Sales strategy adjustment system and method based on omni-channel sales data

Abstract

The invention provides a sales strategy adjustment system and method based on omni-channel sales data, and the system comprises a sales data obtaining module which is used for obtaining the sales data of a commodity, and the sales data comprises a commodity ID, a sales channel, and a time period; the sales data preprocessing module is used for preprocessing the sales data to generate to-be-clustered data; the clustering analysis module is used for carrying out clustering analysis according to the to-be-clustered data and determining sales strategies of all categories, and the sales strategies comprise the stock amount of commodities, the number of stock turnover days, sales targets and sales prices; and the sales strategy optimization module is used for further optimizing and determining an optimal sales strategy according to the sales strategy determined by the clustering analysis module. According to the invention, the optimal configuration of the inventory, the inventory turnover days, the sales target and the sales price can be realized, and the sales efficiency and the profitability are improved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and more specifically, to a sales strategy adjustment system and method based on omni-channel sales data. Background Art

[0002] Existing sales strategy adjustment systems often lack the ability to comprehensively analyze omni-channel sales data, resulting in the inability to accurately capture the sales dynamics of products in different sales channels and time periods. These systems usually rely on a single source of sales data and cannot fully reflect the performance of products in different sales environments. Due to the lack of in-depth analysis of multi-dimensional data, these systems have limitations in inventory management, price setting, sales target setting, etc., and cannot provide companies with flexible and accurate sales strategies. In addition, traditional sales strategy adjustment methods often require a lot of manual intervention, are inefficient, and have difficulty adapting to the rapidity of market changes. In terms of technical principles, these systems usually do not use advanced data mining and machine learning technologies, cannot achieve automated clustering analysis and strategy optimization, and thus cannot effectively extract valuable information from massive sales data to guide sales decisions.

[0003] In the process of implementing the embodiments of the present invention, the inventors found that there are at least the following problems or defects in the prior art: the existing sales strategy adjustment system and method cannot effectively integrate and analyze omni-channel sales data, resulting in a lack of data support for the formulation of sales strategies, and cannot achieve dynamic adjustment to adapt to market changes. It is also inefficient when processing large-scale data and is difficult to meet the real-time and accuracy requirements of the modern retail environment. Summary of the invention

[0004] The present invention provides a sales strategy adjustment system and method based on omni-channel sales data.

[0005] In a first aspect of the present invention, a sales strategy adjustment system based on omni-channel sales data is provided, comprising:

[0006] The sales data acquisition module is used to obtain the sales data of the product, including: product ID, sales channel, and time period;

[0007] A sales data preprocessing module, used for preprocessing sales data to generate data to be clustered;

[0008] The cluster analysis module is used to perform cluster analysis based on the data to be clustered and determine the sales strategy for each category. The sales strategy includes: the inventory volume of the goods, the inventory turnover days, the sales target, and the sales price;

[0009] The sales strategy optimization module is used to further optimize and determine the optimal sales strategy based on the sales strategy determined by the cluster analysis module.

[0010] Furthermore, the sales data preprocessing module includes:

[0011] De-duplicate the sales data and group the de-duplicated sales data according to the time period to obtain the sales ID vector of each product sales channel within the time period.

[0012] S: S=(s1,t1),(s2,t2),…,(sn,tn)

[0013] Where n is the number of sales of a certain product in a time period, si∈N+, ti∈channel;

[0014] Channels include: shopping malls, e-commerce, physical stores, N+ is a positive integer set;

[0015] Set the dimension of the data to be clustered, and determine the data to be clustered X∈Rm according to the sales vector S, where m is the total number of products.

[0016] Furthermore, the cluster analysis module includes:

[0017] The cluster analysis generation submodule is used to obtain a cluster diagram, which includes: different categories of goods and their corresponding sales strategies;

[0018] The analysis result submodule is used to obtain the analysis results of the clustering graph.

[0019] Furthermore, the cluster analysis generation submodule includes:

[0020] Randomly generate K initial points, where K is the total number of points in the data to be clustered;

[0021] Calculate the weighted distance between the points in the data to be clustered and the initial points;

[0022] Divide each point into the cluster where the nearest initial point is located, so that the K initial points are divided into K cluster centers and K clusters are generated;

[0023] Recalculate the cluster center in each cluster;

[0024] Recalculate the weighted distance between each point and the cluster center in the cluster;

[0025] If the recalculated cluster center is the same as the cluster center calculated in the previous round, the calculation ends;

[0026] Otherwise, repeat the above steps until the calculation result of the cluster center meets the threshold requirement.

[0027] Furthermore, the calculation formula for the weighted distance between the calculated point and the initial point is:

[0028]

[0029] Among them, d i are two points x in the data to be clustered X i ,x j The weighted distance between ik For point x i Corresponding to the k-th dimension data, x jk For point x j Corresponding to the data of the kth dimension; ω k For two points x i ,x j The weight of the product ID in the kth dimension.

[0030] Furthermore, the analysis result submodule includes:

[0031] Clustering result judgment submodule, clustering result analysis submodule;

[0032] The clustering result judgment submodule is used to judge whether the clustering result of the clustering graph is accurate. If so, the clustering result analysis submodule is triggered;

[0033] If not, the cluster center reconstruction module is triggered;

[0034] The cluster center reconstruction module is used to re-cluster analysis until the cluster result judgment submodule determines whether the clustering result of the cluster graph is accurate, and then triggers the cluster result analysis submodule;

[0035] The clustering result analysis submodule is used to select a clustering diagram whose weighted distance between the cluster center and the data to be clustered is greater than a set value according to the clustering diagram, and determine the sales strategy corresponding to the cluster center in the clustering diagram, wherein the sales strategy includes: sales price, inventory turnover days, inventory quantity, and sales target;

[0036] The sales strategy of the cluster center is the sales strategy for products in the same category.

[0037] Furthermore, the cluster center reconstruction module is used for re-clustering analysis, and the specific steps include:

[0038] Re-clustering analysis:

[0039] Step I: Initialize the sales price of the goods in each cluster, calculate the average price of the goods in the cluster; record the average sales price of the goods in the cluster as the sales price;

[0040] Step II: recalculate the distance between each cluster based on the sales price;

[0041] The distance between the clusters is calculated as follows:

[0042] Step i, add the commodities in cluster 1 to cluster 2 in sequence, calculate the average value between the price of each commodity in cluster 1 and the sum of the prices of commodities in cluster 2, and use this value as the distance between cluster 1 and cluster 2 after the first addition;

[0043] Step ii, add the next product in cluster 1 to cluster 2 in sequence, calculate the average of the price of each product in cluster 1 and the price of the products in cluster 2, and obtain the distance between cluster 1 and cluster 2;

[0044] Step iii, after the sales price of the goods in cluster 1 is added to cluster 2, the goods with the cheapest sales price in cluster 1 are used as the centroid of cluster 1 and cluster 2;

[0045] Step iv, repeat step iii until the distance change between cluster 1 and cluster 2 is less than a set value;

[0046] Step III, repeat steps I to II, obtain the distance value between any two clusters, select the largest one as the number of clusters, and re-cluster according to the number of clusters until the clustering result judgment submodule determines whether the clustering result of the clustering graph is accurate.

[0047] Furthermore, the process of determining the optimal sales strategy includes the following steps:

[0048] Calculate the weighted distance mean of each category in the data to be clustered and the cluster center. The calculation formula for the weighted distance mean of the data to be clustered is:

[0049]

[0050] In the formula, D is the weighted distance mean of the data to be clustered, a i is the weight, d i is the distance from point i to the cluster center, and n is the number of sales of the product within a time period; obtain the sales price of each category; calculate the difference between the sales price of each category and the average price of the sales price of the cluster center; use the category with the smallest difference, its cluster center's product inventory, turnover days, and sales target as the sales strategy for the product to be sold.

[0051] Furthermore, the cluster diagram includes:

[0052] Coordinate axis, horizontal axis, vertical axis; the horizontal axis is the product ID information, the vertical axis is the sales price of the product, and the coordinate axis is the inventory of the product;

[0053] Each point is associated with the inventory quantity and sales price of its corresponding product.

[0054] In a second aspect of the present invention, a sales strategy adjustment method based on omni-channel sales data is provided, comprising:

[0055] Step S1, obtaining sales data of the product, the sales data including: product ID, sales channel, and time period;

[0056] Step S2, preprocessing the sales data to generate data to be clustered;

[0057] Step S3, performing cluster analysis based on the data to be clustered, and determining sales strategies for each category, the sales strategies including: inventory volume, inventory turnover days, sales target, and sales price of the goods;

[0058] Step S4, performing cluster analysis based on the data to be clustered, and determining the sales strategy of the sales channel corresponding to each category;

[0059] Step S5, determining the optimal sales strategy according to the sales price, inventory turnover days and sales target of each category of sales strategy.

[0060] The above-mentioned embodiments of the present invention have at least the following beneficial effects: by using omni-channel sales data for comprehensive analysis and clustering, it is possible to accurately capture the sales dynamics of goods in different sales channels and time periods, and provide enterprises with more accurate and flexible sales strategies. This method can optimize inventory management, dynamically adjust sales prices, and formulate more reasonable sales targets, improve sales efficiency and market response speed. At the same time, through automated clustering analysis and strategy optimization, the system can reduce manual intervention, reduce costs, and improve the efficiency and accuracy of data processing.

[0061] In addition, the present invention can ensure the accuracy and stability of clustering results by calculating the weighted distance between the point and the initial point, and applying the cluster center reconstruction module. This precise cluster analysis can further refine the sales strategy, so that the same category of goods can obtain customized sales strategies according to their sales characteristics. This method can improve the pertinence and effectiveness of sales strategies, enhance the market competitiveness of enterprises, and provide decision makers with more intuitive and clear sales data analysis results to assist them in making more scientific business decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0063] Figure 1 A schematic diagram of the structure of a sales strategy adjustment system based on omni-channel sales data provided by an embodiment of the present invention;

[0064] Figure 2A flow chart of a sales strategy adjustment method based on omni-channel sales data provided by an embodiment of the present invention;

[0065] Figure 3 The schematic diagram schematically shows the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0067] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present invention can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0068] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0069] Reference below Figure 1 , Figure 1 This is a schematic diagram of the structure of a sales strategy adjustment system based on omni-channel sales data provided by an embodiment of the present invention. Figure 1 As shown, a sales strategy adjustment system 100 based on omni-channel sales data includes:

[0070] The sales data acquisition module 101 is used to acquire the sales data of the commodity, and the sales data includes: commodity ID, sales channel, and time period;

[0071] The sales data preprocessing module 102 is used to preprocess the sales data to generate data to be clustered;

[0072] The cluster analysis module 103 is used to perform cluster analysis based on the data to be clustered and determine the sales strategy for each category. The sales strategy includes: the inventory quantity of the goods, the inventory turnover days, the sales target, and the sales price;

[0073] The sales strategy optimization module 104 is used to further optimize and determine the optimal sales strategy based on the sales strategy determined by the cluster analysis module.

[0074] It should be noted that the system includes a sales data acquisition module, whose function is to collect sales data of goods, which covers the product ID, sales channel and time period. Here, sales data refers to all information related to the sales of goods, including but not limited to sales volume, sales volume, customer feedback, etc., while sales channels include shopping malls, e-commerce, physical stores and other sales channels.

[0075] Specifically, the sales data acquisition module can collect data through interface docking, data import or web crawler technology. For example, for e-commerce platforms, sales data can be directly obtained through the API interface; for physical stores, data may need to be exported through the POS system. In terms of parameter settings, the system can set time periods such as days, weeks, months, etc. according to user needs, as well as product ID identification rules and sales channel classification standards.

[0076] Preferably, the sales data acquisition module can also include data verification and cleaning functions to ensure that the collected data is accurate. For example, the system can automatically remove data records with outliers or missing key information, and format the data so that subsequent preprocessing and analysis work can proceed smoothly. In addition, the system can also set up an automatic update mechanism to adapt to market changes and data updates.

[0077] In some embodiments, the sales data preprocessing module includes:

[0078] De-duplicate the sales data and group the de-duplicated sales data according to the time period to obtain the sales ID vector of each product sales channel within the time period.

[0079] S: S=(s1,t1),(s2,t2),…,(sn,tn)

[0080] Where n is the number of sales of a certain product in a time period, si∈N+, ti∈channel;

[0081] Channels include: shopping malls, e-commerce, physical stores, N+ is a positive integer set;

[0082] Set the dimension of the data to be clustered, and determine the data to be clustered X∈Rm according to the sales vector S, where m is the total number of products.

[0083] It should be noted that the sales data preprocessing module of the system is responsible for processing the collected sales data in order to generate data to be clustered. Here, data to be clustered refers to a data set that has been preprocessed and is ready for clustering analysis. The purpose of preprocessing is to remove noise and inconsistency in the data and to convert the data into a format suitable for analysis.

[0084] Specifically, the sales data preprocessing module first deduplicates the sales data to ensure that the sales record of each product is unique. Then, the deduplicated sales data is grouped according to the time period to obtain the sales ID vector of each product in different sales channels. The time period here can be of any length, for example, it can be one day, one week, or one month, depending on business needs and the purpose of data analysis. The sales ID vector refers to a vector consisting of the sales records of each product in different sales channels within a specific time period, where each element represents a sales event and contains the product ID and sales channel information.

[0085] Preferably, the sales data preprocessing module may also include data standardization and normalization steps. Data standardization refers to scaling the data so that it falls into a small specific interval, such as [0,1]. Normalization refers to adjusting the data values ​​to a state where the mean is 0 and the standard deviation is 1. These steps help improve the accuracy and efficiency of cluster analysis.

[0086] Furthermore, the preprocessing module can also set parameters to determine the dimensions of the data to be clustered, which is usually set based on the total number of products to ensure that each dimension of the data can represent the characteristics of a product. In actual operation, these parameters can be adjusted according to the characteristics of the product and business needs to achieve the best clustering effect.

[0087] In some embodiments, the cluster analysis module includes:

[0088] The cluster analysis generation submodule is used to obtain a cluster diagram, which includes: different categories of goods and their corresponding sales strategies;

[0089] The analysis result submodule is used to obtain the analysis results of the clustering graph.

[0090] It should be noted that the system includes a cluster analysis module, whose function is to perform cluster analysis based on preprocessed sales data and determine sales strategies for each category. Cluster analysis is a statistical method that divides samples in a data set into several groups, with the goal of making objects in the same group more similar, while objects in different groups less similar. Sales strategies include key factors such as inventory volume, inventory turnover days, sales targets, and sales prices.

[0091] Specifically, the cluster analysis module includes a cluster analysis generation submodule and an analysis result submodule. The cluster analysis generation submodule is responsible for generating a cluster diagram, which is a visualization tool for displaying different categories of goods and their corresponding sales strategies. The analysis result submodule is used to obtain the analysis results of the cluster diagram for further decision support.

[0092] More specifically, in terms of parameter settings, you can select a suitable clustering algorithm, such as K-means, hierarchical clustering, etc., according to business needs and data characteristics, and set algorithm parameters, such as the number of clusters K value, and distance metric standards.

[0093] Preferably, the operation steps of the cluster analysis module can be further refined. For example, in the cluster analysis generation submodule, a random initialization method can be used to determine the initial cluster center, and then the position of the cluster center can be adjusted by an iterative optimization algorithm until the convergence condition is met. In the analysis result submodule, machine learning technology can be introduced to improve the accuracy of the clustering results, such as using a support vector machine (SVM) or a neural network to verify and adjust the clustering results.

[0094] Furthermore, it is also possible to consider introducing interactive data visualization tools so that non-technical users can intuitively understand and analyze clustering results, thereby better supporting the formulation of sales strategies.

[0095] In some embodiments, the cluster analysis generation submodule includes:

[0096] Randomly generate K initial points, where K is the total number of points in the data to be clustered;

[0097] Calculate the weighted distance between the points in the data to be clustered and the initial points;

[0098] Divide each point into the cluster where the nearest initial point is located, so that the K initial points are divided into K cluster centers and K clusters are generated;

[0099] Recalculate the cluster center in each cluster;

[0100] Recalculate the weighted distance between each point and the cluster center in the cluster;

[0101] If the recalculated cluster center is the same as the cluster center calculated in the previous round, the calculation ends;

[0102] Otherwise, repeat the above steps until the calculation result of the cluster center meets the threshold requirement.

[0103] It should be noted that the system specifically includes a cluster analysis generation submodule in the cluster analysis module, which is used to obtain a cluster diagram, which includes different categories of goods and their corresponding sales strategies. Here, the cluster diagram is a graphical representation method for displaying the results of cluster analysis, including the classification of goods and the characteristics of each category, such as sales strategies.

[0104] Specifically, the cluster analysis generation submodule first randomly generates K initial points, where K represents the total number of points in the data to be clustered. These initial points are used as the starting center points of clustering for the subsequent clustering process. Next, the weighted distance between the points in the data to be clustered and these initial points is calculated, and each point is divided into the cluster where the nearest initial point is located, forming K clusters.

[0105] More specifically, in terms of parameter setting, the calculation of weighted distance involves the allocation of weights, which can be determined according to different attributes of the goods (such as sales volume, price, etc.) to reflect the importance of each attribute in the clustering process.

[0106] Preferably, the operation steps of the cluster analysis generation submodule can be further refined. For example, when calculating the weighted distance, different distance measurement methods such as Euclidean distance or Manhattan distance can be used. The specific method can be selected according to the characteristics of the data and business requirements. In the clustering process, a threshold can be set to determine whether the cluster center is stable. If the change of the cluster center in several consecutive iterations is less than this threshold, it can be considered that the clustering process has converged.

[0107] Furthermore, if the clustering results are not satisfactory, you can consider using different clustering algorithms, such as DBSCAN or AGNES, to improve the clustering effect and accuracy. These algorithms may be more effective in processing data sets of different shapes and densities.

[0108] In some embodiments, the weighted distance between the calculated point and the initial point is calculated as follows:

[0109]

[0110] Among them, d i are two points x in the data to be clustered X i ,x j The weighted distance between ik For point x i Corresponding to the k-th dimension data, x jk For point x j Corresponding to the data of the kth dimension; ω k For two points x i ,x j The weight of the product ID in the kth dimension.

[0111] It should be noted that the system pays special attention to the calculation formula of the weighted distance between the calculated point and the initial point in the cluster analysis generation submodule. Here, weighted distance means that in cluster analysis, considering the different importance of data of different dimensions, each dimension in the distance formula is weighted to obtain a comprehensive measurement value. This formula is a key step in cluster analysis and is used to determine the similarity between data points.

[0112] Specifically, the formula for calculating the weighted distance is:

[0113]

[0114] Among them, d ij Represents two points x in the data to be clustered X i and x j The weighted distance between ik For point x i Corresponding to the k-th dimension data, x jk For point x j Corresponding to the data of the kth dimension; w k For two points x i and x j The weight corresponding to the product ID in the kth dimension. In parameter setting, the weight w k The weight can be set according to the importance of the product ID in the sales data. For example, if the sales volume of a certain product ID has a greater impact on the sales strategy, a higher weight can be given.

[0115] Preferably, the operation step of calculating the weighted distance can be further refined or an alternative solution can be provided. k The setting can be based on statistical analysis of historical sales data and determined by calculating the contribution of each product ID to the total sales.

[0116] Furthermore, in addition to Euclidean distance, other types of distance metrics can also be considered, such as cosine similarity or Manhattan distance, which may be more appropriate when processing different types of data. In practical applications, the most appropriate distance metric can be selected according to the characteristics of the data and business needs, and the weights can be adjusted dynamically to improve the accuracy and effect of clustering analysis.

[0117] In some embodiments, the analysis result submodule includes:

[0118] Clustering result judgment submodule, clustering result analysis submodule;

[0119] The clustering result judgment submodule is used to judge whether the clustering result of the clustering graph is accurate. If so, the clustering result analysis submodule is triggered;

[0120] If not, the cluster center reconstruction module is triggered;

[0121] The cluster center reconstruction module is used to re-cluster analysis until the cluster result judgment submodule determines whether the clustering result of the cluster graph is accurate, and then triggers the cluster result analysis submodule;

[0122] The clustering result analysis submodule is used to select a clustering diagram whose weighted distance between the cluster center and the data to be clustered is greater than a set value according to the clustering diagram, and determine the sales strategy corresponding to the cluster center in the clustering diagram, wherein the sales strategy includes: sales price, inventory turnover days, inventory quantity, and sales target;

[0123] The sales strategy of the cluster center is the sales strategy for products in the same category.

[0124] It should be noted that the system further includes an analysis result submodule in the cluster analysis module, whose function is to process the analysis results of the cluster graph. Here, the analysis result submodule refers to the component in the system responsible for evaluating the accuracy of the cluster results and triggering subsequent analysis accordingly. The cluster result judgment submodule and the cluster result analysis submodule are two key parts of the analysis result submodule, which work together to ensure the effectiveness of the cluster results and determine the sales strategy accordingly.

[0125] Specifically, the clustering result judgment submodule in the analysis result submodule is used to judge whether the clustering result of the clustering graph is accurate. If the clustering result is judged to be accurate, the clustering result analysis submodule is triggered for further analysis; if it is inaccurate, the cluster center reconstruction module is triggered for re-clustering analysis.

[0126] More specifically, the clustering result analysis submodule is responsible for selecting cluster diagrams whose weighted distance between the cluster center and the data to be clustered is greater than the set value based on the cluster diagram, and determining the sales strategy corresponding to the cluster center. The weighted distance here means that in the clustering process, considering the different importance of data of different dimensions, each dimension in the distance formula is weighted to obtain a comprehensive measurement value.

[0127] Preferably, the operation steps of the analysis result submodule can be further refined or alternative solutions can be provided. For example, the clustering result judgment submodule can use statistical methods, such as Silhouette Coefficient or Davies-Bouldin Index, to quantify the compactness and separation of the clustering results, thereby evaluating the accuracy of the clustering results.

[0128] Furthermore, the cluster center reconstruction module can use different clustering algorithms, such as K-means++, to improve the selection of initial cluster centers and reduce the deviation of clustering results. In the cluster result analysis submodule, a dynamic threshold can be set, which can be automatically adjusted according to historical data and market changes to meet the clustering needs in different situations. These refined steps and alternatives help to improve the flexibility and adaptability of cluster analysis and ensure that sales strategies are formulated more accurately and effectively.

[0129] In some embodiments, the cluster center reconstruction module is used for re-clustering analysis, and the specific steps include:

[0130] Re-clustering analysis:

[0131] Step I: Initialize the sales price of the goods in each cluster, calculate the average price of the goods in the cluster; record the average sales price of the goods in the cluster as the sales price;

[0132] Step II: recalculate the distance between each cluster based on the sales price;

[0133] The distance between the clusters is calculated as follows:

[0134] Step i, add the commodities in cluster 1 to cluster 2 in sequence, calculate the average value between the price of each commodity in cluster 1 and the sum of the prices of commodities in cluster 2, and use this value as the distance between cluster 1 and cluster 2 after the first addition;

[0135] Step ii, add the next product in cluster 1 to cluster 2 in sequence, calculate the average of the price of each product in cluster 1 and the price of the products in cluster 2, and obtain the distance between cluster 1 and cluster 2;

[0136] Step iii, after the sales price of the goods in cluster 1 is added to cluster 2, the goods with the cheapest sales price in cluster 1 are used as the centroid of cluster 1 and cluster 2;

[0137] Step iv, repeat step iii until the distance change between cluster 1 and cluster 2 is less than a set value;

[0138] Step III, repeat steps I to II, obtain the distance value between any two clusters, select the largest one as the number of clusters, and re-cluster according to the number of clusters until the clustering result judgment submodule determines whether the clustering result of the clustering graph is accurate.

[0139] It should be noted that the system is specially designed with a cluster center reconstruction module, which is used to re-perform cluster analysis when the clustering results are inaccurate. Here, the cluster center reconstruction module refers to a component within the system that is responsible for optimizing the clustering process by adjusting the cluster centers when the initial cluster analysis fails to produce satisfactory results. This module is part of the analysis result submodule and ensures the accuracy and reliability of the cluster analysis.

[0140] Specifically, the operation steps of the cluster center reconstruction module include several stages of re-clustering analysis. First, step I involves initializing the sales price of the goods in each cluster and calculating the average price of the goods in the cluster, which is recorded as the sales price. Step II is to recalculate the distance between the clusters based on the sales price, which includes adding the goods in cluster 1 to cluster 2 in turn and calculating the average price between the two as a measure of distance. Step III is to repeat steps I and II until the accuracy of the clustering results is confirmed. In terms of parameter setting, a threshold can be set to determine whether the change in the distance between clusters is small enough to decide whether to stop the clustering process.

[0141] Preferably, the operation steps of the cluster center reconstruction module can be further refined or alternatives can be provided. For example, in step I, when initializing the sales price, the median or mode can be considered as the initial value, rather than just the average, to reduce the impact of outliers. In step II, when calculating the distance between clusters, a more complex distance measurement method can be used, such as the Mahalanobis distance, which takes into account the covariance structure of the data.

[0142] Furthermore, a dynamic adjustment mechanism can be introduced to automatically adjust the threshold according to the results of the previous clustering, making the clustering process more flexible and adaptable. These refined steps and alternatives help improve the accuracy and efficiency of clustering analysis and ensure that the final sales strategy is more in line with the actual market situation.

[0143] In some embodiments, the process of determining the optimal sales strategy includes the following steps:

[0144] Calculate the weighted distance mean of each category in the data to be clustered and the cluster center. The calculation formula for the weighted distance mean of the data to be clustered is:

[0145]

[0146] In the formula, D is the weighted distance mean of the data to be clustered, a i is the weight, d i is the distance from point i to the cluster center, and n is the number of sales of the product within a time period; obtain the sales price of each category; calculate the difference between the sales price of each category and the average price of the sales price of the cluster center; use the category with the smallest difference, its cluster center's product inventory, turnover days, and sales target as the sales strategy for the product to be sold.

[0147] It should be noted that in the process of determining the optimal sales strategy, the system involves calculating the weighted distance mean and cluster center of the data to be clustered. Here, the weighted distance mean refers to the distance from each data point to the cluster center, and these distances are weighted averaged according to the number of sales of each point. This method can more accurately reflect the distribution of goods in different categories, thereby providing more accurate data support for the formulation of sales strategies.

[0148] Specifically, the process of determining the optimal sales strategy includes the following steps: First, the weighted distance mean and cluster center of each category in the data to be clustered are calculated, where the calculation formula of the weighted distance mean is:

[0149]

[0150] In the formula, is the weighted distance mean of the data to be clustered, w i is the weight, d i is the distance from the point to the cluster center, and n is the number of sales of the product in the time period. Next, obtain the sales price of each category and calculate the difference between the sales price of each category and the average price of the cluster center sales price. The sales price here refers to the pricing of the product in different sales channels, and the average price refers to the average sales price of the product category represented by the cluster center.

[0151] Preferably, the steps for determining the optimal sales strategy can be further refined or alternatives can be provided. For example, when calculating the weighted distance mean, a time decay factor can be introduced to reflect that recent sales data has a greater impact on the current sales strategy.

[0152] Furthermore, machine learning methods such as random forests or gradient boosting machines can be used to predict sales prices and optimize sales strategies accordingly. When determining the optimal sales strategy, external factors such as market trends, seasonal factors, and promotional activities can also be considered to achieve more comprehensive and dynamic sales strategy adjustments. These detailed steps and alternatives help improve the adaptability and effectiveness of sales strategies and ensure competitiveness in a changing market environment.

[0153] In some embodiments, the cluster graph includes:

[0154] Coordinate axis, horizontal axis, vertical axis; the horizontal axis is the product ID information, the vertical axis is the sales price of the product, and the coordinate axis is the inventory of the product;

[0155] Each point is associated with the inventory quantity and sales price of the corresponding product.

[0156] It should be noted that the system involves the concepts of coordinate axis, horizontal axis, and vertical axis in the construction of cluster diagram. Here, cluster diagram is a data visualization tool used to show the relationship between product ID information, sales price, and inventory. The horizontal axis represents the product ID information, the vertical axis represents the sales price of the product, and the coordinate axis represents the inventory of the product. This graphical representation method can help users intuitively understand the performance and position of different products in the sales strategy.

[0157] Specifically, the construction of the clustering diagram includes the following steps: First, determine the product ID information on the horizontal axis, which can be the product number, name or other unique identifier. Second, determine the product sales price on the vertical axis, which can be the pricing of the product in different sales channels. Finally, determine the inventory on the coordinate axis, which can be the storage quantity of the product in the warehouse. In terms of parameter setting, the scales of the horizontal and vertical axes can be adjusted according to the sales data and inventory data of the product to ensure the accuracy and readability of the clustering diagram.

[0158] Preferably, the operation steps of the cluster diagram can be further refined or alternative solutions can be provided. For example, when displaying product ID information, color coding or icons can be used to distinguish different categories of products to improve the information richness of the cluster diagram. When indicating the selling price, different marks (such as dots, squares, etc.) can be used to indicate different price ranges, so that users can quickly identify price levels. When displaying inventory levels, consider using three-dimensional graphics or heat maps to more intuitively display the level of inventory.

[0159] Furthermore, interactive features can be introduced to allow users to zoom in or out on specific areas of the cluster map by clicking or dragging, so as to analyze the sales and inventory of goods in more detail. These refined steps and alternatives help to improve the practicality and user-friendliness of the cluster map, making the formulation of sales strategies more intuitive and efficient.

[0160] The above-mentioned embodiments of the present invention have the following beneficial effects: the system and method described in the present invention can effectively integrate and analyze data from different sales channels, including product ID, sales channel and time period, so as to provide more accurate sales strategies for product sales. Through the sales data preprocessing module, the system can remove duplicate data and group them by time period, and generate sales ID vectors for each product sales channel, which helps to analyze sales trends and patterns in more detail. The application of the cluster analysis module can determine different categories of products and their corresponding sales strategies, including inventory, inventory turnover days, sales targets and sales prices, so as to achieve personalization and optimization of sales strategies.

[0161] In addition, the system and method can optimize the process of determining sales strategies. By calculating the weighted distance mean of the data to be clustered and the cluster center, and analyzing the difference between the sales price of each category and the sales price of the cluster center, the system can identify the optimal sales strategy. This method can ensure that the sales strategies of the same category of goods are highly relevant and applicable. At the same time, through the cluster center reconstruction module, the system can adjust inaccurate clustering results until the most accurate cluster diagram is found, further enhancing the effectiveness and adaptability of the sales strategy.

[0162] like Figure 2 As shown, in some embodiments, a sales strategy adjustment method 200 based on omni-channel sales data includes:

[0163] Step S1, obtaining sales data of the product, the sales data including: product ID, sales channel, and time period;

[0164] Step S2, preprocessing the sales data to generate data to be clustered;

[0165] Step S3, performing cluster analysis based on the data to be clustered, and determining sales strategies for each category, the sales strategies including: inventory volume, inventory turnover days, sales target, and sales price of the goods;

[0166] Step S4, performing cluster analysis based on the data to be clustered, and determining the sales strategy of the sales channel corresponding to each category;

[0167] Step S5, determining the optimal sales strategy according to the sales price, inventory turnover days and sales target of each category of sales strategy.

[0168] It is understandable that the steps and references in the sales strategy adjustment method 200 based on omni-channel sales data are the same. Figure 1 Therefore, the modules, features and beneficial effects described above for the sales strategy adjustment system based on omni-channel sales data are also applicable to the sales strategy adjustment method 200 based on omni-channel sales data and the operations contained therein, which will not be repeated here.

[0169] Reference below Figure 3 , which shows a schematic diagram of a structure 300 of an electronic device suitable for implementing some embodiments of the present invention. The electronic devices in some embodiments of the present invention may include, but are not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0170] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0171] Typically, the following devices may be connected to the I / O interface 305: input devices 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 308 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 309. The communication devices 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0172] Furthermore, the storage medium of the embodiment of the present application stores program instructions that can implement all the above methods, wherein the program instructions can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0173] The above descriptions are only some preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention to form a technical solution.

Claims

1. A sales strategy adjustment system based on omni-channel sales data, characterized in that: The system comprises: The sales data acquisition module is used to obtain the sales data of the product, including: product ID, sales channel, and time period; A sales data preprocessing module, used for preprocessing sales data to generate data to be clustered; The cluster analysis module is used to perform cluster analysis based on the data to be clustered and determine the sales strategy for each category. The sales strategy includes: the inventory volume of the goods, the inventory turnover days, the sales target, and the sales price; The sales strategy optimization module is used to further optimize and determine the optimal sales strategy based on the sales strategy determined by the cluster analysis module.

2. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 1, characterized in that: The sales data preprocessing module includes: De-duplicate the sales data and group the de-duplicated sales data according to the time period to obtain the sales ID vector of each product sales channel within the time period. S: S=(s1,t1),(s2,t2),…,(sn,tn) Where n is the number of sales of a certain product in a time period, si∈N+, ti∈channel; Channels include: shopping malls, e-commerce, physical stores, N+ is a positive integer set; Set the dimension of the data to be clustered, and determine the data to be clustered X∈Rm according to the sales vector S, where m is the total number of products.

3. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 1, characterized in that: The cluster analysis module includes: The cluster analysis generation submodule is used to obtain a cluster diagram, which includes: different categories of goods and their corresponding sales strategies; The analysis result submodule is used to obtain the analysis results of the clustering graph.

4. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 3, characterized in that: The cluster analysis generation submodule includes: Randomly generate K initial points, where K is the total number of points in the data to be clustered; Calculate the weighted distance between the points in the data to be clustered and the initial points; Divide each point into the cluster where the nearest initial point is located, so that the K initial points are divided into K cluster centers and K clusters are generated; Recalculate the cluster center in each cluster; Recalculate the weighted distance between each point and the cluster center in the cluster; If the recalculated cluster center is the same as the cluster center calculated in the previous round, the calculation ends; Otherwise, repeat the above steps until the calculation result of the cluster center meets the threshold requirement.

5. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 4, characterized in that: The calculation formula for the weighted distance between the calculated point and the initial point is: Among them, d i are two points x in the data to be clustered X i ,x j The weighted distance between ik For point x i Corresponding to the k-th dimension data, x jk For point x j Corresponding to the data of the kth dimension; ω k For two points x i ,x j The weight of the product ID in the kth dimension.

6. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 3, characterized in that: The analysis result submodule includes: Clustering result judgment submodule, clustering result analysis submodule; The clustering result judgment submodule is used to judge whether the clustering result of the clustering graph is accurate. If so, the clustering result analysis submodule is triggered; If not, the cluster center reconstruction module is triggered; The cluster center reconstruction module is used to re-cluster analysis until the cluster result judgment submodule determines whether the clustering result of the cluster graph is accurate, and then triggers the cluster result analysis submodule; The clustering result analysis submodule is used to select a clustering diagram whose weighted distance between the cluster center and the data to be clustered is greater than a set value according to the clustering diagram, and determine the sales strategy corresponding to the cluster center in the clustering diagram, wherein the sales strategy includes: sales price, inventory turnover days, inventory quantity, and sales target; The sales strategy of the cluster center is the sales strategy for products in the same category.

7. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 6, characterized in that: The cluster center reconstruction module is used for re-clustering analysis. The specific steps include: Re-clustering analysis: Step I: Initialize the sales price of the goods in each cluster, calculate the average price of the goods in the cluster; record the average sales price of the goods in the cluster as the sales price; Step II: recalculate the distance between each cluster based on the sales price; The distance between the clusters is calculated as follows: Step i, add the commodities in cluster 1 to cluster 2 in sequence, calculate the average value between the price of each commodity in cluster 1 and the sum of the prices of commodities in cluster 2, and use this value as the distance between cluster 1 and cluster 2 after the first addition; Step ii, add the next product in cluster 1 to cluster 2 in sequence, calculate the average of the price of each product in cluster 1 and the price of the products in cluster 2, and obtain the distance between cluster 1 and cluster 2; Step iii, after the sales price of the goods in cluster 1 is added to cluster 2, the goods with the cheapest sales price in cluster 1 are used as the centroid of cluster 1 and cluster 2; Step iv, repeat step iii until the distance change between cluster 1 and cluster 2 is less than a set value; Step III, repeat steps I to II, obtain the distance value between any two clusters, select the largest one as the number of clusters, and re-cluster according to the number of clusters until the clustering result judgment submodule determines whether the clustering result of the clustering graph is accurate.

8. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 6, characterized in that: The process of determining the optimal sales strategy includes the following steps: Calculate the weighted distance mean of each category in the data to be clustered and the cluster center. The calculation formula for the weighted distance mean of the data to be clustered is: In the formula, D is the weighted distance mean of the data to be clustered, a i is the weight, d i is the distance from point i to the cluster center, and n is the number of sales of the product within a time period; obtain the sales price of each category; calculate the difference between the sales price of each category and the average price of the sales price of the cluster center; use the category with the smallest difference, its cluster center's product inventory, turnover days, and sales target as the sales strategy for the product to be sold.

9. A sales strategy adjustment system based on omni-channel sales data as claimed in claim 3, characterized in that: The cluster diagram includes: Coordinate axis, horizontal axis, vertical axis; the horizontal axis is the product ID information, the vertical axis is the sales price of the product, and the coordinate axis is the inventory of the product; Each point is associated with the inventory quantity and sales price of its corresponding product.

10. A sales strategy adjustment method based on omni-channel sales data, characterized in that: The method comprises the following steps: Step S1, obtaining sales data of the product, the sales data including: product ID, sales channel, and time period; Step S2, preprocessing the sales data to generate data to be clustered; Step S3, performing cluster analysis based on the data to be clustered, and determining sales strategies for each category, the sales strategies including: inventory volume, inventory turnover days, sales target, and sales price of the goods; Step S4, performing cluster analysis based on the data to be clustered, and determining the sales strategy of the sales channel corresponding to each category; Step S5, determining the optimal sales strategy according to the sales price, inventory turnover days and sales target of each category of sales strategy.

Images