Inventory monitoring method, system, computer equipment and medium

By analyzing the historical sales data of e-commerce platforms, identifying seasonal patterns and classifying products, and adopting targeted inventory strategies, the problem of improper inventory management is solved, sufficient supply and cost reduction during peak sales periods are achieved, and sales efficiency is improved and profitability.

CN117829740BActive Publication Date: 2024-11-22深圳泽熙网络科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410248201.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-05
Publication Date
2024-11-22
Estimated Expiration
2044-03-05

AI Technical Summary

Technical Problem

In the field of e-commerce, it is difficult for companies to accurately predict product sales, resulting in improper inventory management, which may lead to insufficient or backlog of inventory, affecting sales and costs.

Method used

By obtaining the historical sales data of e-commerce platforms, identifying seasonal patterns, classifying products into seasonal and non-seasonal products, adopting a predicted inventory strategy based on sales volume forecast values, and making forecast inventory adjustments to seasonal products to avoid insufficient Or backlog, adopt an average inventory strategy for non-seasonal commodities to save costs.

Benefits of technology

Ensure sufficient supply during peak sales periods, improve sales efficiency, reduce waiting time, reduce inventory holding and capital occupation costs, optimize resource allocation, and improve corporate profitability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117829740B_ABST
    Figure CN117829740B_ABST
Patent Text Reader

Abstract

The present invention provides an inventory monitoring method, system, computer equipment and medium, and the monitoring method includes the following steps: S1: obtaining historical sales data of goods on sale on an e-commerce platform; S2: identifying the seasonal pattern of goods based on the historical sales data, and classifying the goods based on the seasonal pattern, wherein the classification results include seasonal goods and non-seasonal goods; S3: monitoring the inventory of seasonal goods and / or non-seasonal goods, and adopting corresponding inventory strategies; wherein, for seasonal goods, calculating the sales volume forecast value of the seasonal goods, and adopting a forecast inventory strategy based on the sales volume forecast value; for non-seasonal goods, adopting an average inventory strategy; and adopting different inventory monitoring strategies for seasonal goods and non-seasonal goods, respectively, to improve the efficiency and flexibility of inventory management.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of e-commerce, and in particular relates to an inventory monitoring method, system, computer equipment and medium. Background Art

[0002] In the field of e-commerce, the management of commodity inventory is crucial to the operation of an enterprise. At present, enterprises usually rely on sales experience to predict the sales of commodities, and then formulate corresponding inventory management and sales strategies. However, this method has some problems. For example, when there are too many types of commodities, it is difficult to cover all types of commodities by relying on manual experience, and the prediction results are not accurate enough. In addition, it is difficult to accurately grasp when to increase or decrease inventory.

[0003] On the one hand, if you increase inventory in advance, you will be able to cope with the peak sales season more calmly, but the increased inventory will take up more storage capacity, and the extra storage capacity may affect the stocking of other products. On the other hand, if you delay increasing inventory, it may be difficult to cope with the upcoming peak sales season. Summary of the invention

[0004] In view of this, the present application aims to solve at least one of the problems in the above technology to a certain extent. To this end, the present application provides an inventory monitoring method, system, computer device and medium to ensure sufficient supply during peak sales period while reducing inventory costs.

[0005] In order to achieve the above-mentioned purpose, one aspect of the present application provides an inventory monitoring method, comprising the following steps: S1: obtaining historical sales data of goods on sale on an e-commerce platform; S2: identifying the seasonal pattern of goods based on the historical sales data, and classifying the goods based on the seasonal pattern, wherein the classification results include seasonal goods and non-seasonal goods; S3: monitoring inventory of seasonal goods and / or non-seasonal goods, and adopting corresponding inventory strategies; wherein, for seasonal goods, calculating the sales volume forecast value of the seasonal goods, and adopting a predictive inventory strategy based on the sales volume forecast value; for non-seasonal goods, adopting an average inventory strategy; adopting a predictive inventory strategy for seasonal goods can replenish stocks in time according to sales forecasts, avoid insufficient inventory or backlogs, and thus improve sales efficiency, while adopting an average inventory strategy for non-seasonal goods can save costs and avoid inventory backlogs.

[0006] Preferably, S1 also includes: if the commodity on sale is a first-time listing, there is no historical sales data, and similar commodities are searched based on preset specification parameters of the commodity on sale, wherein the similar commodities are commodities on the same e-commerce platform, and the similar commodities include commodities on sale and historical commodities; the historical sales data of the similar commodities are obtained, and S2 to S3 are executed; this makes up for the gap in the acquisition of data on commodities that are listed for the first time, helps predict the sales of new commodities and formulate appropriate inventory management strategies.

[0007] Preferably, the preset specification parameters at least include the size, weight, color and purpose of the product.

[0008] Preferably, if the number of similar products is greater than two, features are extracted from the preset specifications of the similar products, and each similar product is represented as a feature vector based on the features, and the feature vector v=[x1, x2, ..., x n ], where x1, x2, ..., x n are the values ​​of the 1st to nth features in the feature vector respectively; based on the feature vector, the similarity between the products is calculated, the similar products closest to the products on sale are determined, and S2 to S3 are executed; based on the calculated product similarity, the similar products closest to the new products can be determined, providing a more accurate historical sales data reference for the new products, helping to predict the sales of the new products and formulate appropriate inventory management strategies; wherein, the Euclidean distance formula is used to calculate the distance between any two products, and the calculation formula of the Euclidean distance is as follows:

[0009] ;

[0010] Where v1 is the first eigenvector, v2 is the second eigenvector, n is the dimension of the eigenvector, and v 1i is the value of v1 in the i-th dimension, v 2i is the value of v2 in the i-th dimension.

[0011] Preferably, S2 also includes: extracting seasonal characteristics from historical sales data, wherein the seasonal characteristics include a seasonal index SI of commodity sales quantity and sales time t The calculation formula is as follows:

[0012] ;

[0013] Among them, X t is the sales volume in time t, tis the average of all historical sales within time t; taking seasonal characteristics as input, a seasonal demand model is established; using the seasonal demand model to predict and identify new sales data; calculating the inventory change based on the current inventory, and adjusting the new inventory based on the inventory change.

[0014] Preferably, the sales cycle of seasonal commodities is divided into at least a seasonal cycle and a non-seasonal cycle based on the sales volume change trend in the seasonal demand model, wherein the sales volume in the seasonal cycle is greater than the sales volume in the non-seasonal cycle; n The previous cycle Q n-1 In the seasonal cycle Q, the inventory change is increased; in the seasonal cycle Q n The next cycle Q n+1 By dividing the sales cycle of seasonal goods into seasonal cycles and non-seasonal cycles, and increasing or decreasing the inventory changes according to the sales trend in the seasonal cycle, the inventory level can be adjusted more accurately and the inventory management strategy can be optimized. In view of the high sales volume in the seasonal cycle, the inventory changes are increased in the previous cycle of the seasonal cycle, which helps to ensure sufficient inventory supply, meet the demand during the peak sales period, and improve sales efficiency and customer satisfaction.

[0015] Preferably, based on the sales volume forecast value, the seasonal cycle Q n Including the day with the highest forecast value; before the day with the highest forecast value, increasing the inventory change; the day with the highest forecast value usually corresponds to the peak sales period, increasing the inventory change can effectively avoid the problem of insufficient inventory caused by the peak sales period and avoid the loss of sales opportunities.

[0016] On the other hand, the present application provides an inventory monitoring system, including: a data acquisition module, used to obtain historical sales data of goods on sale on the e-commerce platform; a product classification module, used to identify the seasonal patterns of goods and classify the goods based on the seasonal patterns; an inventory monitoring module, used to adopt corresponding inventory monitoring strategies for seasonal goods and non-seasonal goods.

[0017] On the other hand, the present application provides a computer device, comprising at least a processor and a memory, wherein the processor and the memory are communicatively connected, wherein at least one computer program is stored in the memory, and when the computer program is executed by the processor, the above-mentioned monitoring method is implemented.

[0018] On the other hand, the present application provides a computer-readable storage medium storing at least one computer instruction, and when the computer instruction is executed, the above-mentioned monitoring method is implemented.

[0019] The technical solution provided by this application can achieve the following beneficial effects:

[0020] 1. Adjust inventory levels based on sales forecasts, especially increase inventory changes before peak sales periods to ensure adequate product supply, so as to meet potential high sales demand, avoid loss of sales opportunities due to insufficient inventory, and improve customer satisfaction and loyalty.

[0021] 2. By increasing the inventory change, it is possible to maintain a good inventory level during the sales peak period, so as to better meet customer needs and improve sales efficiency, which helps to deliver products in time, reduce waiting time, increase sales, and optimize the company's sales performance.

[0022] 3. Accurate inventory management and capital operation can avoid inventory backlogs and capital waste, reduce inventory holding costs and capital occupation costs. Reasonable inventory adjustments and capital operations can also help companies optimize resource allocation and improve capital utilization efficiency, thereby enhancing the profitability of the company. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0024] Figure 1 is a flow chart of a monitoring method provided by an embodiment of the present invention;

[0025] Figure 2 is a flow chart of a method for acquiring historical sales data provided by an embodiment of the present invention;

[0026] Figure 3 is a flow chart of a method for distinguishing seasonal goods from non-seasonal goods provided by an embodiment of the present invention;

[0027] Figure 4 is a flow chart of an inventory adjustment method provided by an embodiment of the present invention;

[0028] Figure 5 is a flow chart of another inventory adjustment method provided by an embodiment of the present invention;

[0029] Figure 6 is a flow chart of an average inventory adjustment method provided by an embodiment of the present invention;

[0030] Figure 7 is a flow chart of a method for screening similar products provided by an embodiment of the present invention;

[0031] Figure 8is a schematic diagram of the structure of a monitoring system provided by an embodiment of the present invention;

[0032] Fig. 9 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention.

[0033] Among them: 101-data acquisition module, 102-commodity classification module, 103-inventory monitoring module, 201-processor, 202-memory. DETAILED DESCRIPTION

[0034] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0035] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0036] An embodiment of an inventory monitoring method is provided herein, such as Figure 1 As shown, the monitoring method comprises the following steps:

[0037] S1: Obtain the historical sales data of products on sale on the e-commerce platform.

[0038] In one example, the e-commerce platform includes a database, which stores historical sales data of the e-commerce platform, including order information, product information, customer information, transaction time, etc. Figure 2As shown, S101: log in to the database management system of the e-commerce platform, confirm the information including historical sales data in the database, such as order table, transaction table, etc., use database management tools to connect to the database, it can be imagined that database management tools include MySQLWorkbench, pgAdmin, etc., S102: use SQL query language to prepare query statements, select corresponding tables and fields to obtain the required historical sales data, the query statement includes order number, product number, quantity, price and order date, for example, BETWEEN '2023-01-01' AND '2023-12-31', to query all order data with order date in 2023, and then S103: execute SQL query statements to obtain historical sales data from the database.

[0039] S2: Identify the seasonal pattern of the goods based on the historical sales data, and classify the goods based on the seasonal pattern, wherein the classification results include seasonal goods and non-seasonal goods.

[0040] In one example, if Figure 3 As shown, S201: arrange the historical sales data in chronological order. When calculating the seasonal pattern, S202: use the seasonal index to more accurately represent the sales volume changes of each month relative to the overall average. Assume that the inventory of commodity A is monitored, and only the sales data of commodity A in 2023 is compared to extract the sales time and sales quantity in the sales data. The sales data in 2023 is as shown in Table 1 below:

[0041] ,

[0042] Table 1

[0043] Seasonality Index SI t The calculation formula is as follows:

[0044] ;

[0045] Among them, X t is the sales volume in time t, t is the average of all historical sales in time t. According to the above table, we can get: X1=50, X2=60, ... X 12 =60, 12 =75, we can calculate SI1=0.667, SI2=0.8, ...SI 12 =0.8, as shown in Table 2 below:

[0046] ,

[0047] Table 2

[0048] The seasonal index represents the change in sales volume in each month relative to the overall average. The seasonal index is positively correlated with the sales share, and commodities can be classified based on the seasonal index. It can be understood that the sales time in Table 1 is in months. If other forms of historical sales data are in weeks, the seasonal cycle is also in weeks.

[0049] Correspondingly, the seasonal pattern is based on the seasonal index. The seasonal pattern of commodity A can reflect that the sales peak of commodity A is in summer, among which the month with the highest sales volume is July, and the sales volumes in other seasons are relatively low, and the month with the lowest sales volume is January. It can be imagined that the difference in the seasonal index of each month can express the relative changes in sales volumes between different months. The larger the difference, the greater the fluctuation of the sales volume in the corresponding month relative to the annual average, thereby helping to identify seasonal patterns.

[0050] S203: If the difference of the seasonal index is greater than or equal to the first threshold, the product is defined as a seasonal product. Based on the data in Table 2 above, the difference of the seasonal index between January and July is 1.2. Assuming the first threshold is 1, product A is defined as a seasonal product.

[0051] In other examples, there are different ways to define a product as a seasonal product. For example, the difference between the seasonal index of at least two months and the seasonal index of any month is greater than or about the first threshold. Based on the data in Table 2 above, the difference between the seasonal index of January and July is 1.2, which is greater than the first threshold; the difference between the seasonal index of January and August is 1.066, which is greater than the first threshold. Then the difference between the seasonal index of July and August and the seasonal index of January is greater than the first threshold, so product A is defined as a seasonal product. The above example takes the month with the lowest sales volume as any month. In another example, the month with the highest sales volume is taken as any month. The difference between the seasonal index of July and January is 1.2, which is greater than the first threshold; the difference between the seasonal index of July and February is 1.067, which is greater than the first threshold, so product A is defined as a seasonal product.

[0052] Correspondingly, for some months, the smaller the difference in seasonal index, the more stable the sales volume and the smaller the fluctuation. If the difference in seasonal index between any two months is less than the first threshold, the product is defined as a non-seasonal product.

[0053] S3: Monitor inventory of seasonal and / or non-seasonal goods and adopt corresponding inventory strategies;

[0054] Among them, for seasonal products, the sales volume forecast value of seasonal products is calculated, and a predictive inventory strategy is adopted based on the sales volume forecast value; for non-seasonal products, an average inventory strategy is adopted.

[0055] In an example, the seasonal demand model is based on the SARIMA model. Different inventory strategies are adopted for seasonal goods and non-seasonal goods. For seasonal goods, an initial parameter of a SARIMA model is selected, and the SARIMA model is applied to historical sales data for fitting. The fitting process estimates the parameter estimates of the model. The parameter estimates include autoregressive coefficients, seasonal autoregressive coefficients, moving average coefficients, and seasonal moving average coefficients. The fitted SARIMA model is used to complete the sales volume forecast.

[0056] Specifically, a SARIMA(1,1,1)(1,1,1)12 model is selected and the following parameter estimates are obtained: =0.5, = -0.3, =0.2, = -0.1, the time index of the last observation is T, the mean μ of the model is 100, and the residual of the last observation is =5, the residual of the second-to-last observation =3, use the recursive formula to get the sales forecast value. For i=1,2,...12, the sales forecast value for the next 12 months is calculated as follows:

[0057] ;

[0058] in, is the last observation in the historical sales data, It is the 12th last observation.

[0059] It is understandable that other examples may also adopt other models to complete the sales volume forecast value.

[0060] Furthermore, a forecast inventory strategy is adopted based on the sales forecast value, increasing inventory in months when the sales forecast value increases, and reducing inventory in months when the sales forecast value decreases. It can be understood that if the time unit is quarter, inventory will be increased in quarters when the sales forecast value increases, and inventory will be reduced in quarters when the sales forecast value decreases.

[0061] Furthermore, for seasonal commodities, seasonal cycles and non-seasonal cycles are introduced. Seasonal cycles and non-seasonal cycles constitute the complete sales cycle of seasonal commodities. This example takes part of the sales data of commodity A in Table 1 as an example. After calculating the seasonal index of commodity A in each month, the seasonal cycle and non-seasonal cycle of commodity A are distinguished based on the seasonal index.

[0062] A second threshold is set. If the seasonal index of a certain month is greater than or equal to the second threshold, the month is defined as part of the seasonal cycle. If the seasonal index of a certain month is less than the second threshold, the month is defined as part of the non-seasonal cycle.

[0063] Based on the data in Table 2 above, assuming that the second threshold is 1.5, the seasonal indexes of June, July and August are greater than the second threshold, the seasonal cycle of commodity A is June-August, and the non-seasonal cycle of commodity A is January-May and September-December, among which the sales quantity in any month in the seasonal cycle is greater than the sales quantity in any month in the non-seasonal cycle.

[0064] like Figure 4 As shown, in the seasonal cycle Q n The previous cycle Q n-1 Specifically, commodity A enters the seasonal cycle from June, n=6, then the user is reminded to adjust the inventory in May, S301: the inventory change is calculated according to the sales forecast values ​​for May and June and the current inventory, inventory change = May sales forecast + June sales forecast - current inventory, S302: the user increases the corresponding inventory according to the received inventory change prompt.

[0065] like Figure 5 As shown, in the seasonal cycle Q n The next cycle Q n+1 Specifically, commodity A enters the non-seasonal cycle since September, n=8, then the user is provided with the opportunity to adjust the inventory when entering September, S311: the inventory change is calculated based on the sales forecast values ​​for August and September and the current inventory, inventory change = current inventory - August sales forecast value - September sales forecast value, S312: the user reduces the corresponding inventory based on the received inventory change prompt.

[0066] In one example, taking days as the time unit, we get Q n The daily sales volume forecast value within the period, when n=6, is provided to the user again before the highest forecast value day in June to increase the corresponding inventory.

[0067] In one example, for non-seasonal products, an average inventory strategy is adopted, such as Figure 6As shown, S321: calculate the monthly average sales volume based on historical sales data, and use the monthly average sales volume as the sales volume forecast value of non-seasonal products, and S322: adjust the inventory of non-seasonal products to the monthly average sales volume. It can be understood that if there is an obvious growth or decrease trend in sales volume during the sales volume monitoring process, the inventory is increased or decreased accordingly. In some examples, obvious growth is defined as monthly sales volume greater than or equal to the third threshold, and obvious decrease trend is defined as monthly sales volume less than the third threshold.

[0068] It should be noted that the definition of the time period in this embodiment is that the user can independently select quarters or months as the time unit based on actual sales conditions. It is understandable that in some special cases, days can also be selected as the time unit.

[0069] When an e-commerce platform has newer products, when the product is first listed, the e-commerce platform's database does not have the product's historical sales data. In order to accurately manage the inventory of the new product, the new product's historical sales data refers to the historical sales data of similar products, such as Figure 7 As shown, S401: perform keyword screening; S402: convert preset specification parameters of the product into feature vectors, and calculate the similarity of the products; S403: take the product with the highest similarity as the same category product of the new product.

[0070] Specifically, first determine the preset specification parameters of the new product. The preset specification parameters are displayed on the e-commerce platform. The preset specification parameters include the size, weight, color, and purpose of the product. It can be imagined that the preset specification parameters come from the supplier's product information and can also be obtained through online searches, product catalogs and other channels. Further, clearly define the standards for size, weight, color, and purpose parameters. For example, size may include length, width, and height; weight may be identified by a weight range; color may be a specific color name or color code; and purpose may be the main function of the product or the target user group.

[0071] In one example, Euclidean distance is used as a similarity metric to measure the similarity between products. Specifically, the similarity between products in terms of size, weight, color, and usage parameters is evaluated. Euclidean distance is used to screen products on the e-commerce platform, and products that are most similar to new products are selected as similar products. A third threshold is set to control the level of similarity.

[0072] In order to shorten the calculation process and obtain the calculation results faster, a two-step screening method is adopted. The first step of the screening method is the keyword screening method. The keyword is generally the name of the product. The search engine matches the corresponding products in the database according to the input keyword. These matched products have a larger range. It is understandable that in order to increase the exposure of the product, the product name usually contains more field information, that is, the name of a product includes the product and other products. If a product name includes more than two products, the two or more products may be of the same type or may be supporting products. In the next step, products other than the same type are screened out.

[0073] The second step is to calculate the similarity. First, the preset specification parameters of the product are converted into feature vectors, where the preset specification parameters of the product include size, weight, color and purpose. The size of the product is converted into a numerical feature. In an example, the size of the product (cm) is used as the numerical value of the feature; the material of the product is converted into a numerical feature, and different numerical codes are assigned to different materials. For example, metal is encoded as 0, wood is encoded as 1, and plastic is encoded as 2; the color of the product is converted into a numerical feature, and different numerical codes are assigned to different colors. For example, blue is encoded as 0, red is encoded as 1, and green is encoded as 2. The purpose of the product is converted into a numerical feature, and different numerical codes are assigned to different purposes. For example, a desk is encoded as 0, a dining table is encoded as 1, and a chair is encoded as 2.

[0074] For product A and product B, the preset specification parameters of product A are: size: 120cm x 80cm x 75cm, material: natural wood, color: red, purpose: dining table; the preset specification parameters of product B are: 100cm x 60cm x 70cm, material: metal, color: blue, purpose: desk, then the feature vector of product A is: [120,80,75,1, 1,1], and the feature vector of product B is: [100,60,70,0, 0,0], where the values ​​in the feature vector represent the length, width, and height of the size, the material code (wood is 1, metal is 0), the color code (red is 1, blue is 0), and the function and purpose code (dining is 1, office or study is 0).

[0075] The calculation formula of Euclidean distance is as follows:

[0076] ;

[0077] Where v1 is the first eigenvector, v2 is the second eigenvector, n is the dimension of the eigenvector, and v 1i is the value of v1 in the i-th dimension, v 2i is the value of v2 in the i-th dimension.

[0078] Substitute the feature vectors of product A and product B: d(v 1, v2)=28.77. According to the given feature vector, the Euclidean distance between product A and product B is 28.77. The smaller the distance, the more similar the two products are. In some examples, if the number of similar products is greater than two, the product B with the smallest Euclidean distance is selected as the closest similar product of the new product A, and the historical sales data of product B is used as the historical sales data of product A. Repeat S2 to S3.

[0079] The preset specification parameters of the product include several parameters. In one example, the user can assign different weights to each parameter according to actual needs. For example, the weight of size is set to 0.2, the weight of weight is set to 0.2, the weight of color is set to 0.2, and the weight of purpose is set to 0.4.

[0080] In another example, the user can select some parameters to be converted into feature vectors according to actual needs. For example, the user only selects size, weight, and purpose to be converted into feature vectors. Referring to the preset specification parameters of product A and product B, the feature vector of product A is: [120, 80, 75, 1, 1], and the feature vector of product B is: [100, 60, 70, 0, 0].

[0081] Based on the above example, as shown in the figure, another example of the present application further provides an inventory monitoring system, wherein the monitoring system includes:

[0082] The data acquisition module 101 is used to obtain the historical sales data of the products on sale on the e-commerce platform; the product classification module 102 is used to identify the seasonal pattern of the products and classify the products based on the seasonal pattern; the inventory monitoring module 103 is used to adopt corresponding inventory monitoring strategies for seasonal products and non-seasonal products.

[0083] Specifically, the data acquisition module 101 communicates with the e-commerce platform interface to obtain historical sales data, including sales quantity, sales time, etc., and stores it in the system database. The product classification module 102 establishes a seasonal demand model by analyzing historical sales data, market trends and seasonal factors, and classifies products according to the model to distinguish between seasonal products and non-seasonal products. The inventory monitoring module 103 is for seasonal products. This module predicts sales volume according to the seasonal demand model and increases inventory during seasonal peaks to meet demand. For non-seasonal products, an average inventory strategy is adopted to calculate the monthly average sales volume based on historical sales data, and adjust the inventory volume to the monthly average sales volume.

[0084] Based on the above example, as shown in the figure, another example of the present application also provides a computer device, wherein the computer device includes a processor 201 and a memory 202, the processor 201 and the memory 202 are communicatively connected, and at least one computer program is stored in the memory 202. When the computer program is executed by the processor 201, an inventory monitoring method for seasonal goods and non-seasonal goods is implemented.

[0085] Based on the above examples, another example of the present application further provides a computer-readable storage medium storing at least one computer instruction. When the computer instruction is executed, an inventory monitoring method for seasonal goods and non-seasonal goods is implemented.

[0086] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for inventory monitoring, characterized in that: The following steps are involved: S1: Obtain the historical sales data of the products on sale on the e-commerce platform; S2: Identifying the seasonal pattern of the commodity based on the historical sales data, and classifying the commodities based on the seasonal pattern, wherein the classification results include seasonal commodities and non-seasonal commodities; including: arranging the historical sales data in chronological order; calculating the seasonal index of each month using the following formula: , Among them, X t is the sales volume in time t, is the average of all historical sales in time t; Defining commodities whose seasonal index difference is greater than or equal to a first threshold as seasonal commodities, and vice versa as non-seasonal commodities, including: defining commodities whose seasonal index difference between at least two months and the seasonal index of any month is greater than or equal to the first threshold as seasonal commodities, and defining commodities whose seasonal index difference between any two months is less than the first threshold as non-seasonal commodities; S3: Monitor inventory of seasonal and / or non-seasonal goods and adopt corresponding inventory strategies; Among them, for seasonal commodities, the sales volume forecast value of the seasonal commodities is calculated, and a forecast inventory strategy is adopted based on the sales volume forecast value; the calculation of the sales volume forecast value of the seasonal commodities includes: determining the initial parameters of the SARIMA model; applying the SARIMA model with the determined initial parameters to the historical sales data for fitting, and obtaining the parameter estimation value of the SARIMA model; based on the fitted SARIMA model, using a recursive formula to calculate the sales volume forecast value; For non-seasonal products, adopt an average inventory strategy; S1 also includes: If the product on sale is a first-time product and there is no historical sales data for the product on sale, keyword screening is performed. If the screening result includes at least two products, the preset specification parameters of the products are converted into feature vectors, and the similarity of the products is calculated; the product with the highest similarity is used as a similar product of the new product, and the preset specification parameters include at least the size, weight, color and purpose of the product, and S2 to S3 are executed.

2. The inventory monitoring method according to claim 1, characterized in that: S2 also includes: Taking seasonal characteristics as input, establishing a seasonal demand model; the seasonal characteristics include the seasonal index; Use the seasonal demand model to predict and identify new sales data; The inventory change amount is calculated according to the current inventory, and the new inventory is adjusted based on the inventory change amount.

3. The inventory monitoring method according to claim 2, characterized in that: Also includes: Based on the sales volume change trend in the seasonal demand model, the sales cycle of seasonal products is divided into at least a seasonal cycle and a non-seasonal cycle, wherein the sales volume in the seasonal cycle is greater than the sales volume in the non-seasonal cycle within the same unit time; In the seasonal cycle Q n The previous cycle Q n-1 , increasing the inventory change; In the seasonal cycle Q n The next cycle Q n+1 , reducing the inventory change amount.

4. The inventory monitoring method according to claim 3, characterized in that: Also includes: Based on the sales volume forecast value, the seasonal cycle Q n Includes the day with the highest predicted value; Before the day with the highest predicted value, increase the inventory change amount.

5. An inventory monitoring system, characterized in that: include: The data acquisition module is used to obtain the historical sales data of the products on sale on the e-commerce platform; The commodity classification module is used to identify the seasonal pattern of commodities and classify the commodities based on the seasonal pattern; wherein the classification results include seasonal commodities and non-seasonal commodities; including: arranging the historical sales data in chronological order; calculating the seasonal index of each month using the following formula: , Among them, X t is the sales volume in time t, is the average of all historical sales in time t; Defining commodities whose seasonal index difference is greater than or equal to a first threshold as seasonal commodities, and vice versa as non-seasonal commodities, including: defining commodities whose seasonal index difference between at least two months and the seasonal index of any month is greater than or equal to the first threshold as seasonal commodities, and defining commodities whose seasonal index difference between any two months is less than the first threshold as non-seasonal commodities; The inventory monitoring module is used to adopt corresponding inventory monitoring strategies for seasonal goods and non-seasonal goods; wherein, for seasonal goods, the sales volume forecast value of the seasonal goods is calculated, and a forecast inventory strategy is adopted based on the sales volume forecast value; the calculation of the sales volume forecast value of the seasonal goods includes: determining the initial parameters of the SARIMA model; applying the SARIMA model with the determined initial parameters to the historical sales data for fitting to obtain the parameter estimation value of the SARIMA model; based on the fitted SARIMA model, using a recursive formula to calculate the sales volume forecast value; For non-seasonal products, adopt an average inventory strategy; The data acquisition module is also used to perform keyword screening if the product on sale is a first-time product on the shelves and there is no historical sales data for the product on sale. If the screening result includes at least two products, the preset specification parameters of the products are converted into feature vectors to calculate the similarity of the products; the product with the highest similarity is used as a similar product of the new product, the preset specification parameters at least include the size, weight, color and purpose of the product, and the product classification module is executed to the inventory monitoring module.

6. A computer device comprising at least a processor and a memory, characterized in that: The processor is communicatively connected to a memory, wherein at least one computer program is stored in the memory, and when the computer program is executed by the processor, the inventory monitoring method according to any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium storing at least one computer instruction, characterized in that: When the computer instructions are executed, the inventory monitoring method according to any one of claims 1 to 4 is implemented.