CN111724176A - Store flow adjustment method, device, device and computer-readable storage medium - Google Patents

Abstract

The present disclosure provides a method, apparatus, device, and computer-readable storage medium for adjusting store traffic. The method includes: predicting the data of the store to be predicted according to the prediction model, and outputting the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the store in the bottleneck period; comparing the traffic threshold and the actual traffic of the store, and based on The preset allocation method allocates traffic to stores according to the comparison results. In the method, device, device and computer-readable storage medium provided by the present disclosure, a prediction model is obtained by training according to the existing bottleneck period store data, and then the flow threshold value of other stores when they reach the bottleneck period is predicted through the prediction model, so that the flow threshold value can be calculated according to the flow threshold value. Regulating the traffic distribution to the store can reduce the waste of traffic on the basis of ensuring the sales of the store, and can improve the overall traffic conversion rate.

Description

Store flow adjustment method, device, device and computer-readable storage medium

technical field

The present disclosure relates to a store flow distribution technology, and in particular, to a store flow adjustment method, apparatus, device, and computer-readable storage medium.

Background technique

With the development of Internet technology, online shopping has attracted more and more attention. At present, there are already many mature electronic malls, and many online stores are set up in these electronic malls, and users can shop products in the online stores.

Generally speaking, the sales of a store are related to the number of users who enter the store to browse, that is, to traffic, and to the conversion rate of browsing users to purchase users, and also to the order price when users purchase products. In order to increase the sales of the store, in the prior art, precision marketing is often used to recommend products of interest to users, thereby increasing the traffic of the store and achieving the purpose of increasing the sales of the store.

However, the inventor found that the sales of stores did not increase proportionally with the traffic, and the sales of many stores did not increase significantly after the traffic reached a certain value. Therefore, only using precision marketing to attract traffic to the store has the problem that even if it brings traffic to the store, the traffic will not contribute to the increase in sales of the store, which will lead to the waste of this part of the traffic resources.

SUMMARY OF THE INVENTION

The present disclosure provides a method, device, device, and computer-readable storage medium for adjusting store traffic, so as to solve the problem in the prior art that only accurate marketing is used to attract traffic to stores. It will contribute to the store's increase in sales, which will lead to the problem of wasting this part of traffic resources.

A first aspect of the present disclosure is to provide a method for adjusting store traffic, including:

Predict the data of the store to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the store in the bottleneck period;

Compare the traffic threshold with the actual traffic of the store, and allocate traffic to the store according to the comparison result based on a preset allocation method.

Another aspect of the present disclosure is to provide a store flow adjustment device, comprising:

A prediction module, configured to predict the data of the store to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the stores in the bottleneck period;

The distribution module is configured to compare the flow threshold with the actual flow of the store, and distribute the flow to the store based on the comparison result based on a preset distribution method.

Yet another aspect of the present disclosure is to provide a store traffic regulation device, including:

memory;

processor; and

Computer program;

Wherein, the computer program is stored in the memory, and is configured to be executed by the processor to implement the method for adjusting store traffic according to the first aspect above.

Yet another aspect of the present disclosure is to provide a computer-readable storage medium on which a computer program is stored, the computer program being executed by a processor to implement the method for adjusting store traffic according to the first aspect above.

The technical effects of the store traffic adjustment method, device, device and computer-readable storage medium provided by the present disclosure are:

The store traffic adjustment method, device, device and computer-readable storage medium provided by the present disclosure include: predicting the data of the store to be predicted according to the prediction model, and outputting the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is based on the data of the store in the bottleneck period. Obtained by model training; compares the traffic threshold with the actual traffic of the store, and allocates traffic to the store based on the comparison result based on the preset allocation method. In the method, device, device and computer-readable storage medium provided by the present disclosure, a prediction model is obtained by training according to the existing data of stores in the bottleneck period, and then the traffic thresholds of other stores when they reach the bottleneck period are predicted by the prediction model, so that the traffic thresholds can be calculated according to the traffic thresholds. Regulating the traffic distribution to the store can reduce the waste of traffic on the basis of ensuring the sales of the store, and can improve the overall traffic conversion rate.

Description of drawings

1A is a schematic diagram showing the relationship between store traffic and sales according to an exemplary embodiment;

FIG. 1B is a schematic diagram showing the relationship between store traffic and sales according to another exemplary embodiment;

2 is a flowchart of a method for adjusting store traffic according to an exemplary embodiment of the present invention;

3 is a flowchart of a method for adjusting store traffic according to another exemplary embodiment of the present invention;

4 is a structural diagram of a store flow adjustment device according to an exemplary embodiment of the present invention;

5 is a structural diagram of a store flow adjustment device according to another exemplary embodiment of the present invention;

FIG. 6 is a structural diagram of a store flow adjustment device according to an exemplary embodiment of the present invention.

Detailed ways

FIG. 1A is a schematic diagram showing the relationship between store traffic and sales according to an exemplary embodiment; FIG. 1B is a schematic diagram showing the relationship between store traffic and sales according to another exemplary embodiment.

At present, in order to increase the sales of online stores, the mall where the store is located can attract traffic to the store and attract users to the store in various ways. However, the inventor found that in a certain stage, the sales volume of the store will increase with the increase of store traffic (as shown in Figure 1A), and in another stage, although the traffic in the store is increasing, the sales volume does not increase significantly trend (see the second stage in Figure 1B). If the traditional traffic drainage method is still used to attract users, this part of users will not be able to bring obvious benefits to the store, that is, this part of the user traffic is invalid, resulting in a waste of traffic.

The flow here may be a flow value within a preset time period, for example, a daily flow value, a weekly flow value, and the like.

FIG. 2 is a flowchart of a method for adjusting store traffic according to an exemplary embodiment of the present invention.

As shown in Figures 1A and 1B, for the stage where sales increase with the increase of traffic, we can call it a growth stage, and stores in this stage can be called potential growth stage stores. For the stage where the sales no longer increase with the increase of traffic, we can call it the slow stage, and the stores in this stage can be called the bottleneck stage stores. With the increase of traffic, sales will increase to a certain extent. When the traffic increases to a certain level, the sales of the store will no longer increase. The method provided by this embodiment utilizes the inflection point between these two stages to allocate traffic to the store. Thereby, the store is at the maximum sales value, and there will be no problem of wasting traffic.

Step 201: Predict the store data to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by model training according to the store data in the bottleneck period.

The prediction model may be pre-trained, and the equipment for training the model may be the same as or different from the equipment for executing the method of this embodiment. The sample data of the stores in the bottleneck period can be collected in advance, and the model can be trained according to the data of the stores in the bottleneck period. The specific training method may be similar to the model training method in the prior art.

Specifically, the sample data of each store in the bottleneck period may have a label, and the label may be the flow value corresponding to the inflection point when the store enters the slow period from the growth period. The sample data can be input into the model, and the flow value can be input into the model as a label, so as to train the parameters in the model, and the sample data can be processed based on these parameters to obtain the corresponding label of the sample data. During training, the parameters in the training model can be initialized, and then the parameters can be adjusted based on the data of the stores in the bottleneck period and their corresponding labels. After training with a large amount of data of the stores in the bottleneck period, accurate model parameters can be obtained. The model processes other store data, and the model can output the corresponding traffic threshold.

Further, the data of the store to be predicted can be predicted according to the prediction model obtained by training. The prediction model is stored in an electronic device with computing function, so that the electronic device can predict the flow of the store based on the prediction model. For example, the electronic device can be a computer, and the data of the store to be predicted can be input into the prediction model stored in the computer. , the prediction model can output the traffic threshold corresponding to the store to be predicted.

In practical application, the method provided in this embodiment may also be packaged in an electronic device, so that the electronic device can execute the method provided in this embodiment. The electronic device can obtain the data of the store to be predicted, and the data can be stored in the electronic device or in other devices, and the electronic device can read the data of the store to be predicted. Databases for storing store data to be predicted can be set up, and the electronic devices can access these databases.

The user may also select some specified store data as the store data to be predicted. For example, if the user wishes to predict the traffic threshold of store A, the data of store A can be used as the data of the store to be predicted.

Specifically, the data of the store to be predicted can be input into the prediction model, and the prediction model calculates the data according to the internal parameters, obtains the traffic threshold corresponding to the store, and outputs the traffic threshold. The traffic threshold refers to the traffic value corresponding to the inflection point when the store enters the slow period from the growth period. When the traffic value of the store is less than the traffic threshold, the sales of the store can increase with the increase of traffic. When the sales of the store is greater than or equal to the traffic threshold , the store's sales no longer increase with the increase in traffic.

Step 202 , compare the traffic threshold with the actual traffic of the store, and allocate traffic to the store according to the comparison result based on a preset allocation method.

Further, the traffic can be allocated to the store according to the traffic threshold, thereby increasing the sales brought by the traffic to the store, and the problem of wasting traffic will not occur.

In practical application, when the store traffic exceeds the traffic threshold, the store's sales will not increase with the increase of traffic; when the store traffic is less than the traffic threshold, the store's sales will increase with the increase of traffic. Therefore, it is possible to It is considered that when the store traffic fluctuates up and down the traffic threshold, the store's traffic conversion rate is the highest, that is, the traffic utilization rate is the highest. The conversion rate refers to the proportion of the user traffic that enters the store and the user traffic who purchases the product.

Among them, through the deployment of store traffic, the store traffic can be stabilized near the traffic threshold, so that the utilization rate of the store traffic is the highest. For example, the current traffic of the store can be monitored. If the current traffic is greater than the traffic threshold, the traffic allocated to the store can be reduced, and if the current traffic is less than the traffic threshold, the traffic allocated to the store can be increased. If the current traffic is equal to the traffic threshold, the amount of traffic allocated to the store may not be changed.

Specifically, in addition to the mall allocating traffic to the store through traffic drainage, there will also be users who actively enter the store. Therefore, the average daily traffic and the results of the traffic threshold and the actual traffic of the store can be integrated to make a reasonable traffic distribution to the store. . In addition, parameters for allocating traffic can be set, and based on these parameters, traffic can be allocated to the store. For example, weight values corresponding to different distribution methods can be set, the number of traffic corresponding to each distribution method can be determined according to these weight values, and then based on the different distribution methods and their corresponding traffic numbers, traffic is allocated to the store.

The method provided in this embodiment is used to adjust store traffic, and the method is executed by a device provided with the method provided in this embodiment, and the device is usually implemented in hardware and/or software.

The store traffic adjustment method provided in this embodiment includes: predicting the data of the store to be predicted according to the prediction model, and outputting the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the stores in the bottleneck period; comparing the traffic The threshold value and the actual flow of the store, and based on the preset distribution method, the flow is allocated to the store according to the comparison result. In the method provided in this embodiment, a prediction model is obtained by training according to the existing data of stores in the bottleneck period, and then the traffic thresholds of other stores when they reach the bottleneck period are predicted through the prediction model, so that the traffic distribution to the stores can be regulated and controlled according to the traffic thresholds. On the basis of ensuring store sales, reducing the waste of traffic can improve the overall traffic conversion rate.

FIG. 3 is a flowchart of a method for adjusting store traffic according to another exemplary embodiment of the present invention.

As shown in FIG. 3 , the method for adjusting store traffic provided by this embodiment includes:

Step 301: Divide the store data in the bottleneck period according to the time corresponding to the data to form multiple data sets.

In the method provided in this embodiment, the step of training a prediction model may also be included. The data of the stores in the bottleneck period can be obtained, and the model can be trained based on the data to obtain the prediction model. Since the stores in the bottleneck period have transitioned from the growth period to the slow period, the stores in the bottleneck period have known traffic threshold information. The store data can reflect the characteristics of the store. Therefore, it can be considered that the store data has an associated relationship with the traffic threshold. By analyzing a large amount of store data, the relationship between the store data and the traffic threshold can be determined to obtain a prediction model.

Among them, the store data has a time attribute, that is, traffic and sales information corresponding to a certain time period, and other information. For example, in the first month, the average daily traffic value of the store is 1000, the sales volume is 50000, and the store rating is 4.7. In the second month, the average daily traffic value of the store is 1300, the sales volume is 55000, and the store rating is 4.8.

Specifically, according to the time attribute of the data, the store data in the bottleneck period can be divided to obtain multiple data sets. Specifically, the generation time corresponding to the store data in the bottleneck period may be divided into multiple time windows of length L, and the store data corresponding to each time window is a data set. For example, if the store data in the bottleneck period is the data of the store for two years, the data of the stores in the bottleneck period can be divided into 8 data sets according to 3 months as a set.

Step 302: Train the model according to each data set to obtain a prediction model.

Among them, the store data in the bottleneck period is divided into multiple data sets, so that the amount of basic data used for training the model is large. Moreover, it can be trained according to the data of stores in different periods, and it is more adaptable when using the trained model to predict the results.

Specifically, since the traffic threshold corresponding to the store is known, after dividing the data corresponding to the store, the traffic threshold corresponding to the obtained data set is also known. Therefore, the labels corresponding to each dataset are still the traffic thresholds corresponding to the stores.

In practical applications, the data set and traffic threshold can be input into the model, so as to adjust the built-in parameters of the model and complete the model training.

Wherein, a preset time can be preset, and the data at the preset time and later in the data set is used as the training set, and the prediction model is trained according to the training set, and then the data at the preset time and later is used as the test set, using The prediction model obtained from the test training. Since the data used in the training set and the test set are the same, the consistency of the data distribution of the training set and the test set can be guaranteed to a large extent.

Specifically, during the test, the result corresponding to the test set data can be output, and the result can be compared with the label corresponding to the test set data, that is, the flow threshold, to obtain the test error. If the test error is greater than the preset allowable error , the above preset time can be adjusted, the training set can be re-determined, and the prediction model can be re-trained.

If the test error is less than or equal to the allowable error, the current prediction model can be used as the final prediction model, and the prediction model can be used to make predictions based on the store data to be predicted.

Further, if the number of times of re-determining the training set exceeds the preset value, the preset value can be set according to the requirements. At this time, it can be considered that the training model has failed, and the store data in the bottleneck period is re-selected, and then based on the new store data. Step 301 is executed.

Step 303: Predict the data of the store to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted.

The specific principles and implementation manners of step 303 are similar to those of step 201, and are not repeated here.

The type of store data to be predicted is consistent with the type of store data in the bottleneck period used to train the prediction model, including at least one of the following:

Store opening time, main categories, main brands, basic attributes of main brands, average daily sales, average daily traffic, average order price, attention, and ratings. The basic attributes of the main brand include: brand level (such as first-tier, second-tier, and third-tier brands), brand type (such as online and offline brands, international and domestic brands), brand characteristics (such as business, leisure, trendy brands, etc. ).

Specifically, if some information in the predicted store data is missing, 0 can be used to replace the corresponding attribute information, so that the type of store data to be predicted entered into the prediction model is consistent with the type of store data in the bottleneck period.

Step 304, compare the traffic threshold with the actual traffic of the store.

The specific principle and implementation manner of comparing the flow threshold in step 304 and step 202 are similar to the actual flow of the store, and details are not repeated here.

Step 305, if the flow threshold is greater than the actual flow, increase the flow of the store; if the flow threshold is less than the actual flow, reduce the flow of the store.

Further, if the comparison result is that the traffic threshold is greater than the actual traffic, it can be considered that the current store is in a growth period, and more traffic is allocated to the store, so the sales of the store still have room for growth, so the traffic of the store can be increased. Specifically, according to the difference between the traffic threshold and the actual traffic, it can be determined to increase the amount of allocated traffic for the store.

In practical applications, if the comparison result is that the traffic threshold is less than the actual traffic, it can be considered that the current store is in a slow period. Even if more traffic is allocated to the store, the sales will not increase significantly. Therefore, the traffic of the store can be reduced. Specifically, according to the difference between the traffic threshold and the actual traffic, it can be determined to reduce the amount of traffic allocated for the store.

Among them, the reasonable allocation of traffic to the store according to the traffic threshold can improve the overall conversion rate of traffic and avoid traffic waste.

Specifically, parameters for allocating traffic can also be set, for example, weight values corresponding to different allocation methods, and parameters used when allocating traffic are adjusted according to the difference between the traffic threshold and the actual traffic.

Step 306: Determine the predicted sales of the store based on the preset prediction method, and monitor the actual sales of the store.

Wherein, in the method provided in this embodiment, after adjusting the dedicated traffic distribution to the store, the sales volume of the store can also be predicted. For example, a forecasting model may be preset, and the forecasted sales of the store may be determined based on the forecasting model. The prediction model can be any one of regression, random forest, neural network model, etc.

Specifically, it is also possible to estimate the total flow of the store after adjusting the distribution of traffic, and to determine the average value of orders in the store according to historical data, as well as the traffic conversion rate corresponding to the store. Multiply to get the store's forecast sales.

Further, the actual sales of the store can also be monitored. The electronic device implementing this embodiment can be connected to the backend server of the mall to obtain store sales information stored in the server. The specific connection method can be connected through a network, which can be wired or wireless.

In practical applications, the sales of the store can be predicted in advance for a period of time, and after the period of time has passed, the actual sales generated during this period of time can be obtained. That is, the time when the predicted sales are generated should be consistent with the time when the actual sales are generated, so that the two are comparable.

Among them, the predicted sales value can be compared with the actual sales value. If the difference between the predicted sales and the actual sales is greater than the store prediction error, go to step 307 , otherwise go to step 308 .

Step 307, revise the preset allocation method and the forecast model according to the forecasted sales and the actual sales.

Specifically, the store prediction error corresponding to the store sales can be preset. If the difference between the predicted sales value and the actual sales value exceeds the store prediction error, it can be considered that there is a problem with the set flow distribution method or the parameters in the prediction model. , so the preset allocation and forecast model can be revised based on forecast sales and actual sales.

Step 308: Determine the overall predicted sales volume according to the predicted sales volume corresponding to each store.

If the difference between the predicted sales and the actual sales is less than or equal to the store prediction error, it can be considered that the current prediction model and distribution method are accurate for the store's data.

At this time, in order to more accurately determine whether the error generated by the prediction model and distribution method is acceptable for the entire mall, the overall predicted sales can also be determined according to the predicted sales corresponding to each store. Specifically, the predicted sales generated by each store in the same time period can be superimposed to obtain the overall predicted sales.

Step 309: Determine the overall actual sales according to the actual sales corresponding to each store.

Further, the actual sales generated by each store in the same time period can be superimposed to obtain the overall actual sales. This time period should be the same as the time period that generates the forecasted sales.

The execution order of step 308 and step 309 is not limited.

Step 310 , if the difference between the overall predicted sales and the overall actual sales is greater than the overall predicted error, modify the preset allocation method according to the overall predicted sales and the overall actual sales.

In practical application, the overall forecasted sales and the overall actual sales can be compared, and the overall forecast error can also be preset. If the difference between the overall forecasted sales and the overall actual sales is greater than the overall forecast error, it can be considered as a preset allocation method. It is not very accurate, resulting in a small error in the corresponding sales of each store, but a large error in the overall sales. Therefore, the preset allocation method can be revised according to the overall predicted sales and the overall actual sales.

Among them, the method provided in this embodiment can monitor the sales of the store after adjusting the flow allocated by the store, and further modify the prediction module and the preset distribution method according to the monitoring, so that the whole method is more accurate.

FIG. 4 is a structural diagram of a store flow adjustment device according to an exemplary embodiment of the present invention.

As shown in FIG. 4 , the store flow adjustment device provided in this embodiment includes:

The prediction module 41 is configured to predict the data of the store to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the stores in the bottleneck period;

The distribution module 42 is configured to compare the flow threshold with the actual flow of the store, and distribute the flow to the store according to the comparison result based on a preset distribution method.

The store traffic adjustment device provided in this embodiment includes a prediction module, which is used to predict the data of the store to be predicted according to the prediction model, and output the traffic threshold corresponding to the store to be predicted; wherein, the prediction model is obtained by performing model training according to the data of the stores in the bottleneck period. ; The allocation module is used to compare the traffic threshold with the actual traffic of the store, and allocate traffic to the store based on the comparison result based on the preset allocation method. The device provided in this embodiment can train a prediction model according to the existing data of stores in the bottleneck period, and then use the prediction model to predict the traffic thresholds of other stores when they reach the bottleneck period, so that the traffic distribution to the stores can be regulated according to the traffic thresholds. On the basis of ensuring store sales, reducing the waste of traffic can improve the overall traffic conversion rate.

The specific principle and implementation manner of the store flow adjustment device provided in this embodiment are similar to the embodiment shown in FIG. 2 , and details are not described herein again.

FIG. 5 is a structural diagram of a store flow adjustment device according to another exemplary embodiment of the present invention.

As shown in FIG. 5 , on the basis of the foregoing embodiment, in the apparatus provided in this embodiment, the allocation module 42 is specifically used for:

If the flow threshold is greater than the actual flow, increase the flow of the store;

If the flow threshold is smaller than the actual flow, reduce the flow of the store.

Optionally, the bottleneck period store data is consistent with the to-be-predicted store data category, including at least one of the following:

Store opening time, main categories, main brands, basic attributes of the main brands, average daily sales, average daily traffic, average order price, attention, and ratings.

Optionally, the basic attributes of the main brand include:

Brand level, brand type, brand characteristics.

The device provided in this embodiment further includes a training module 43 for:

Divide the store data in the bottleneck period according to the time corresponding to the data to form multiple data sets;

The model is trained according to each of the data sets to obtain the prediction model.

The training module 43 includes:

The training unit 431 is configured to use the data corresponding to the preset time in the data set as a training set, and train a model according to the training set to obtain the prediction model;

a testing unit 432, configured to use the data corresponding to the preset time in the data set as a test set, and test the prediction model according to the test set to obtain a test error;

The adjustment unit 433 is configured to adjust the preset time if the test error is greater than the allowable error, and the training unit 431 continues to execute the preset time in the data set according to the adjusted preset time Then the corresponding data is used as the training set step.

If the test error is less than or equal to the error threshold, the prediction module 41 uses the prediction model to make predictions according to the store data to be predicted.

The device also includes a correction module 44 for:

Determine the predicted sales of the store based on a preset prediction method, and monitor the actual sales of the store;

If the difference between the predicted sales and the actual sales is greater than the store prediction error, the preset allocation method and the prediction model are revised according to the predicted sales and the actual sales.

If the difference between the predicted sales and the actual sales is less than or equal to the store prediction error, the correction module 44 is further configured to:

Determine the overall predicted sales according to the predicted sales corresponding to each of the stores;

Determine the overall actual sales according to the actual sales corresponding to each of the stores;

If the difference between the overall predicted sales and the overall actual sales is greater than the overall prediction error, the preset allocation method is revised according to the overall predicted sales and the overall actual sales.

The specific principle and implementation manner of the apparatus provided in this embodiment are similar to the embodiment shown in FIG. 3 , and details are not described herein again.

FIG. 6 is a structural diagram of a store flow adjustment device according to an exemplary embodiment of the present invention.

As shown in FIG. 6 , the store traffic adjustment device provided in this embodiment includes:

memory 61;

processor 62; and

Computer program;

Wherein, the computer program is stored in the memory 61 and configured to be executed by the processor 62 to implement any one of the above-mentioned methods for adjusting store traffic.

This embodiment also provides a computer-readable storage medium on which a computer program is stored,

The computer program is executed by the processor to implement any of the store traffic regulation methods described above.

Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by program instructions related to hardware. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps including the above method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: The technical solutions described in the foregoing embodiments can still be modified, or some or all of the technical features thereof can be equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention. scope.

Claims (12)

1. A store traffic adjustment method, comprising:

predicting the data of the shop to be predicted according to the prediction model, and outputting a flow threshold corresponding to the shop to be predicted; the prediction model is obtained by performing model training according to bottleneck shop data;

and comparing the flow threshold with the actual flow of the shop, and distributing the flow to the shop according to a comparison result based on a preset distribution mode.

2. The method according to claim 1, wherein the allocating traffic to the stores according to the comparison result based on a preset allocation manner comprises:

if the flow threshold value is larger than the actual flow, increasing the flow of the shop;

and if the flow threshold value is smaller than the actual flow, reducing the flow of the shop.

3. The method of claim 1, wherein the bottleneck-in-store data is in accordance with the store-to-be-forecasted data category, comprising at least one of:

the method comprises the following steps of starting time, main business category, main business brand, basic attribute of the main business brand, daily average sales, daily average flow, average order price, amount of concern and rating.

4. The method of claim 3, wherein the base attributes of the hosted brand include:

brand level, brand type, brand features.

5. The method of claim 1, further comprising:

dividing the bottleneck shop data according to the time corresponding to the data to form a plurality of data sets;

and training a model according to each data set to obtain the prediction model.

6. The method of claim 5, wherein training the model from each of the data sets to obtain the predictive model comprises:

taking the corresponding data after the preset time in the data set as a training set, and training a model according to the training set to obtain the prediction model;

taking the corresponding data after the preset time in the data set as a test set, and testing the prediction model according to the test set to obtain a test error;

and if the test error is larger than the allowable error, adjusting the preset time, and continuing to execute the step of taking the corresponding data after the preset time in the data set as a training set according to the adjusted preset time.

7. The method of claim 6, wherein if the test error is less than or equal to an error threshold value, predicting according to the shop data to be predicted by using the prediction model.

8. The method according to claim 1, further comprising, after allocating the traffic to the store according to the comparison result based on a preset allocation manner,:

determining the predicted sales amount of the shop based on a preset prediction mode, and monitoring the actual sales amount of the shop;

and if the difference value between the predicted sales and the actual sales is larger than the prediction error of the shop, correcting the preset distribution mode and the prediction model according to the predicted sales and the actual sales.

9. The method of claim 8, wherein if the difference between the predicted sales and the actual sales is less than or equal to the store prediction error, the method further comprises:

determining the overall predicted sales according to the predicted sales corresponding to each shop;

determining the whole actual sales amount according to the actual sales amount corresponding to each shop;

and if the difference value between the overall predicted sales and the overall actual sales is larger than the overall prediction error, correcting the preset distribution mode according to the overall predicted sales and the overall actual sales.

10. A store flow adjustment device, comprising:

the prediction module is used for predicting the data of the shop to be predicted according to the prediction model and outputting a flow threshold corresponding to the shop to be predicted; the prediction model is obtained by performing model training according to bottleneck shop data;

and the distribution module is used for comparing the flow threshold value with the actual flow of the shop and distributing the flow to the shop according to a comparison result based on a preset distribution mode.

11. A store traffic conditioning apparatus, comprising:

a memory;

a processor; and

a computer program;

wherein the computer program is stored in the memory and configured to be executed by the processor to implement the method of any of claims 1-9.

12. A computer-readable storage medium, having stored thereon a computer program,

the computer program is executed by a processor to implement the method according to any one of claims 1-9.

Images