Index data self-inspection robot system for marketing business early warning and monitoring

Abstract

The invention relates to the technical field of data analysis, in particular to an index data self-inspection robot system for marketing business early warning and monitoring, and the system comprises a robot marketing data collection module which obtains various sales data including sales in a preset analysis interval; the robot marketing data classification module is used for decomposing the sequence, dividing a period in combination with fluctuation characteristics of the data, and then performing classification; the robot marketing data analysis module obtains sales heterogeneity according to difference characteristics of data in different categories; comparing sequence decomposition results to obtain marketing data anomaly; obtaining sales data consistency according to the similarity of the marketing data anomalies of different analysis intervals; and the robot marketing data early warning monitoring module is used for monitoring and early warning enterprise marketing business states according to the sales data consistency. The invention aims to improve the early warning and monitoring accuracy of the robot on the marketing business based on the index data.

Description

Technical Field

[0001] The present application relates to the technical field of data analysis, and in particular to an indicator data self-checking robot system for early warning monitoring of marketing business. Background Art

[0002] Marketing is the forefront of a company's business services. Marketing business early warning monitoring is a system that monitors marketing activities and related business data in real time. It aims to identify potential problems and abnormal trends and issue warnings in advance so that the company can take prompt action to avoid or mitigate negative impacts. The monitoring system usually includes continuous tracking of some indicators, such as sales, transaction volume, etc., and detects anomalies in the data through statistical analysis of the indicators.

[0003] In the marketing business field, existing methods are usually limited to a single data dimension, focusing only on the abnormal characteristics of the current marketing business, while ignoring the relative characteristics of marketing data and the impact of changes in marketing strategies within a time period. This limitation leads to the low accuracy of traditional marketing trend analysis in warning of marketing business problems, and also weakens the reliability of indicator data self-checking. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides an indicator data self-inspection robot system for marketing business early warning monitoring to solve the existing problems.

[0005] The present application provides an indicator data self-checking robot system for marketing business early warning monitoring, the system comprising:

[0006] Robot marketing data collection module: used to take a sequence consisting of various sales data including sales within a preset analysis interval as various sales data sequences;

[0007] Robotic marketing data classification module: used to decompose the sales data sequence corresponding to the sales amount, and divide all sales data sequences into periods based on the fluctuation characteristics of the sales data; all sales data at the same time in each period are regarded as a multidimensional point, and the multidimensional points in each period are classified;

[0008] Robotic marketing data analysis module: used to obtain the sales heterogeneity of each cycle according to the difference characteristics of sales data in different categories in each cycle; to obtain the abnormality of marketing data of each cycle by combining the results of sequence decomposition of each cycle with the other cycles and the sales heterogeneity; to obtain the consistency of sales data of each analysis interval according to the similarity of the abnormality of marketing data of each analysis interval and all cycles of the other analysis intervals;

[0009] Robot marketing data early warning monitoring module: used to monitor and warn the enterprise marketing business status in the next analysis interval based on the consistency of sales data in all analysis intervals.

[0010] The specific steps of dividing all sales data series into periods include:

[0011] All peak points in the period item of the result of the sequence decomposition are obtained, and the time period corresponding to two consecutive peak points is taken as a period of various sales data.

[0012] The classification of the multidimensional points in each cycle is specifically as follows:

[0013] The multidimensional points in each cycle are divided into two categories according to the distance measurement between the multidimensional points. The category with more multidimensional points is recorded as the first category, and the other category is recorded as the second category.

[0014] The sales heterogeneity of each cycle is obtained as follows:

[0015] For each period, the coefficient of data difference for each period was obtained based on the difference in the number of multidimensional points between the two categories;

[0016] Based on the distance distribution between the same sales data of two categories in each period, the purchase difference coefficient of each period is obtained;

[0017] The data variance coefficient of each period is merged with the purchase variance coefficient to obtain the sales heterogeneity of each period.

[0018] The purchase difference coefficient of each period is obtained as follows:

[0019] Based on the distance measurement between the various sales data sequences of the second category and the first category, the relative differences of various sales data are obtained; based on the difference between the relative difference corresponding to the sales volume of each period and the other relative differences, the purchase difference coefficient of each period is obtained.

[0020] The abnormality of the marketing data in each cycle is obtained as follows:

[0021] For each analysis interval, the sales trend similarity coefficient and the sales trend difference coefficient of each period are obtained according to the trend item of the result of the sequence decomposition of each period and the other periods and the sales heterogeneity;

[0022] The inverse proportional mapping result of the sales trend difference coefficient and the sales trend similarity coefficient of each period is used as the abnormality of the marketing data of each period.

[0023] The obtaining of the sales trend similarity coefficient and the sales trend difference coefficient of each period includes:

[0024] For each analysis period, the sales trend similarity coefficient of each period is obtained based on the similarity between the corresponding trend items of sales of each period and all other periods; the sales trend difference coefficient of each period is obtained based on the difference in sales heterogeneity between each period and all other periods.

[0025] The sales data consistency of each analysis interval is specifically the inverse proportional mapping result of the distance distribution between the abnormality of all marketing data between each analysis interval and all other analysis intervals.

[0026] The monitoring and early warning of the enterprise marketing business status in the next analysis interval is specifically as follows:

[0027] According to the distribution of sales data consistency of all analysis intervals, the sales data consistency of the next analysis interval is predicted;

[0028] When the normalized result of the sales data consistency in the next analysis interval is less than or equal to the preset abnormality threshold, the robot issues an early warning; otherwise, the robot does not issue an early warning.

[0029] The specific steps of predicting the consistency of sales data in the next analysis interval include:

[0030] The sales data consistency of all analysis intervals is used as the input of the sequence prediction algorithm to obtain the sales data consistency of the next analysis interval.

[0031] This application has at least the following beneficial effects:

[0032] The present application first obtains a variety of sales data in a preset time period and forms a sequence; provides a data basis for the subsequent feature analysis of sales data in different time periods; decomposes the sequence, which helps to extract more accurate sequence features, divides the data sequence into periods, measures the periodic characteristics of data changes and fluctuations, and can improve the accuracy of subsequent data predictions; takes all sales data at the same time in each period as a multidimensional point, uses visualization technology to more intuitively display the relationship and pattern between the data, classifies the multidimensional points in each period, and measures the deviation of different data in the same time; constructs sales heterogeneity, considers the impact of marketing strategies in different time periods and the associated change characteristics of multidimensional marketing data; constructs marketing data anomaly, conducts associated trend feature analysis on marketing business data in each time period, and reflects the impact of marketing characteristics on marketing business data in different time periods; constructs sales data consistency, further considers the regular characteristics affected by marketing strategies in different time periods, and then accurately reflects the abnormal characteristics of marketing business data, accurately predicts the indicator data for the next time period, and improves the accuracy of the self-inspection robot system for marketing business warnings. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A system block diagram of an indicator data self-checking robot system for marketing business early warning monitoring provided by one embodiment of the present application;

[0035] Figure 2 A schematic diagram of a flow chart for obtaining sales data consistency provided in one embodiment of the present application;

[0036] Figure 3 A specific flow chart for monitoring and warning the marketing business status of an enterprise provided in one embodiment of the present application. DETAILED DESCRIPTION

[0037] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail the specific implementation method, structure, features and effects of an indicator data self-checking robot system for marketing business early warning monitoring proposed in the present application. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0038] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0039] The following is a detailed description of a specific scheme of an indicator data self-checking robot system for marketing business early warning monitoring provided by the present application in conjunction with the accompanying drawings.

[0040] See also Figure 1 , which shows a block diagram of an indicator data self-inspection robot system for marketing business early warning monitoring provided by an embodiment of the present application, the system includes: a robot marketing data collection module 101, a robot marketing data classification module 102, a robot marketing data analysis module 103, and a robot marketing data early warning monitoring module 104.

[0041] The robot marketing data collection module 101 takes a sequence consisting of various sales data including sales amount within a preset analysis interval as various sales data sequences.

[0042] This application uses the retail sales data of an enterprise as an example to identify and warn of possible abnormal data. It should be noted that the implementer can select data related to marketing in other industries and use the same method as this application for analysis.

[0043] The sales data of the enterprise within a preset time period is collected, including sales revenue, product sales quantity and customer flow, and the sequences consisting of sales revenue, sales volume and customer flow in the preset analysis interval in ascending chronological order are recorded as sales revenue sequence, sales volume sequence and customer flow sequence respectively.

[0044] In some embodiments of the present application, the preset time period is a set of all non-holidays whose time interval with the current moment is less than or equal to the preset time period, the preset time period is 1 year, and the data collection interval is 1 day; 30 days are used as an analysis interval; considering that a large increase in sales during the holiday time period is an expected and reasonable phenomenon, then these peaks are not comparable with other data, and therefore the sales peaks generated by holidays cannot be regarded as outliers, so the sales data of non-holiday days in the preset time period are collected; implementers can adjust it at their own discretion.

[0045] Since missing values ​​may occur due to external factors during the collection process, some embodiments of the present application use the isnull() function of the Pandas library to judge the missing values ​​of the above sequence. If there are missing values ​​in the sequence, the mean filling method is used to fill the missing values. It should be noted that on the basis of implementing the above missing value filling, other methods such as median filling or mode filling can also be used, and the implementer can choose at will.

[0046] Robot marketing data classification module 102: Sequence decomposition of the sales data sequence corresponding to the sales volume is performed, and all sales data sequences are divided into periods in combination with the fluctuation characteristics of the sales data; all sales data at the same time in each period are taken as a multidimensional point, and the multidimensional points in each period are classified.

[0047] Since sales data is crucial to the strategic decision-making and resource allocation of enterprises, it can help enterprises optimize product portfolios, adjust market strategies and sales channels, and changes in sales can reflect market trends. Sales data may show periodic changes due to different shopping habits of consumers in different time periods, such as different purchasing patterns on weekends and weekdays. Based on this, all sales data series are divided into periods according to sales series:

[0048] The sales series is decomposed to obtain the trend item and the cycle item of the sales series; all the peak points in the cycle item are obtained, and the time period corresponding to two consecutive peak points is regarded as a cycle of various sales data.

[0049] It should be noted that there are many methods for sequence decomposition and acquisition of peak points. The implementer can choose a suitable method to obtain them, and this application does not impose any restrictions on this.

[0050] In some embodiments of the present application, a seasonal and trend decomposition using Loess (STL) algorithm is used for sequence decomposition. The input of the STL algorithm is the sales sequence of each year, and the output is the trend term, periodic term and residual term of the sales sequence. Each item obtained by decomposition is a time series with the same length as the decomposed sequence. Among them, the periodic term shows the periodicity of the change of the sales sequence, and the automatic multiscale-based peak detection (AMPD) algorithm is used to obtain the peak point in the periodic term. It should be noted that the STL sequence decomposition and AMPD algorithm are well-known technologies, and the specific process will not be repeated.

[0051] Under normal circumstances, there is a certain correlation between various sales data in marketing business. For example, when the customer flow is large on a certain day, the customer purchase rate is more likely to increase, and then the sales volume and sales amount will increase. When the customer flow is small, the sales volume and sales amount will usually be small. It may be affected by the fluctuation of the average unit price or purchase rate of customers. Although different types of sales data do not show a completely strict increase and decrease in the same proportion, the general trend of change between different types of sales data is the same. Based on this, all sales data at the same time in each cycle is regarded as a multidimensional point. According to the distance measurement between the multidimensional points, the multidimensional points in each cycle are divided into two categories. The category with more multidimensional points is recorded as the first category, and the other category is recorded as the second category.

[0052] In some embodiments of the present application, the sales volume, sales volume and passenger flow data of the same day constitute a three-dimensional point, and all three-dimensional points in a cycle are mapped to a three-dimensional rectangular coordinate system. In the normal state of the marketing business, the distance between the three-dimensional points is relatively close, and the distribution will be relatively concentrated. In the abnormal state of the marketing business, for example, the passenger flow in a certain day is large and the sales volume is small, resulting in the three-dimensional point corresponding to this day deviating from other three-dimensional points. Therefore, the fuzzy C-means clustering algorithm of machine learning is used to classify the three-dimensional points of each cycle. Among them, the number of clusters is set to 2, that is, the three-dimensional points of each cycle are divided into two categories. It should be noted that the implementer may also choose other classification methods to classify the three-dimensional points, and this application does not limit this; in addition, the fuzzy C-means clustering algorithm is an existing well-known technology and will not be repeated in this application.

[0053] Furthermore, the category with a large number of three-dimensional points is recorded as the first category, and the category with a small number of three-dimensional points is recorded as the second category.

[0054] Robot marketing data analysis module 103: According to the difference characteristics of sales data in different categories in each cycle, the sales heterogeneity of each cycle is obtained; the results of the sequence decomposition of each cycle and the other cycles are integrated with the sales heterogeneity to obtain the abnormality of the marketing data of each cycle; according to the similarity of the abnormality of the marketing data of each analysis interval and all the cycles of the other analysis intervals, the consistency of the sales data of each analysis interval is obtained.

[0055] For each period, the data difference coefficient of each period is obtained according to the difference in the number of multidimensional points between the second category and the first category; the relative difference of various sales data is obtained according to the distance measurement between various sales data sequences of the second category and the first category; the purchase difference coefficient of each period is obtained according to the difference between the relative difference corresponding to the sales volume of each period and the other relative differences; the data difference coefficient of each period is merged with the purchase difference coefficient to obtain the sales heterogeneity of each period.

[0056] It should be noted that the difference represents the numerical difference between variables, specifically: difference, absolute value of difference, ratio, etc.; the distance measurement between sequences is a method of measuring the similarity or difference of sequences, specifically: Euclidean distance, Manhattan distance, DTW distance, etc.; the methods of fusing variables include: addition, multiplication, and mixed addition and multiplication operations; this application does not impose any restrictions on this.

[0057] In some embodiments of the present application, the ratio of the number of multidimensional points in the second category to that in the first category in each cycle is used as the data difference coefficient of each category; the size of the number of multidimensional points in the first category reflects the number of business days under normal conditions, and the larger the data difference coefficient, the more days with abnormal business in a cycle.

[0058] In addition, under normal marketing business conditions, the customer purchase rate is relatively stable, that is, when the customer flow is large, the sales volume should be large. Therefore, by analyzing the difference between the sales data of the two categories, the difference in the stability of the customer purchase rate is measured, and then the business status within the cycle is reflected. The sequences of the sales data in the first category and the second category in ascending time order are recorded as the first subsequence of sales and the second subsequence of sales, respectively, and the DTW distance between the first subsequence of sales and the second subsequence of sales is taken as the relative difference in sales; accordingly, the relative difference in sales volume and the relative difference in customer flow are obtained. The absolute values ​​of the differences between the relative difference in sales volume and the relative difference in sales volume and the relative difference in customer flow are calculated respectively, and then the average is calculated to obtain the purchase difference coefficient for each cycle. The larger the purchase difference coefficient, the more unstable the customer purchase rate corresponding to the first category and the second category in each cycle.

[0059] Furthermore, the product of the data variance coefficient and the purchase variance coefficient of each period is taken as the sales heterogeneity of each period. The greater the sales heterogeneity, the more unstable the sales status of the business in that period.

[0060] It should be understood that the increase in sales heterogeneity may be caused by changes in the company's marketing strategy or differences in operational management, such as random promotional activities, insufficient supply of goods, etc., which in turn lead to random rapid growth or decline in sales. If the sales series corresponding to consecutive cycles show similar fluctuation characteristics, for example, the company's marketing strategy causes the data at specific locations in different cycles to show short-term rapid changes, this feature may appear abnormal in one cycle. If the short-term rapid change feature occurs regularly, it can be explained as a normal phenomenon. Therefore, the smaller the difference in the sales coefficient of variation between different cycles, the more regular the short-term rapid change feature of sales appears.

[0061] For each analysis interval, the sales trend similarity coefficient of each period is obtained based on the similarity between the corresponding trend items of sales of each period and all other periods; the sales trend difference coefficient of each period is obtained based on the difference in sales heterogeneity between each period and all other periods; the inverse proportional mapping result of the sales trend difference coefficient of each period and the sales trend similarity coefficient is used as the marketing data abnormality of each period.

[0062] It should be noted that the calculation of similarity between sequences specifically includes: Pearson correlation coefficient, cosine similarity, Jaccard similarity, etc.; usually, ratios, reciprocals, etc. are used to measure the inverse proportional mapping results between variables; this application does not impose any restrictions on this.

[0063] In some embodiments of the present application, the average of the absolute values ​​of the cosine similarities of the corresponding trend items of the sales series of each period and all other periods is calculated as the sales trend similarity coefficient of each period; then the average of the absolute values ​​of the differences in sales heterogeneity between each period and all other periods is taken as the sales trend difference coefficient of each period.

[0064] The ratio of the sales trend difference coefficient to the sales trend similarity coefficient of each period is taken as the abnormality of the marketing data of each period.

[0065] It should be understood that the larger the sales trend similarity coefficient, the more similar the overall trend changes of sales series in different periods. The smaller the sales trend difference value, the more likely it is that sales series in different periods may have similar variation states, the more regular the characteristics of short-term rapid changes in sales are, and the marketing business may be relatively normal. The larger the sales trend difference value, the more likely it is that the characteristics of short-term rapid changes in sales only appear in some periods, the more abnormal the marketing business may be, and the greater the abnormality of the marketing data, indicating that the possibility of abnormality in the marketing data is greater.

[0066] The anomaly of marketing data focuses on the business data status within the sales cycle. Furthermore, the sales between different analysis intervals have certain similar characteristics as a whole:

[0067] The inverse proportional mapping result of the distance distribution of all marketing data abnormalities between each analysis interval and all other analysis intervals is used as the sales data consistency of each analysis interval.

[0068] In some embodiments of the present application, the abnormality of marketing data of all cycles in each analysis interval is used to form a marketing data characterization sequence in ascending order of time. The marketing business characterization sequence reflects the characteristics of the sales change state of the enterprise within the analysis interval. Since the consumption habits of users are closer in different time periods, the reciprocal of the average Jaccard distance between each analysis interval in the collected data and the corresponding marketing data characterization sequence of all other analysis intervals is calculated as the sales data consistency of each analysis interval.

[0069] Among them, the schematic diagram of the process of obtaining sales data consistency is as follows: Figure 2 shown.

[0070] It should be understood that the higher the consistency of the sales data obtained, the better the overall marketing status of the company in each analysis period. If the consistency of the sales data is low, there may be problems with the company's marketing business in each analysis period.

[0071] The robot marketing data early warning monitoring module 104 monitors and warns the enterprise marketing business status in the next analysis interval based on the consistency of sales data in all analysis intervals.

[0072] This application analyzes the sales data series, sales volume series, and customer flow series of the enterprise, and calculates the consistency of sales data in each year. The consistency of sales data reflects the possibility of abnormalities in the marketing strategy or operation process of the enterprise to a certain extent. The sales data consistency of all analysis intervals is predicted using the ARIMA model to predict the sales data consistency of the next analysis interval, and the Z-Score method is used for normalization.

[0073] It should be noted that the implementer may choose other methods to predict the consistency of sales data in the next analysis interval, or may choose other methods to normalize the data, and this application does not impose any restrictions on this.

[0074] Conduct early warning monitoring of the enterprise's marketing business based on the normalized results of the sales data consistency in the next analysis period;

[0075] If the consistency of the normalized sales data is less than or equal to the preset abnormal threshold, it means that there is an abnormality in the marketing business status. The robot will issue an early warning to remind relevant company staff to adjust the marketing business in a timely manner. If the consistency of the normalized sales data is greater than the preset abnormal threshold, it means that the marketing business status is normal and the robot will not issue an early warning.

[0076] In some embodiments of the present application, the abnormal threshold value is 0.6; the implementer can adjust it at will.

[0077] Among them, the specific flow chart for monitoring and early warning of the enterprise marketing business status is as follows: Figure 3 shown.

[0078] In summary, the embodiment of the present application first obtains various sales data of a preset time period and forms a sequence; provides a data basis for the subsequent feature analysis of sales data in different time periods; decomposes the sequence, which helps to extract more accurate sequence features, divides the data sequence into periods, measures the periodic characteristics of data changes and fluctuations, and can improve the accuracy of subsequent data predictions; takes all sales data at the same time in each period as a multidimensional point, uses visualization technology to more intuitively display the relationship and pattern between the data, classifies the multidimensional points in each period, and measures the deviation of different data in the same time; constructs sales heterogeneity, considers the impact of marketing strategies in different time periods and the associated change characteristics of multidimensional marketing data; constructs marketing data anomaly, performs associated trend feature analysis on marketing business data in each time period, and reflects the impact of marketing characteristics on marketing business data in different time periods; constructs sales data consistency, further considers the regular characteristics affected by marketing strategies in different time periods, and then accurately reflects the abnormal characteristics of marketing business data, accurately predicts the indicator data of the next time period, and improves the accuracy of the self-inspection robot system for marketing business warnings.

[0079] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0080] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0081] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Modifications to the technical solutions recorded in the aforementioned embodiments, or equivalent replacement of some of the technical features therein, 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 application, and should all be included in the protection scope of the present application.

Claims

1. An indicator data self-checking robot system for marketing business early warning monitoring, characterized in that: The system includes: Robot marketing data collection module: used to take a sequence consisting of various sales data including sales within a preset analysis interval as various sales data sequences; Robotic marketing data classification module: used to decompose the sales data sequence corresponding to the sales amount, and divide all sales data sequences into periods based on the fluctuation characteristics of the sales data; all sales data at the same time in each period are regarded as a multidimensional point, and the multidimensional points in each period are classified; Robotic marketing data analysis module: used to obtain the sales heterogeneity of each cycle according to the difference characteristics of sales data in different categories in each cycle; to obtain the abnormality of marketing data of each cycle by combining the results of sequence decomposition of each cycle with the other cycles and the sales heterogeneity; to obtain the consistency of sales data of each analysis interval according to the similarity of the abnormality of marketing data of each analysis interval and all cycles of the other analysis intervals; Robot marketing data early warning monitoring module: used to monitor and warn the enterprise marketing business status in the next analysis interval based on the consistency of sales data in all analysis intervals.

2. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 1, characterized in that: The specific steps of dividing all sales data series into periods include: All peak points in the period item of the result of the sequence decomposition are obtained, and the time period corresponding to two consecutive peak points is taken as a period of various sales data.

3. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 1, characterized in that: The multidimensional points in each cycle are classified as follows: The multidimensional points in each cycle are divided into two categories according to the distance measurement between the multidimensional points. The category with more multidimensional points is recorded as the first category, and the other category is recorded as the second category.

4. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 3, characterized in that: The sales heterogeneity of each cycle is obtained as follows: For each period, the coefficient of data difference for each period was obtained based on the difference in the number of multidimensional points between the two categories; Based on the distance distribution between the same sales data of two categories in each period, the purchase difference coefficient of each period is obtained; The data variance coefficient of each period is merged with the purchase variance coefficient to obtain the sales heterogeneity of each period.

5. The indicator data self-checking robot system for marketing business early warning monitoring as claimed in claim 4, characterized in that: The purchase difference coefficient of each period is obtained as follows: Based on the distance measurement between the various sales data sequences of the second category and the first category, the relative differences of various sales data are obtained; based on the difference between the relative difference corresponding to the sales volume of each period and the other relative differences, the purchase difference coefficient of each period is obtained.

6. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 1, characterized in that: The abnormality of the marketing data in each cycle is obtained as follows: For each analysis interval, the sales trend similarity coefficient and the sales trend difference coefficient of each period are obtained according to the trend item of the result of the sequence decomposition of each period and the other periods and the sales heterogeneity; The inverse proportional mapping result of the sales trend difference coefficient and the sales trend similarity coefficient of each period is used as the abnormality of the marketing data of each period.

7. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 6, characterized in that: The sales trend similarity coefficient and sales trend difference coefficient of each period are obtained, including: For each analysis period, the sales trend similarity coefficient of each period is obtained based on the similarity between the corresponding trend items of sales of each period and all other periods; the sales trend difference coefficient of each period is obtained based on the difference in sales heterogeneity between each period and all other periods.

8. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 1, characterized in that: The sales data consistency of each analysis interval is specifically an inverse proportional mapping result of the distance distribution between all marketing data abnormalities between each analysis interval and all other analysis intervals.

9. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 1, characterized in that: The monitoring and early warning of the enterprise marketing business status in the next analysis interval is specifically as follows: According to the distribution of sales data consistency of all analysis intervals, the sales data consistency of the next analysis interval is predicted; When the normalized result of the sales data consistency in the next analysis interval is less than or equal to the preset abnormality threshold, the robot issues an early warning; otherwise, the robot does not issue an early warning.

10. The indicator data self-checking robot system for marketing business early warning monitoring according to claim 9, characterized in that: The specific steps of predicting the consistency of sales data in the next analysis interval include: The sales data consistency of all analysis intervals is used as the input of the sequence prediction algorithm to obtain the sales data consistency of the next analysis interval.

Images