CN113592539A - Shop order trend prediction method and device based on artificial intelligence and storage medium - Google Patents
Shop order trend prediction method and device based on artificial intelligence and storage medium Download PDFInfo
- Publication number
- CN113592539A CN113592539A CN202110760703.4A CN202110760703A CN113592539A CN 113592539 A CN113592539 A CN 113592539A CN 202110760703 A CN202110760703 A CN 202110760703A CN 113592539 A CN113592539 A CN 113592539A
- Authority
- CN
- China
- Prior art keywords
- order
- shop
- data
- artificial intelligence
- business
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 30
- 239000011159 matrix material Substances 0.000 claims description 28
- 238000003064 k means clustering Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 238000003062 neural network model Methods 0.000 claims description 8
- 238000007477 logistic regression Methods 0.000 claims description 4
- 230000008901 benefit Effects 0.000 abstract description 5
- 238000004458 analytical method Methods 0.000 description 16
- 238000010586 diagram Methods 0.000 description 11
- 238000010801 machine learning Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001960 triggered effect Effects 0.000 description 2
- 241000531116 Blitum bonus-henricus Species 0.000 description 1
- 235000008645 Chenopodium bonus henricus Nutrition 0.000 description 1
- 239000004278 EU approved seasoning Substances 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 235000013361 beverage Nutrition 0.000 description 1
- 239000012459 cleaning agent Substances 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013479 data entry Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000011194 food seasoning agent Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 235000008446 instant noodles Nutrition 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 235000011888 snacks Nutrition 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000003442 weekly effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Economics (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a shop order trend prediction method, a shop order trend prediction device and a storage medium based on artificial intelligence, wherein the shop order trend prediction method comprises the following steps: dividing the shop into a plurality of associated business circle sets according to preset rules; generating observation characteristic data according to all historical orders of all shops in the associated business district set in a set observation period; inputting the observation characteristic data into a business district prediction model so that the business district prediction model outputs business district order prediction data and a corresponding business district confidence coefficient; and judging whether the business circle confidence coefficient is greater than a preset business circle confidence coefficient threshold value, and if the business circle confidence coefficient is greater than the preset business circle confidence coefficient threshold value, obtaining shop order prediction data of one shop in the associated business circle set at least according to the business circle order prediction data. The method, the device and the storage medium for predicting the order trend of the stores based on the artificial intelligence have the advantages that the method, the device and the storage medium for predicting the order trend of the stores based on the artificial intelligence are provided, and the influence on the stores caused by the store environment is taken into consideration according to the historical data of the stores.
Description
Technical Field
The application relates to the field of e-commerce data management, in particular to a shop order trend prediction method and device based on artificial intelligence and a storage medium.
Background
The existing small and medium-sized convenience stores usually carry out off-line commodity purchase at suppliers through own channels, and due to the characteristics of the small and medium-sized convenience stores, the small and medium-sized convenience stores cannot purchase goods in large batches, so that effective bargaining can not be carried out with the suppliers, and meanwhile, due to the requirement of wholesale of the suppliers on the purchase quantity, the small and medium-sized convenience stores need to guarantee a certain scale for each purchase, so that the inventory problem is caused.
From the perspective of suppliers, the scattered purchasing mode of small and medium-sized convenience stores leads to the increase of the warehousing cost of the suppliers, so that the supply price is kept high.
In the related art, as disclosed in chinese patent publication No. CN112801759A, a buyer, a seller and a carrier are integrated together through an e-commerce platform, so that orders of multiple buyers can be 'listed' in a single logistics order, thereby implementing centralized purchasing and centralized distribution, and reducing the purchasing cost of small and medium-sized convenience stores.
However, due to the occurrence of the 'order-sharing' mode, orders of stores are more random, in order to meet the discrete requirements of the stores, the system needs to analyze orders which are possibly generated in the future of the stores to carry out configuration of goods sources and warehouses, and the analysis scheme of the existing system is more to analyze and predict from historical data of the stores, so that external factors, particularly the commercial influence of nearby friends on the stores, are ignored.
Disclosure of Invention
In order to solve the defects of the prior art, the application provides a store order trend prediction method based on artificial intelligence, which comprises the following steps: dividing the shop into a plurality of associated business circle sets according to preset rules; generating observation characteristic data according to all historical orders of all shops in the associated business district set in a set observation period; inputting the observation characteristic data into a business district prediction model so that the business district prediction model outputs business district order prediction data and corresponding business district confidence; and judging whether the business circle confidence coefficient is larger than a preset business circle confidence coefficient threshold value or not, and if the business circle confidence coefficient is larger than the preset business circle confidence coefficient threshold value, obtaining the shop order prediction data of one shop in the associated business circle set at least according to the business circle order prediction data.
Further, the step of dividing the plurality of stores into the associated business district sets according to preset rules comprises the following steps: establishing a two-dimensional coordinate system according to the geographic position; acquiring coordinate values of the shop in the two-dimensional coordinate system; performing K-Means clustering operation on the coordinate values of the stores in the two-dimensional coordinate system; and dividing the associated business circle set according to the result of the K-Means clustering operation.
Further, the step of dividing the plurality of stores into the associated business district sets according to preset rules further comprises the following steps: acquiring historical order data of the shop; calculating the average order value of the shop according to the historical order data of the shop, wherein the average order value is the average value of the order values of all purchase orders of the shop in the observation period; establishing a three-dimensional coordinate system by taking the average order value as a third dimension; acquiring coordinate values of the shop in the three-dimensional coordinate system; performing K-Means clustering operation on the coordinate values of the stores in the three-dimensional coordinate system; and dividing the associated business circle set according to the K-Means clustering operation result.
Further, the store order trend prediction method based on artificial intelligence further comprises the following steps: and taking all historical orders of all shops in the associated business district set within a set observation period as training set data.
Further, the quotient circle prediction model is a BP neural network model.
Further, the quotient circle prediction model is a logistic regression analysis model.
Further, the observation feature data is a matrix expressing stores, commodities and their corresponding relations.
As another aspect of the present application, the present application further provides an artificial intelligence-based store order trend prediction apparatus, including: a memory for storing a computer program; and the processor is used for realizing the store order trend prediction method based on artificial intelligence when executing the computer program.
As another aspect of the present application, the present application further provides a computer client storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the artificial intelligence based store order trend prediction method as described above.
The application has the advantages that: the artificial intelligence based store order trend prediction method, the artificial intelligence based store order trend prediction device and the storage medium are provided, wherein the artificial intelligence based store order trend prediction method, the artificial intelligence based store order trend prediction device and the storage medium not only take account of influence of a business environment where a store is located on the store according to historical data of the store.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, serve to provide a further understanding of the application and to enable other features, objects, and advantages of the application to be more apparent. The drawings and their description illustrate the embodiments of the invention and do not limit it. In the drawings:
FIG. 1 is a schematic diagram of the steps of an automated order generation method according to one embodiment of the present application;
FIG. 2 is a schematic diagram of an order prediction model according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a sales prediction model according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a price prediction model according to one embodiment of the present application;
FIG. 5 is a schematic diagram of steps of a store order trend prediction method according to one embodiment of the present application;
FIG. 6 is a schematic diagram of a rectangle of observed feature data according to one embodiment of the present application;
FIG. 7 is a schematic diagram of a three-dimensional coordinate set clustered by K-Means clustering;
FIG. 8 is a schematic diagram of a business turn prediction model according to an embodiment of the present application;
FIG. 9 is a schematic illustration of a method of store merchandise analysis according to one embodiment of the present application;
FIG. 10 is a schematic diagram of an order prediction model according to an embodiment of the present application;
FIG. 11 is a matrix diagram of input data for an order prediction model according to one embodiment of the present application;
FIG. 12 is a block diagram of an apparatus implementing one embodiment of the method of the present application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present application will be described in detail below with reference to the embodiments with reference to the attached drawings.
From an overall point of view, the present application constructs a system that employs the following method, which mainly achieves three aspects of functionality:
1. a suggested purchase order is automatically provided for the shop according to the corresponding data of the shop, so that the management cost of a shop operator is reduced, and meanwhile, the system is convenient to efficiently configure supplier, warehouse and logistics resources.
2. Analysis is performed from a range of similar stores to analyze the trends of the orders as a whole and to analyze possible orders from the stores based thereon.
3. And analyzing the operation mode of the shop so as to assist in analyzing the future order trend of the shop of the type in the similar historical data of the shop.
Specifically, referring to fig. 1 to 4, as a first aspect of the present application, an automated order generation method of the present application includes: collecting inventory data and sales data of stores; generating shop state characteristic data according to the stock data and the sales data of the shop; inputting the store state characteristic data into an order prediction model so that the order prediction model outputs a suggested order data and a corresponding suggested confidence coefficient; and judging whether the suggested confidence coefficient is larger than a preset suggested confidence coefficient threshold value, and if the suggested confidence coefficient is larger than the suggested confidence coefficient threshold value, generating a purchase order at least according to the suggested order data.
The specific scheme is that the inventory data and the sales data of the stores are collected on the same day of the stores.
In order to provide the suggestion of purchase orders for the shop in time, the system preferably triggers the execution of the automated order generation method of the present application daily, i.e. the system runs the program daily according to user settings or system settings.
The trigger time may be set at 22 points in view of the business habits of convenience stores, or, as another preferable scheme, a trigger condition may be set, and when the trigger condition is satisfied, a program implementing the automated order generation method is automatically executed. For example, after the store uploads inventory data and sales data after daily inventory, a program is automatically triggered; for another example, when a store user uses a user terminal to send a request for an automated order to a server (the user may click a recommendation button in an APP interface of the user terminal), a program is triggered.
The stock data and sales data may be managed by a program of a system platform, or may be connected to other goods statistics programs owned by the store through a data port developed by the system.
Inventory data and sales data as referred to herein refer to the SKU code and corresponding quantity data, respectively, of the items in inventory or sold that day.
According to a general analysis scheme, historical inventory data and sales data are generally adopted to respectively train a machine learning module, then the forecast of the inventory data and the sales data is respectively carried out, then the condition of possibly picking up goods is judged according to the forecast result, and then corresponding order suggestions and the like are generated according to the condition of shortage of goods.
However, due to uncertainty of sales data, unless all goods are included in the input data, the input data is huge and invalid data is large, and finally the trained machine learning model cannot converge.
In addition, although stock data is relatively stable with respect to sales data, the machine learning model cannot be converged due to a problem that the number of categories of products is large like the stock data.
For the above reasons, there is a great technical obstacle to directly adopting inventory data and sales data for machine learning model training and prediction.
Based on the above, the technical scheme of the application adopts a new technical concept, and particularly, as one scheme, after the stock data and the sales data are collected; and acquiring SKU data and sales quantity of the commodities with sales quantity positioned in the first five digits and the existing inventory quantity corresponding to the first five commodities from the sales data (the current day), and forming a matrix with five rows and three columns, wherein the matrix is used as shop state characteristic data.
As an extension, the number of rows of the matrix may be set according to the size of the store scale, and in a store with a large transaction amount, the number of categories of products sold per day is large, and the number of rows of the matrix may be tens of rows or tens of rows.
As a preferable scheme, the total number of the commodities in the store inventory is S, and a head value M is calculated to be S × K, where K is a head percentage, which is determined according to a value interval in which S is located.
When S is more than 0 and is more than or equal to 300 percent, the value range of the head percentage is 7 percent; when S is more than 300 and is more than or equal to 1000, the value range of the head percentage is 15 percent; when S is more than 0 and is more than or equal to 300 percent, the value range of the head percentage is 20 percent. The number of rectangle lines N is equal to the rounding value of the header value M.
The training order prediction model can adopt two schemes, and store state characteristic data introduced above is input. The difference is the source of the output data in the training set.
The first mode is as follows: and processing historical data of the shop, wherein the actual historical purchase order is used as output data, specifically, the output data is a matrix formed by SKU data of commodities in the purchase order and the purchase quantity, the matrix is determined to have two columns, specifically, how many rows are generated according to the actual situation of the purchase order, and the shop state characteristic data of the day before the actual historical purchase order is generated is used as input data, so that a group of training data is formed.
By adopting the scheme, the suggested order data output by the order prediction model is output as a matrix consisting of SKU data and the purchase quantity. The method has the advantages that a purchase order mode can be directly generated, but due to the uncertainty of an output matrix, an order prediction model is equivalent to an empirical model, the final output is greatly influenced by the sequence of input training sets and the setting of model parameters during training, the accuracy fluctuation is large, a high confidence threshold value needs to be set to ensure that the output has reference significance, and the problem of program circulation is caused during running. In this way, the order prediction model may be a CNN neural network model as a preferred scheme.
The second mode is as follows: the order prediction model is constructed into a prediction machine learning model, namely, the shop state characteristic data before the current day is used as input data, the shop state characteristic data on the current day is used as output data, and model training is carried out on the order prediction model, so that the output data and the output data are regular data and a determined matrix, and training convergence is easy. In this way, as a preferred scheme, the order prediction model may adopt a BP neural network model.
Preferably, the order prediction model is constructed in a second way, in which the output is not direct purchase order data, but store status characteristic data (actually a matrix) of the next day as described above, and in the matrix, SKU data, sales numbers and stock numbers of the front-ranked items that may appear on the second day are predicted.
The system passes through the sales quantity and inventory quantity data of all the commodities in the prediction matrix to judge whether the commodities meet the preset relative relationship. Specifically, the correlation is sales number > stock number × balance coefficient; wherein, the value range of the equilibrium coefficient is 0.27 to 0.7. Preferably, the balance coefficient takes a value of 0.5, and when the correlation relation judges that the sales quantity is more than 50% of the stock quantity, the commodity is selected to the purchase order.
Preferably, an order prediction model is trained for each store, namely one order prediction model corresponds to each store. The data for each store is input into a corresponding order prediction model.
As a preferred scheme, the automated order generation method based on artificial intelligence further comprises: and if the suggested confidence is less than or equal to the suggested confidence threshold, returning to the step of collecting inventory data and sales data, namely processing by using the order prediction model again and outputting new output data and confidence.
As a further scheme, if the recommendation confidence level cannot meet the recommendation confidence level threshold all the time, for example, the recommendation confidence level is not met after exceeding the preset number of times (of course, the preset number of times may be 0, that is, the next step is directly performed, and the order prediction model is not returned for data processing), the store state characteristic data of the current day is used as the prediction store state characteristic data, that is, the purchase order is directly generated by using the data of the current day.
The purchase order can be sent to the shop user for reference after being generated, automatic ordering can be achieved if the shop user determines that the purchase order is, and of course, unmanned management can be achieved by the system through ordering directly according to the purchase order under the condition that the shop user authorizes the shop user.
Preferably, the step of generating the purchase order according to the suggested order data in consideration of price fluctuation of the e-commerce platform comprises the following steps: generating a purchase commodity list according to the suggested order data; acquiring the instant purchase price of the commodities in the purchased commodity list; obtaining a future predicted price of the commodity in the purchased commodity list; calculating the total price difference value of the instant purchasing total price and the future prediction total price of all the commodities in the purchasing commodity list; and judging whether the total price difference value is larger than the difference threshold value, and if the total price difference value is smaller than the difference threshold value, generating the total price of the purchase order at the instant purchase price. If the total price difference value is larger than the difference threshold value, prompting the store user to give an order in the day and the day, and suggesting the commodity category possibly with out-of-stock in the order data (screened by the difference value of the stock quantity and the sales quantity in the prediction matrix), and selecting the order in the day or automatically giving the order in the day by the user.
Alternatively, if the total price difference is greater than the difference threshold, inputting sales data to the sales prediction model such that the sales prediction model outputs a future predicted sales; calculating to obtain future predicted inventory according to the current inventory data and the future predicted sales volume; if the future forecasted inventory is less than or equal to 0, the current instant purchase price generates a total price for the purchase order, and if the future forecasted inventory is greater than 0, the total price for the purchase order is generated at the future forecasted price. When the total price of the purchase order is generated at the future predicted price, the system places an order for the store user the next day or prompts the store user to place an order the next day.
The advantage of using the sales forecasting model alone is that the sales forecasting of a single commodity is influenced by other variables compared with the order forecasting model with multiple characteristic dimensions.
While the machine learning model for a single commodity SKU trained with historical sales data for individual commodities, data entry and entry are more regular, which converges easily. Preferably, the sales prediction model is a BP neural network model, the input data of the model is historical sales data of a certain commodity before the current day, and the output data of the model is predicted sales (i.e. future predicted sales) of the commodity on the next day.
This is done primarily to compare the cost variation of orders placed on the current day with the cost variation of orders placed on the next day, since in most cases the generation of purchase orders is based on future (next day) forecasts, avoiding incurring immediate losses for future demand.
As a further preferred solution, future price prediction can be performed by inputting the instant purchase price into a price prediction model; obtaining a predicted price and a price confidence coefficient output by a price prediction model; if the price confidence coefficient is larger than the price confidence coefficient threshold value, outputting the predicted price as the future predicted price; and if the price confidence is less than or equal to the price confidence threshold, outputting the average price in the specified period as the future predicted price.
Similar to sales prediction models, machine learning models for individual commodities are easier to train. The price prediction model is also a BP neural network model, the input data of the model is historical price data of a certain commodity before the current day, and the output data of the model is the predicted price of the commodity on the next day. That is, one price prediction neural network corresponds to one commodity SKU.
The price confidence threshold may actually be set to a smaller value because the price of the item will not vary significantly due to cost considerations and the like. In addition, the commodity price herein refers to the "pin-out" price of the e-commerce platform, rather than the supplier price or the retail price of the store.
In addition, as an alternative, when the price confidence is less than or equal to the price confidence threshold, the specified period may be one month or one week.
Through the scheme, automatic order generation and management can be realized from the perspective of the store, so that a background data base which is convenient for store users to realize automatic ordering, price comparison and other functions through store terminals is provided.
As another aspect of the present application, in order to realize analysis of a trend of a store order within a business district, as shown in fig. 5, a method for predicting a trend of a store order of the present application includes the steps of: dividing the shop into a plurality of associated business circle sets according to preset rules; generating observation characteristic data according to all historical orders of all shops in the associated business district set in a set observation period; inputting the observation characteristic data into a business district prediction model so that the business district prediction model outputs business district order prediction data and a corresponding business district confidence coefficient; and judging whether the business circle confidence coefficient is greater than a preset business circle confidence coefficient threshold value, and if the business circle confidence coefficient is greater than the preset business circle confidence coefficient threshold value, obtaining shop order prediction data of one shop in the associated business circle set at least according to the business circle order prediction data.
The related business circle set is essentially a set of shops and can be expressed as the ID of the shop from the perspective of data representation; of course, the name of the store may be expressed, but the characters themselves are not suitable for data processing, and preferably, the unique store ID code of the store is used, or the unique account ID of the store user may be used, and the roles of the two are the same.
As a specific scheme, the following method is adopted to realize the division of the associated business district sets, and specifically the method comprises the following steps: establishing a two-dimensional coordinate system according to the geographic position; acquiring coordinate values of shops laid in a two-dimensional coordinate system; performing K-Means clustering operation by using coordinate values of the stores in a two-dimensional coordinate system; and dividing the associated business circle set according to the result of the K-Means clustering operation.
By the method, a clustering result of shops on the geographical position can be obtained, but the business attribute of the shop cannot be reflected only from the geographical position. Clustering cannot be achieved from a commercial perspective.
As a preferable scheme based on the above, the business association between the stores can be reflected by adding one business attribute dimension and carrying out three-dimensional clustering operation on the stores.
As a specific scheme, the method comprises the following steps: acquiring historical order data of a shop; calculating the average order value of the shop according to the historical order data of the shop, wherein the average order value is the average value of the order values of all purchase orders of the shop in an observation period; establishing a three-dimensional coordinate system by taking the average order value as a third dimension; acquiring coordinate values of shops laid in a three-dimensional coordinate system; performing K-Means clustering operation by using coordinate values of the stores in a three-dimensional coordinate system; and dividing the associated business district set according to the K-Means clustering operation result.
The scheme is ingenious in that the data of the average value of the purchase orders can reflect the scale of the shop, the purchase intention of the shop on an e-commerce platform and the viscosity of customers at the same time. The data is not available as a dimension for other items of data, and the dimension can better divide the associated business circle set according to the practical application effect.
Preferably, the observation period is quarterly or annual. Therefore, the characteristics of the shop can be reflected more stably for a longer time period.
Fig. 7 shows a result of three-dimensional clustering, which is similar to the state reflected by the re-mapping, and the associated business circle set divided by the three-dimensional clustering operation more accurately reflects the business circles formed by the shops that can influence each other.
Therefore, the plurality of business district prediction models trained according to the division result of the associated business district set can output required analysis data more accurately.
Specifically, the shop order trend prediction method further comprises the following steps: and taking all historical orders of all shops in the associated business district set within a set observation period as training set data.
The observation period may be set according to the requirement of data analysis, such as weekly analysis or daily analysis, and as a preferable scheme, the observation period is set to be one week.
As shown in fig. 6, assuming that 6 stores are shared in a certain related business district set, and the observation period is one week, historical order data of one week is collected at this time, and the following data is acquired for each store ID: order data, total order amount (sum of all order amounts), quantity SKU, and value SKU. Then, they are formed into a matrix in the manner of fig. 6 to be used as input data of the business circle prediction model, that is, the observation feature data is a matrix expressing stores, commodities and their corresponding relations, wherein the corresponding relations are attribute corresponding relations after the commodities are collected.
More specifically, the observation feature data is constructed in a matrix manner, the matrix comprises five columns which are respectively the shop ID, the order data, the order total amount, the quantity SKU and the value SKU of the corresponding same shop, and the rows of the matrix respectively correspond to different shops in the associated business district set.
The quantity SKU is the SKU code of the type of commodity with the largest quantity after all orders are aggregated, and the value SKU is the SKU code of the type of commodity with the highest aggregated price after all orders are aggregated.
Meanwhile, preferably, the output data of the business district prediction model (business district order prediction data) is also a same matrix, that is, the business district order prediction data is also a matrix expressing stores, commodities and their corresponding relations, that is, the matrix introduced above. The information of the number of possible orders, the total amount of the orders and the like of a certain shop in the next observation period can be obtained through the matrix.
The quotient circle prediction model can be a BP neural network model or a logistic regression analysis model.
By adopting the scheme of the associated business district set and the business district prediction model introduced above, the store classifications with strong business association can be collected together for analysis, so that the order trend of the stores can be analyzed and predicted from the perspective of the whole business district.
As another aspect of the present application, as shown in fig. 9, the present application also provides a store merchandise analysis method including the steps of: acquiring historical purchase order data of a shop; generating three-dimensional attribute coordinates of the shop according to historical purchase order data of the shop; performing K-Means clustering operation by using the coordinate value of the three-dimensional attribute coordinate of the shop; dividing the result of the K-Means clustering operation into a plurality of shop attribute sets; constructing order prediction models corresponding to different store attribute sets; and inputting the historical purchase order data of the shop to an order prediction model corresponding to the attribute set of the shop to which the historical purchase order data belongs.
Specifically, the three-dimensional attribute coordinate construction includes the following steps: setting a commodity classification table to classify commodities into fast-moving commodities, life commodities and stationery commodities; classifying the commodities in the historical purchase order data of one shop into the classifications of the commodity classification table respectively according to the commodity classification table; calculating the total classification price of the commodities in the classification of each commodity classification table of the shop; and establishing a coordinate system of the three-dimensional attribute coordinate by taking the three classifications of the commodity classification table as coordinate axes respectively, and taking the total classification price of the stores in the three classifications as coordinate values.
Preferably, the system can set the length of the time period for collecting the historical purchase orders according to the analysis requirement, such as the time period of the season or the year. Preferably, the length of the collection time period is set to year if a more stable order prediction model and store attribute set is to be obtained.
When the time period length is year, the above method specifically is: the method comprises the steps of summarizing annual purchase orders of a shop, classifying the commodities in the summary into three classifications according to the three classifications in a commodity classification table, and respectively counting the total classification prices of all the commodities in the three classifications, wherein the total classification prices of the shop under the three classifications are coordinate values of the shop in a three-dimensional attribute coordinate, and the counting unit of the total classification prices is hundred yuan in view of the length of a time period, so that the coordinate values are not too large, and the coordinate points of a representative circuit are not too discrete due to the computing unit in clustering operation.
Through the three-dimensional attribute coordinate establishment and the clustering operation, the stores can be divided into different store attribute sets. According to an ideal state, according to the design concepts of fast-moving goods, life goods and stationery goods, the shops are divided into corresponding business circle types, cell types and school types, wherein the fast-moving goods in the shopping orders of the business circle types are the main categories of the purchased commodities; the life category in the shop purchase order of the district type is the main purchase commodity category; the stationery in the school-type shop purchase order is the main purchase commodity category. Alternatively, the fast-extinction class may include: beverages, snacks, instant noodles, and the like; the life categories may include: seasonings, cleaning agents and articles of daily use; the stationery items may include: stationery, toys, etc.
In the process of actual data sorting and analysis, the attributes of a plurality of shops are found to be complex, if the stores are classified into a quotient circle type, a cell type and a school type, the classified highest classification total price can be adopted to belong to the classification, namely the classification, but the model training is difficult due to the fact that simple classification is found through later-stage model construction and verification. For example, even a shop located in school type has a problem that the purchase amount of fast-moving goods is larger than that of stationery goods. Therefore, simple classification cannot bring practical value to later analysis and model construction.
By adopting the scheme, the stores can be divided into the store attribute set according to the actual situation through dimension division and three-dimensional clustering, and the store attribute set reflects the actual attributes rather than classification attributes formed by artificial division.
After the division of the store attribute sets, the stores in the store attribute sets can be considered to have similar attribute characteristics, and based on the attribute characteristics, an order prediction model is constructed for each store attribute set.
The order prediction model is a prediction model whose inputs are historical data and whose outputs are prediction data and corresponding confidence levels.
From a training perspective, all historical purchase orders for all stores in the store attribute set are sorted into a matrix as shown in FIG. 11. Only one row of the matrix corresponds to a purchase order, and the columns are divided into an order amount, a quantity SKU, and a value SKU. And taking the historical purchase order of one shop as corresponding training data to train the order prediction model.
And the value SKU is the SKU value of the commodity with the highest payment amount in the purchase order.
Preferably, an observation period may be set, the order prediction model is not limited by the observation period when being trained, and the input data may be historical data in the observation period when the order prediction model is used, such as a week or a month.
The order prediction model may be a BP neural network model or a logistic regression analysis model.
In this way, the order prediction model belonging to a certain store attribute set can predict the amount of money of the next purchase order of the store, the product with the largest purchase quantity and the product with the highest purchase total price by inputting the history data of a certain store in the store attribute set.
Preferably, the method further comprises the following steps: and judging whether the confidence coefficient of the shop order prediction data is less than or equal to a preset threshold value for times of more than or equal to 3, and if the confidence coefficient of the shop order prediction data is more than or equal to 3, taking the shop order data with the maximum confidence coefficient output by the current order prediction model as analysis data. Due to the adoption of the store attribute set classification mode, generally speaking, the confidence problem does not occur, and the data with the highest confidence output by the model can also be used as the analysis data.
Through the three methods, the recommended order data, the business circle order prediction data and the store operation type are obtained from the store operation self-perspective.
The system compares and analyzes the suggested order data, the business district order prediction data and the shop order prediction data to obtain further application functions and analysis functions.
As one preferable scheme, after the suggested order data is generated, the corresponding business district order prediction data and the corresponding shop order prediction data can be compared, and the difference between the business district order prediction data and the shop order prediction data is analyzed, so that the accuracy of the suggested order data is determined.
As another scheme, the business circle order prediction data and the shop order prediction data are compared, so that the deviation between the business circle and the business direction is analyzed, and a suggestion is given to the user to enable the business direction to be matched with the business circle.
As another aspect of the present application, as shown in fig. 12, the present application also provides a server 100, i.e., a device executing a program, which includes a memory 101 and a processor 102. Wherein the memory 101 is adapted to store a computer program and the processor 102 is adapted to carry out the steps of the method as provided above when executing the computer program.
The present application also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method as provided above.
The present application also provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method as provided above.
The above description is only a preferred embodiment of the present application and is not intended to limit the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (10)
1. A shop order trend prediction method based on artificial intelligence is characterized by comprising the following steps:
the shop order trend prediction method based on artificial intelligence comprises the following steps:
dividing the shop into a plurality of associated business circle sets according to preset rules;
generating observation characteristic data according to all historical orders of all shops in the associated business district set in a set observation period;
inputting the observation characteristic data into a business district prediction model so that the business district prediction model outputs business district order prediction data and corresponding business district confidence;
and judging whether the business circle confidence coefficient is larger than a preset business circle confidence coefficient threshold value or not, and if the business circle confidence coefficient is larger than the preset business circle confidence coefficient threshold value, obtaining the shop order prediction data of one shop in the associated business circle set at least according to the business circle order prediction data.
2. The store order trend prediction method based on artificial intelligence of claim 1, wherein:
the step of dividing the plurality of stores into the associated business district sets according to preset rules comprises the following steps:
establishing a two-dimensional coordinate system according to the geographic position;
acquiring coordinate values of the shop in the two-dimensional coordinate system;
performing K-Means clustering operation on the coordinate values of the stores in the two-dimensional coordinate system;
and dividing the associated business circle set according to the result of the K-Means clustering operation.
3. The store order trend prediction method based on artificial intelligence of claim 2, wherein:
the step of dividing the plurality of stores into the associated business district sets according to preset rules further comprises the following steps:
acquiring historical order data of the shop;
calculating the average order value of the shop according to the historical order data of the shop, wherein the average order value is the average value of the order values of all purchase orders of the shop in the observation period;
establishing a three-dimensional coordinate system by taking the average order value as a third dimension;
acquiring coordinate values of the shop in the three-dimensional coordinate system;
performing K-Means clustering operation on the coordinate values of the stores in the three-dimensional coordinate system;
and dividing the associated business circle set according to the K-Means clustering operation result.
4. The store order trend prediction method based on artificial intelligence of claim 1, wherein:
the shop order trend prediction method based on artificial intelligence further comprises the following steps:
and taking all historical orders of all shops in the associated business district set within a set observation period as training set data.
5. The store order trend prediction method based on artificial intelligence of claim 4, wherein:
the business circle prediction model is a BP neural network model.
6. The store order trend prediction method based on artificial intelligence of claim 4, wherein:
the quotient circle prediction model is a logistic regression analysis model.
7. The store order trend prediction method based on artificial intelligence of claim 1, wherein:
the observation characteristic data is a matrix expressing stores, commodities and corresponding relations of the stores and the commodities.
8. The store order trend prediction method based on artificial intelligence of claim 7, wherein:
the business district order prediction data is a matrix expressing stores, commodities and corresponding relations of the stores and the commodities.
9. The shop order trend prediction device based on artificial intelligence is characterized in that:
the shop order trend prediction device based on artificial intelligence comprises:
a memory for storing a computer program;
a processor for implementing the artificial intelligence based store order trend prediction method of any one of claims 1 to 8 when executing the computer program.
10. A computer client storage medium, characterized in that: the computer-readable storage medium has stored therein a computer program which, when executed by a processor, implements the artificial intelligence based store order trend prediction method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110760703.4A CN113592539B (en) | 2021-06-30 | 2021-06-30 | Store order trend prediction method and device based on artificial intelligence and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110760703.4A CN113592539B (en) | 2021-06-30 | 2021-06-30 | Store order trend prediction method and device based on artificial intelligence and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113592539A true CN113592539A (en) | 2021-11-02 |
| CN113592539B CN113592539B (en) | 2024-09-24 |
Family
ID=78246093
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110760703.4A Active CN113592539B (en) | 2021-06-30 | 2021-06-30 | Store order trend prediction method and device based on artificial intelligence and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113592539B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP4231212A1 (en) * | 2022-02-21 | 2023-08-23 | Jio Platforms Limited | System and method for consumption estimation among users |
| CN117371826A (en) * | 2023-12-07 | 2024-01-09 | 福建科德信息技术服务有限公司 | Enterprise management method and system based on big data |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10240799A (en) * | 1997-02-27 | 1998-09-11 | Nri & Ncc Co Ltd | Support device for new store opening evaluation |
| JP2016167149A (en) * | 2015-03-09 | 2016-09-15 | Kddi株式会社 | Program, equipment, and method that estimate number of persons with which contract is expected to be made at shop according to market area |
| JP2016170778A (en) * | 2015-03-10 | 2016-09-23 | 技研商事インターナショナル株式会社 | Market area analysis system |
| CN106203746A (en) * | 2016-09-30 | 2016-12-07 | 携程计算机技术（上海）有限公司 | Hotel group divides and the method for requirement forecasting |
| CN106528611A (en) * | 2016-09-28 | 2017-03-22 | 西南交通大学 | Analysis method based on internet comment data |
| CN108596648A (en) * | 2018-03-20 | 2018-09-28 | 阿里巴巴集团控股有限公司 | A kind of commercial circle determination method and device |
| CN109684211A (en) * | 2018-12-18 | 2019-04-26 | 北京顺丰同城科技有限公司 | A kind of order dispatch system pressure test method and device based on shop dimension |
| CN111178975A (en) * | 2019-12-31 | 2020-05-19 | 北京顺达同行科技有限公司 | Business circle dividing method and device, electronic equipment and storage medium |
-
2021
- 2021-06-30 CN CN202110760703.4A patent/CN113592539B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10240799A (en) * | 1997-02-27 | 1998-09-11 | Nri & Ncc Co Ltd | Support device for new store opening evaluation |
| JP2016167149A (en) * | 2015-03-09 | 2016-09-15 | Kddi株式会社 | Program, equipment, and method that estimate number of persons with which contract is expected to be made at shop according to market area |
| JP2016170778A (en) * | 2015-03-10 | 2016-09-23 | 技研商事インターナショナル株式会社 | Market area analysis system |
| CN106528611A (en) * | 2016-09-28 | 2017-03-22 | 西南交通大学 | Analysis method based on internet comment data |
| CN106203746A (en) * | 2016-09-30 | 2016-12-07 | 携程计算机技术（上海）有限公司 | Hotel group divides and the method for requirement forecasting |
| CN108596648A (en) * | 2018-03-20 | 2018-09-28 | 阿里巴巴集团控股有限公司 | A kind of commercial circle determination method and device |
| CN109684211A (en) * | 2018-12-18 | 2019-04-26 | 北京顺丰同城科技有限公司 | A kind of order dispatch system pressure test method and device based on shop dimension |
| CN111178975A (en) * | 2019-12-31 | 2020-05-19 | 北京顺达同行科技有限公司 | Business circle dividing method and device, electronic equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| HONGYAN DAI ET AL.: ""A Clustering-based Sales Forecast Method for Big Promotion Days in O2O On-Demand Retailing"", 《IEEE》, 31 December 2019 (2019-12-31), pages 1079 - 1083, XP033701125, DOI: 10.1109/IEEM44572.2019.8978778 * |
| 吴珂;熊黑钢;朱海勇;张景秋;孟斌;: "北京市朝阳区不同商圈在售商铺价格变化分析", 资源开发与市场, no. 11, 15 November 2013 (2013-11-15) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP4231212A1 (en) * | 2022-02-21 | 2023-08-23 | Jio Platforms Limited | System and method for consumption estimation among users |
| CN117371826A (en) * | 2023-12-07 | 2024-01-09 | 福建科德信息技术服务有限公司 | Enterprise management method and system based on big data |
| CN117371826B (en) * | 2023-12-07 | 2024-03-15 | 福建科德信息技术服务有限公司 | Enterprise management method and system based on big data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113592539B (en) | 2024-09-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7340456B2 (en) | Dynamic feature selection for model generation | |
| JP7121118B2 (en) | Demand forecasting using weighted mixed machine learning models | |
| KR102447055B1 (en) | Method and Apparatus for Predicting Demand of product | |
| CN106067102A (en) | The optimization method of layout for storekeeping and optimization device | |
| CN113592539B (en) | Store order trend prediction method and device based on artificial intelligence and storage medium | |
| CN113554455B (en) | Store commodity analysis method, device and storage medium based on artificial intelligence | |
| CN110322093A (en) | Information processing method, information display method, device and calculating equipment | |
| US20230367768A1 (en) | Anonymization of query information while retaining query structure and sizing information | |
| CN111679814A (en) | Data-driven data center system | |
| Alizadeh et al. | Suitable delivery system in small e-commerce companies | |
| CN117934124A (en) | Big data-based electronic commerce order information intelligent processing system | |
| CN108629467A (en) | A kind of sample message processing method and system | |
| Nanda et al. | A multi-agent coalition-based approach for order fulfilment in e-commerce | |
| Yerragudipadu et al. | An Efficient Novel Approach on Machine Learning Paradigmsfor Food Delivery Company through Demand Forecastıng in societal community | |
| CN115797014A (en) | Home decoration industry online and offline commodity recommendation method and device and storage medium | |
| US11861677B2 (en) | Machine learning model for click through rate prediction using three vector representations | |
| CN113554453A (en) | Automatic order generation method and device based on artificial intelligence and storage medium | |
| Kong et al. | Demand-predictive storage assignment mechanism for flower auction centers | |
| US11803868B2 (en) | System and method for segmenting customers with mixed attribute types using a targeted clustering approach | |
| Sakhare et al. | Predictive Analysis of End to End Inventory Management System for Perishable Goods | |
| CN111768139B (en) | Stock processing method, apparatus, device and storage medium | |
| CN113052381A (en) | E-commerce marketing and management system based on big data | |
| Ariffin et al. | Product Pairing Selection for Promotion using Partitioning Method | |
| da Silva Alves | Predicting product sales in fashion retailing: a data analytics approach | |
| US20230196442A1 (en) | Selecting methods to allocate shoppers for order fulfillment in geographic regions by an online concierge system based on machine-learned efficiencies for different allocation methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |