data, the transfer efficiency evaluation coefficient E of the above logistics transfer station can be considered. ij , so as to obtain more accurate logistics multi-source customer service data, thereby making the local outlier factor evaluation value LOF(p) more accurate, and thus making the subsequent logistics benefit evaluation processing operation better.

[0156] Specifically, after step S50, it also includes iteratively optimizing the logistics multi-source customer service data based on the transfer efficiency evaluation coefficient of the logistics transfer station to obtain updated logistics multi-source customer service data; and re-performing a logistics benefit evaluation processing operation based on the updated logistics multi-source customer service data;

[0157] Specifically, Figure 7 As shown, the iterative optimization of the logistics multi-source customer service data based on the transfer efficiency evaluation coefficient of the logistics transfer station to obtain updated logistics multi-source customer service data includes the following steps:

[0158] Step S501: Based on the transfer efficiency evaluation coefficient of the logistics transfer station and the logistics multi-source customer service data, multiple sub-factor combinations are obtained by combining sub-factors in pairs;

[0159] It should be noted that the above embodiment of the present application is to combine the sub-factors of the transfer efficiency evaluation coefficient of the logistics transfer station (i.e., the transfer station data M ij , the throughput P of each transfer station in the transfer station datajh , regional total throughput R iy and the transfer station coverage area P ij ) are combined in pairs with the sub-factors of logistics multi-source customer service data (i.e., order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal (i.e., logistics company) fees and logistics terminal efficiency index data) to obtain multiple sub-factor combinations, which can provide a data basis for subsequent analysis; this application will not go into details.

[0160] Step S502: Calculating the Pearson correlation coefficient r based on the value of the logistics transfer station transfer efficiency evaluation coefficient sub-factor in the sub-factor combination and the value of the logistics multi-source customer service data sub-factor;

[0161] The Pearson correlation coefficient r is calculated as follows:

[0162]

[0163] Where X is the value of the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station (for example, the value of the sub-factor is: a score value or other forms of descriptive quantitative values); Y is the value of the sub-factor in the logistics multi-source customer service data; is the average and sum of the sub-factors of the transfer efficiency evaluation coefficient of the logistics transfer station is the average of the values ​​of the sub-factors in the logistics multi-source customer service data;

[0164] Step S503: Screen and obtain sub-factor combinations corresponding to the Pearson correlation coefficient r being greater than or equal to a preset Pearson correlation coefficient threshold as target sub-factor combinations;

[0165] Step S504: adding the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station in the target sub-factor combination to the logistics multi-source customer service data to obtain updated logistics multi-source customer service data;

[0166] Then, based on the updated logistics multi-source customer service data, the above step S10 is returned to be executed again to obtain a comprehensive abnormality evaluation score, and a logistics benefit evaluation processing operation is performed according to the comprehensive abnormality evaluation score.

[0167] It should be noted that the above embodiment of the present application adds the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station in the target sub-factor combination to the logistics multi-source customer service data in the target sub-factor combination, and retains the sub-factor of the logistics multi-source customer service data in the target sub-factor combination, thereby obtaining the updated logistics multi-source customer service data; for example: a large logistics company A is responsible for logistics distribution in multiple cities across the country, has multiple transfer stations and a large number of customer orders. In order to improve efficiency, Company A decides to implement the above solution;

[0168] First, the sub-factors of the transfer efficiency evaluation coefficient of the logistics transfer station are combined with the sub-factors of the logistics multi-source customer service data in pairs to form multiple sub-factor combinations as the basis for subsequent analysis; further, by calculating the Pearson correlation coefficient, the company found that there is a strong correlation between certain factor combinations (such as "transfer station throughput" and "logistics terminal transportation path"). For example, it was found that the higher the transfer station throughput of certain transfer stations, the shorter the transportation path, and the faster the response time of customer orders, with a correlation coefficient of 0.85 (indicating a strong positive correlation); Company A set the threshold to 0.7 and only screened sub-factor combinations with correlation coefficients greater than or equal to 0.7, and finally screened out the combination of "transfer station throughput" and "transport path" as the target sub-factor combination; through this analysis, Company A added the throughput data of efficient transfer stations to the logistics multi-source customer service data, and used the updated data to help optimize customers' transportation paths and distribution arrangements, reducing unnecessary transfers and improving overall efficiency.

[0169] In summary, the example of the present invention proposes a regional business environment evaluation and analysis system and method based on big data analysis, firstly collects multi-source customer service data; further, by calculating the approximation between adjacent segmented time nodes, the abnormal growth points in the data are identified; by calculating the local outlier factor (LOF) of the sampled data, the degree of abnormality of the data point is evaluated. During the calculation process, the neighborhood size k is optimized by introducing cross-validation, and the accuracy of the local outlier factor is enhanced by calculating the associated sensitive factor, which can effectively improve the stability and accuracy of anomaly detection; not only improves the performance of the model, but also lays the foundation for subsequent comprehensive anomaly evaluation and logistics benefit analysis; at the same time, in the above scheme, based on the LOF and abnormal growth evaluation value, the comprehensive anomaly evaluation score of each data is calculated, the degree of abnormality of different data points is more intuitively compared, and the operability of the analysis is enhanced; the comprehensive anomaly evaluation score is compared with the preset threshold to identify abnormal customers or logistics links, and generate a report to provide decision support, help optimize logistics processes and customer service, and by identifying anomalies and providing targeted suggestions, it can significantly improve logistics efficiency and enterprise terminal satisfaction, help optimize the overall regional operational efficiency, and improve the regional business operation level;

[0170] When performing specific logistics benefit evaluation and analysis operations, the Spearman rank correlation coefficient ρ is obtained by introducing logistics after-sales variable information and comprehensive abnormal evaluation scores to perform targeted logistics benefit evaluation processing operations; in addition, the transfer efficiency evaluation coefficient of the logistics transfer station is introduced to iteratively optimize the logistics multi-source customer service data, thereby optimizing the regional logistics operation efficiency.

[0171] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. A person skilled in the art may modify the technical solutions described in the above embodiments, or replace part or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A regional business environment evaluation and analysis system based on big data analysis, characterized in that: include: Information collection platform, identification module, sampling calculation module, evaluation calculation module and output processing module; The information collection platform is used to collect logistics multi-source customer service data at subdivided time nodes within a preset time period; the logistics multi-source customer service data includes order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal fees and logistics terminal efficiency index data; The identification module is used to obtain the approximation of the logistics multi-source customer service data at each segmented time node based on the logistics multi-source customer service data at two adjacent segmented time nodes, and obtain the abnormal growth evaluation value G(p) of each data p' in the logistics multi-source customer service data at each segmented time node based on the approximation of the logistics multi-source customer service data at each segmented time node; The sampling calculation module is used to obtain the sampling data p of each data p' in the logistics multi-source customer service data of each subdivided time node, and then calculate the local outlier factor based on the neighborhood k obtained by cross-validation of the sampling data p, and output the local outlier factor evaluation value LOF(p) of the sampling data p; The evaluation and calculation module is used to calculate the comprehensive abnormal evaluation score E(p) of each data in the logistics multi-source customer service data at each segmented time node based on the local outlier factor evaluation value LOF(p) of the sampled data p and the abnormal growth evaluation value G(p) of each data; The output processing module is used to perform logistics benefit evaluation processing operations according to the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node.

2. According to the big data analysis regional business environment evaluation and analysis system according to claim 1, it is characterized by: The evaluation calculation module includes an initial processing module, a selection processing module, a domain calculation module, an association calculation module and an evaluation value calculation module: The initial processing module is used to select a neighborhood k of an initial size; and evaluate the accuracy of the local outlier factor evaluation value based on the neighborhood k through cross-validation based on the neighborhood k; The selection processing module is used to select the optimized neighborhood k based on the accuracy rate; The domain calculation module is used to calculate the domain local density ρ(i) of each of the sampled data p based on the optimized neighborhood k; The association calculation module is used to calculate the association sensitivity factor m of each of the sampled data based on the local density ρ(i) of the field i ; The evaluation value calculation module is used to calculate the value of the neighborhood k and the associated sensitivity factor m based on the optimized neighborhood k and the associated sensitivity factor m. i Calculate and obtain the local outlier factor evaluation value.

3. According to claim 2, a regional business environment evaluation and analysis system based on big data analysis is characterized in that: The calculation method of the local density ρ(i) of the field is: Where N k (i) is the neighborhood of the i-th sample data; d(ij) is the distance function, which represents the distance between sample data i and sample data j; The associated sensitivity factor m i The calculation method is: The local outlier factor evaluation value is calculated as follows: In the formula, m j is the correlation sensitivity factor of the i-th sampling data; m i is the correlation sensitivity factor of the i-th sampling data.

4. According to claim 3, a regional business environment evaluation and analysis system based on big data analysis is characterized in that: The output processing module includes an acquisition module, a Spearman rank calculation module and an evaluation output processing module: The acquisition module is used to collect and acquire logistics after-sales variable information; the logistics after-sales variable information includes logistics satisfaction, return rate and logistics complaint rate, transfer efficiency evaluation coefficient of logistics transfer station, and delivery delay time; The Spearman rank calculation module is used to calculate the Spearman rank correlation coefficient ρ based on the combination of the logistics after-sales variable information and the comprehensive abnormality assessment score; The evaluation output processing module is used to determine whether the Spearman rank correlation coefficient ρ is greater than 0 or less than 0; if so, a logistics benefit evaluation processing operation is performed based on the logistics after-sales variable information corresponding to the Spearman rank correlation coefficient ρ and the comprehensive abnormality evaluation score; if not, a logistics benefit evaluation processing operation is performed based on the comprehensive abnormality evaluation score.

5. According to claim 4, a regional business environment evaluation and analysis system based on big data analysis is characterized in that: The Spearman rank calculation module specifically includes: a traversal processing module and a first calculation module and a second calculation module; The traversal processing module is used to traverse each of the logistics after-sales variable information, and obtain a variable pair based on the combination of the logistics after-sales variable information and the comprehensive abnormality evaluation score corresponding to the logistics after-sales variable information; The first calculation module is used to obtain a first variable sequence set by sorting the values ​​of all the logistics after-sales variable information from high to low; and obtain a second variable sequence set by sorting the values ​​of the comprehensive abnormality evaluation scores from high to low; The second calculation module is used to calculate the ranking difference of the variable pair based on the first variable sequence set and the second variable sequence set; and calculate the Spearman rank correlation coefficient ρ based on the ranking difference; The Spearman rank correlation coefficient ρ is calculated as: Where, d i is the ranking difference; n is the total value of the logistics after-sales variable information and the comprehensive abnormality evaluation score.

6. According to claim 5, a regional business environment evaluation and analysis system based on big data analysis is characterized in that: The acquisition module includes a transfer data acquisition module, a matrix construction module and an evaluation coefficient calculation module: The transit data acquisition module is used to collect and obtain logistics transit data; the logistics transit data includes transit station data M ij , the throughput P of each transfer station in the transfer station data jh , regional total throughput R iy and the transfer station coverage area P ij ; The matrix construction module is used to construct a logistics transfer matrix based on the logistics transfer data; the logistics transfer matrix includes a transfer station throughput matrix P, a regional total throughput matrix vector R and a coverage area matrix A; Each element P in the transfer station throughput matrix P jh represents the unit time throughput of the hth transfer station in the jth transfer station; each element R in the regional total throughput matrix vector R iy represents the total item throughput per unit time in the i-th area; each element A in the coverage area matrix A ij represents the coverage area of ​​the jth transfer station in the i-th area; The evaluation coefficient calculation module is used to calculate the transfer efficiency evaluation coefficient E of the logistics transfer station based on the transfer station throughput matrix P, the regional total throughput matrix vector R and the coverage area matrix A. ij .

7. According to claim 6, a regional business environment evaluation and analysis system based on big data analysis is characterized in that: The transfer efficiency evaluation coefficient E of the logistics transfer station ij The calculation method is: In the formula, ∑ h P jh R represents the sum of the throughput per unit time of all transfer stations in the jth transfer station; iy is the total item throughput of the ith area; A ij is the coverage area of ​​the j-th transfer station in the i-th area.

8. The regional business environment evaluation and analysis system based on big data analysis according to claim 7 is characterized in that: It also includes an optimization processing module; the optimization processing module is used to iteratively optimize the logistics multi-source customer service data based on the transfer efficiency evaluation coefficient of the logistics transfer station to obtain updated logistics multi-source customer service data; and re-perform logistics benefit evaluation processing operations based on the updated logistics multi-source customer service data.

9. The regional business environment evaluation and analysis system based on big data analysis according to claim 8 is characterized in that: The optimization processing module specifically includes a sub-factor module, a Pearson correlation coefficient module, a target sub-factor combination processing module and an update processing module; The sub-factor module is used to obtain multiple sub-factor combinations by combining sub-factors in pairs based on the transfer efficiency evaluation coefficient of the logistics transfer station and the logistics multi-source customer service data; The Pearson correlation coefficient module is used to calculate the Pearson correlation coefficient r based on the value of the logistics transfer station's transfer efficiency evaluation coefficient sub-factor and the value of the logistics multi-source customer service data sub-factor in the sub-factor combination; The Pearson correlation coefficient r is calculated as follows: Where X is the value of the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station; Y is the value of the sub-factor in the logistics multi-source customer service data; is the average and value of the sub-factors of the transfer efficiency evaluation coefficient of the logistics transfer station is the average of the values ​​of the sub-factors in the logistics multi-source customer service data; The target sub-factor combination processing module is used to screen and obtain the sub-factor combination corresponding to the Pearson correlation coefficient r being greater than or equal to a preset Pearson correlation coefficient threshold as the target sub-factor combination; The update processing module is used to add the sub-factor of the transfer efficiency evaluation coefficient of the logistics transfer station in the target sub-factor combination to the logistics multi-source customer service data to obtain updated logistics multi-source customer service data.

10. A method for evaluating and analyzing regional business environment based on big data analysis, characterized in that: The steps are as follows: Collect logistics multi-source customer service data at subdivided time nodes within a preset time period; The logistics multi-source customer service data includes order data, logistics terminal coverage data, logistics terminal throughput data, logistics terminal transportation path data, customer feedback order response data, logistics terminal fees and logistics terminal efficiency index data; The approximation of the logistics multi-source customer service data at each segmented time node is obtained according to the logistics multi-source customer service data at two adjacent segmented time nodes, and the abnormal growth evaluation value G(p) of each data p' in the logistics multi-source customer service data at each segmented time node is obtained according to the approximation of the logistics multi-source customer service data at each segmented time node; Obtain the sampled data p of each data p' in the logistics multi-source customer service data of each segmented time node, then calculate the local outlier factor based on the neighborhood k obtained by cross-validation of the sampled data p, and output the local outlier factor evaluation value LOF(p) of the sampled data p; Based on the local outlier factor evaluation value LOF(p) of the sampled data p and the abnormal growth evaluation value G(p) of each data, the comprehensive abnormal evaluation score E(p) of each data in the logistics multi-source customer service data at each segmented time node is calculated; Logistics benefit evaluation processing operations are performed based on the comprehensive abnormality evaluation score of each data in the logistics multi-source customer service data at each segmented time node.

Images