Article recommendation method, device and system based on big data analysis, and medium

Abstract

The invention provides an article recommendation method, device and system based on big data analysis, and a medium, and belongs to the technical field of data processing, and the method comprises the steps: obtaining a plurality of article classifications in a target region based on historical demand data of the target region; determining a first article of which the demand quantity is in the first N positions in each article classification; classifying each article, and dividing the N first articles into second articles and third articles based on the historical demand fluctuation values of the first articles; the third article is the first article of which the historical demand fluctuation value is greater than a first threshold value; predicting the current demand quantity of the third articles based on the historical demand quantity of the third articles, and selecting a fourth article from the third articles based on a prediction result; and recommending the second article and the fourth article to the user of the target area. According to the article recommendation method, device and system based on big data analysis, and the medium provided by the invention, reasonable allocation of articles can be realized, and waste of resources is reduced.

Description

Technical Field

[0001] The present disclosure belongs to the field of data processing technology, and more specifically, to an item recommendation method and device, system, and medium based on big data analysis. Background Art

[0002] With the development of intelligent technology, vending machines are becoming more and more popular in people's lives. Existing vending machine merchants often rely on experience to decide the products in the vending machines when selecting goods, resulting in the products not matching the actual needs of consumers well, causing some products to be unsalable and some products needed by consumers to be out of stock. Unsalable goods will lead to expiration of goods, which in turn causes waste of resources. When goods are out of stock, they need to be adjusted, which will consume manpower and material resources. Summary of the invention

[0003] The purpose of the present disclosure is to provide an item recommendation method and device, system, and medium based on big data analysis to achieve reasonable allocation of items and reduce waste of resources.

[0004] A first aspect of the embodiments of the present disclosure provides an item recommendation method based on big data analysis, comprising: Obtain multiple item classifications in the target area based on historical demand data of the target area; Determine the first item with the top N demand in each item category; For each item classification, N first items are divided into second items and third items based on the historical demand fluctuation value of the first items; the second items are first items whose historical demand fluctuation value is less than or equal to a first threshold, and the third items are first items whose historical demand fluctuation value is greater than the first threshold; Predicting a current demand for the third item based on a historical demand for the third item, and selecting a fourth item from the third items based on the predicted result of the current demand; The second item and the fourth item are recommended to users in a target area.

[0005] A second aspect of the embodiments of the present disclosure provides an item recommendation device based on big data analysis, including: A first processing module, configured to obtain a plurality of item classifications in a target area based on historical demand data of the target area; The second processing module is used to determine the first item with the top N demand in each item category; a data classification module, configured to classify each item, and divide N first items into second items and third items based on the historical demand fluctuation value of the first items; the second items are first items whose historical demand fluctuation value is less than or equal to a first threshold, and the third items are first items whose historical demand fluctuation value is greater than the first threshold; A data prediction module, used for predicting the current demand of the third item based on the historical demand of the third item, and selecting the fourth item from the third items based on the prediction result of the current demand; The item recommendation module is used to recommend the second item and the fourth item to users in a target area.

[0006] According to a third aspect of an embodiment of the present disclosure, there is provided an item recommendation system based on big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned item recommendation method based on big data analysis when executing the computer program.

[0007] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for recommending items based on big data analysis are implemented.

[0008] The beneficial effects of the method, device, system, and medium for recommending items based on big data analysis provided by the embodiments of the present disclosure are: The disclosed embodiment first analyzes the historical demand data of the target area to obtain multiple item classifications. In each item classification, the top N first items are selected based on the demand. Then, according to the historical demand fluctuation value of the first item, it is divided into a second item with stable demand and a third item with large demand fluctuation. For the second item with stable demand, it can be directly recommended to users in the target area (that is, vending machine merchants). For the third item with large demand fluctuation, the current demand is predicted based on the historical demand, and the fourth item with high current demand is selected from the third items according to the prediction results and recommended to users in the target area. The entire recommendation process is based on big data analysis, making full use of the rich information of historical demand data for decision-making, reducing the influence of subjective judgment, making the recommendation strategy more scientific and reasonable, thereby achieving reasonable configuration of items and reducing waste of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A flowchart of an item recommendation method based on big data analysis provided by an embodiment of the present disclosure; Figure 2 A structural block diagram of an item recommendation device based on big data analysis provided by an embodiment of the present disclosure; Figure 3 A schematic block diagram of an item recommendation system based on big data analysis provided by one embodiment of the present disclosure. DETAILED DESCRIPTION

[0011] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0012] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0013] Please refer to Figure 1 , Figure 1 A flowchart of an item recommendation method based on big data analysis provided in one embodiment of the present disclosure includes: S101: Obtain multiple item categories in a target area based on historical demand data of the target area.

[0014] In this embodiment, the target area may be one or more target areas, and the multiple target areas have similar geographical locations and environments, and thus have similar traffic, consumption scenarios, and consumption habits. For example, the multiple current areas are schools, office buildings, subway stations, or train stations.

[0015] The historical demand data is the demand data of items in the target area over the past period of time. The historical demand data can be obtained based on the sales data of items in vending machines in the target area. For example, the historical demand data can be the sales data of items in the target area over the past month, or the sales data of items in the target area over the past three months. The historical demand data is usually in the form of a table. The fields in the table may include item classification (including food, beverages, daily necessities, etc.), demand quantity, demand amount, demand time, etc. Based on the historical demand data, multiple item classifications in the target area can be obtained. S102: Determine the first item whose demand is ranked in the top N in each item category.

[0016] In this embodiment, for each item category, the items under the category can be sorted in descending order according to demand, and the top N items are selected as the first items. The first item can be considered as a popular item in the category, representing the main demand direction of consumers for this type of item.

[0017] S103: Classify each item and divide N first items into second items and third items based on the historical demand fluctuation value of the first item; the second item is a first item whose historical demand fluctuation value is less than or equal to a first threshold, and the third item is a first item whose historical demand fluctuation value is greater than the first threshold.

[0018] In this embodiment, the demand trends of different items are different. For example, for items such as drinking water and common snacks, since these items are indispensable in people's daily lives, the demand is relatively stable, and consumers will purchase them regularly, usually maintaining a high and stable demand. However, the demand for items such as cold drinks and hand warmers is affected by the season, and the demand for some Internet celebrity products has obvious popular trends.

[0019] Therefore, in order to accurately predict consumer demand for items, this embodiment divides the first item into a second item with stable demand and a third item with large demand fluctuations based on the historical demand fluctuation value of the first item, so as to adopt a targeted prediction strategy.

[0020] The historical demand fluctuation value of the first item can be characterized by calculating the relative standard deviation of the historical demand of the first item. The larger the relative standard deviation, the larger the historical demand fluctuation value of the first item. The first threshold is a preset constant, and the specific value of the first threshold can be obtained by statistically analyzing the historical demand data of multiple items.

[0021] S104: predicting the current demand for the third item based on the historical demand for the third item, and selecting a fourth item from the third items based on the predicted result of the current demand.

[0022] In this embodiment, for the third item with large fluctuations in demand, time series analysis methods (such as ARIMA model, exponential smoothing method, etc.) or machine learning algorithms (such as LSTM neural network, random forest regression, etc.) can be used to model the historical demand of the third item and predict its current demand. Then, according to the prediction results, certain screening criteria are set, such as selecting several items with high predicted demand as the fourth item.

[0023] Through the above process, the fourth item that meets the current market demand can be screened out from the third item, so as to more accurately grasp the market dynamics, avoid blind recommendations, and improve the timeliness and accuracy of recommendations.

[0024] S105: Recommend the second item and the fourth item to users in the target area.

[0025] In this embodiment, the users in the target area are merchants operating vending machines. The second item with stable demand can be directly recommended to the users in the target area. Therefore, on the basis of obtaining the fourth item, the second item and the fourth item can be recommended to the users together, and the users can choose the goods accordingly.

[0026] Specifically, the recommended item information can be presented to the user through various methods such as vending machine interface display, mobile phone application push, and SMS notification.

[0027] It can be concluded from the above that this embodiment first analyzes the historical demand data of the target area to obtain multiple item classifications. In each item classification, the top N first items are selected based on the demand. Then, according to the historical demand fluctuation value of the first item, it is divided into a second item with stable demand and a third item with large demand fluctuation. For the second item with stable demand, it can be directly recommended to users in the target area (that is, vending machine merchants). For the third item with large demand fluctuation, the current demand is predicted based on the historical demand, and the fourth item with high current demand is selected from the third item according to the prediction result and recommended to users in the target area. The entire recommendation process is based on big data analysis, making full use of the rich information of historical demand data for decision-making, reducing the influence of subjective judgment, making the recommendation strategy more scientific and reasonable, and thus achieving reasonable configuration of items.

[0028] In one embodiment of the present disclosure, predicting the current demand for the third item based on the historical demand for the third item includes: Determining the difference order based on the historical demand fluctuation value of the third item; randomly generating a plurality of first combinations of autoregressive orders and moving average orders based on a first limit value; Determine multiple first combinations of autoregressive order and moving average order as initial population, use genetic algorithm to perform multiple iterations to determine the optimal order combination; the optimal order combination is the optimal combination of autoregressive order, moving average order and difference order; Construct ARIMA model based on optimal order combination; Predict the current demand for the third item based on the ARIMA model.

[0029] In this embodiment, the Autoregressive Integrated Moving Average Model (ARIMA model) can be used to predict the current demand for the third item. The ARIMA model regards the time series as the result of the interaction of its own past values, past prediction errors, and current random disturbances. This relationship is described by establishing a mathematical model to predict future values.

[0030] The ARIMA model consists of an autoregressive part, a difference part, and a moving average part. The autoregressive part represents the linear relationship between the current value and the past value of the time series. The difference part is used to deal with the non-stationarity of the time series. If the time series has non-stationary characteristics such as trend or seasonality, the series can be differentiated to make it a stationary series. The moving average part takes into account the correlation between the error terms in the time series. Correspondingly, the ARIMA model is usually represented by ARIMA(p, d, q), where p, d, and q represent the autoregressive order, the difference order, and the moving average order, respectively. The appropriate autoregressive order, the difference order, and the moving average order can ensure the prediction accuracy of the ARIMA model.

[0031] In order to quickly and accurately determine the autoregressive order, the difference order and the moving average order, this embodiment first determines the difference order based on the historical demand fluctuation value of the third item. Specifically, if the demand fluctuation is large, a higher-order difference is required to make the data smooth; if the fluctuation is small, only a lower-order difference is required or even no difference is required.

[0032] After determining the appropriate d value to stabilize the data, the genetic algorithm can be used to determine the appropriate (p, q) combination. Specifically, the first limit value can be set in advance, and the value range of p and q can be determined based on the first limit value. For example, if the first limit value is set to 5, then the value range of p is 0 to 5, and the value range of q is 0 to 5. 50 different (p, q) combinations are randomly generated as the initial population. Based on the initial population, a suitable evaluation index (such as AIC based on the Akaike information criterion or BIC based on the Bayesian information criterion) is selected as the fitness function. For each (p, q) combination, combined with the determined d value, an ARIMA (p, d, q) model is constructed, the historical demand data of the third item is fitted, and the fitness function value is calculated. According to the fitness value of the individual, genetic operations such as selection, crossover and mutation are performed. Multiple iterations can gradually optimize these combinations to find the (p, q) combination that makes the model fit best. The best (p, q) combination is combined with the determined d value to obtain the optimal order combination (p, d, q).

[0033] Based on the optimal order combination (p, d, q), the ARIMA model is constructed, the historical demand data of the third item is fitted, and the parameters of the model are determined. Finally, the fitted ARIMA model is used to predict the current demand of the third item.

[0034] It can be concluded from the above that this embodiment first determines a suitable d value to make the historical demand data of the third item more stable, and searches and optimizes (p, q) based on the stable data, so that the model can better capture the rules and trends in the data, thereby improving the stability and reliability of the model.

[0035] In one embodiment of the present disclosure, the item recommendation method based on big data analysis further includes: Calculate the average value of the historical demand for the third item in each time period based on the set sliding window; The average values ​​of each time period are clustered and analyzed to obtain the first cluster area and the second cluster area; the average value corresponding to the first cluster area is greater than the average value corresponding to the second cluster area; Sort the average values ​​in the first clustering area by time period, and calculate the first time interval between adjacent time periods; If the relative standard deviation of the first time interval is less than or equal to the second threshold, setting the first limit value to the first value; If the relative standard deviation of the first time interval is greater than the second threshold, the first limit value is set to the second value; the first value is greater than the second value.

[0036] In this embodiment, if the historical demand for the third item has obvious seasonality or periodicity, for example, the sales of certain commodities increase significantly in a specific time period (such as cold drinks in summer and hand warmers in winter), larger p and q values ​​are required to capture this seasonal pattern. In this case, a larger first limit value can be set; otherwise, if the historical demand for the third item has no obvious seasonality or periodicity, a smaller first limit value can be set.

[0037] Specifically, when judging whether the historical demand for the third item has obvious seasonality or periodicity, first set a fixed-size window to slide on the time series data of the historical demand, and calculate the average value of the data in the sliding window. For example, if the size of the sliding window is set to one week, then the average demand for the third item in that week is calculated once a week, which can smooth out short-term fluctuations in the data, highlight the trend characteristics of demand in different periods, and provide a more stable data basis for subsequent analysis.

[0038] Then, a clustering algorithm (such as K-Means algorithm) is used to group the average demand values ​​in each time period, and the larger average demand values ​​are grouped into the first cluster area, and the smaller average demand values ​​are grouped into the second cluster area. The first cluster area contains the average demand values ​​during the peak demand period. By sorting these periods by time, the first time interval between two adjacent periods is calculated. The first time interval reflects the time distribution between the peak demand periods. By analyzing the change in the first time interval, the frequency and periodicity of the peak demand can be obtained.

[0039] The change in the first time interval can be characterized by the relative standard deviation of the first time interval. The smaller the relative standard deviation, the smaller the change in the first time interval. If the relative standard deviation is less than or equal to the second threshold, it indicates that the first time interval changes very little, and the historical demand for the third item has obvious seasonality or periodicity, and the first limit can be set to a larger first value; otherwise, if the relative standard deviation of the first time interval is greater than the second threshold, it indicates that the historical demand for the third item does not have obvious seasonality or periodicity, and the first limit can be set to a smaller second value. Among them, the second threshold is a preset constant, and those skilled in the art can flexibly design the specific value of the second threshold according to actual needs.

[0040] It can be concluded from the above that this embodiment can deeply explore the time distribution characteristics of the peak demand period of the third item through sliding window and cluster analysis, and dynamically adjust the first limit value according to the time distribution characteristics of the peak demand period of the third item, so that the subsequently constructed ARIMA model can better adapt to the characteristics of the demand data of the third item, thereby improving the prediction accuracy and stability of the model.

[0041] In one embodiment of the present disclosure, the item recommendation method based on big data analysis further includes: If the number of samples of the historical demand for the third item is less than or equal to a third threshold, determining a fitness function of the genetic algorithm based on the Akaike information criterion; If the number of samples of the historical demand for the third item is greater than a third threshold, a fitness function of the genetic algorithm is determined based on the Bayesian information criterion.

[0042] In this embodiment, the fitness function of the genetic algorithm can be determined based on the Akaike Information Criterion (AIC) or the Bayesian Information Criterion (BIC). The core idea of ​​AIC is to balance the goodness of fit and complexity of the model to avoid overfitting or underfitting. The specific calculation formula is:

[0043] in, Indicates the number of parameters of the model. In this embodiment, the parameters of the ARIMA model include p coefficients of the autoregressive part, q coefficients of the moving average term, and the mean. Therefore =p+q+1; Represents the maximum likelihood estimate of the model, which is used to characterize how well the model fits the data. The larger the value, the better the model fits the data.

[0044] BIC strengthens the penalty for complexity. The specific calculation formula is:

[0045] in, Indicates the sample size. When the sample size is large, will become very large, making the penalty term The impact on BIC value is more significant.

[0046] Considering that different items have different launch times, for third items that have not been launched for a long time, due to the small sample size (less than or equal to the third threshold), the fitness function of the genetic algorithm can be determined based on AIC. AIC has a relatively light penalty on model complexity and pays more attention to the degree of fit of the model to the existing data. It can try models of different complexities under limited data and find a good balance between fitting data and model simplicity, which helps to mine information in small sample data and improve the adaptability of the model. Among them, the third threshold is a preset constant, and those skilled in the art can flexibly design the specific value of the third threshold according to actual needs.

[0047] For third-party products that have been on the market for a long time, the model has enough data for learning and fitting. At this time, the fitness function of the genetic algorithm can be determined based on BIC. BIC significantly increases the penalty for model complexity, which can more effectively prevent model overfitting, making the model selected under a large amount of data more concise and more generalizable.

[0048] It can be concluded from the above that this embodiment flexibly selects the fitness function according to the number of samples of the historical demand for the third item, and can better adapt to different data scales and characteristics.

[0049] In one embodiment of the present disclosure, determining the difference order based on the historical demand fluctuation value of the third item includes: Starting from the difference order equal to 0, the difference order adjustment operation is performed multiple times until the stop condition is met; The differential order adjustment operation includes: performing a differential operation on the historical demand of the third item based on the differential order, and calculating a fluctuation value of the historical demand of the third item after the differential operation; If the historical demand fluctuation value of the third item after the difference operation is greater than the first threshold, the difference order is accumulated by 1; The stopping condition is: after the difference operation, the historical demand fluctuation value of the third item is less than or equal to the first threshold.

[0050] In this embodiment, by continuously trying different orders of difference, the fluctuation value of the historical demand of the third item after the difference operation can be calculated to determine the minimum difference order d that makes the data stable. For example, if the relative standard deviation of the historical demand of the third item is greater than the first threshold after the first-order difference, it indicates that the data fluctuation is still large, and the difference order d is accumulated by 1, that is, d=d+1; the historical demand of the third item is subjected to a second-order difference operation, and if the relative standard deviation of the historical demand of the third item is less than or equal to the first threshold after the second-order difference, it indicates that the data is relatively stable, and the attempt can be stopped, and d=2 is determined.

[0051] It can be concluded from the above that this embodiment automatically and adaptively determines the appropriate difference order according to the actual fluctuation of the historical demand data of the third item, so that the data reaches a stable state, providing a good data foundation for the subsequent selection of the (p, q) combination and the construction of the ARIMA model.

[0052] In one embodiment of the present disclosure, the item recommendation method based on big data analysis further includes: If the target area belongs to the first type of area, N is set to the third value; If the target area belongs to the second type of area, N is set to a fourth value; Among them, the flow of people in the first type of area is smaller than the flow of people in the second type of area, and the third value is smaller than the fourth value.

[0053] In this embodiment, considering that different areas target target customer groups with different flow rates, the target areas are divided into first-category areas and second-category areas based on the flow rates of people.

[0054] The first type of area can be schools, office buildings and other areas. Considering that the first type of area has a small flow of people and is aimed at a specific target customer group, their needs are relatively concentrated and clear, and there is no need to display too many products. For example, the vending machines in the school library are mainly for students. Students have a relatively fixed demand for drinks and snacks. A smaller N value can meet their needs, and there is no need to set a larger N value to recommend other products.

[0055] The second type of area can be subway stations, railway stations and other areas with large traffic flow. The needs of different customers vary greatly. A larger N value can cover the preferences of more different types of customers and meet the needs of more people. For example, at the vending machines at the railway station, the needs of passengers coming and going from the south and north are different. Setting a larger N value can display more items with high demand and increase the possibility of customers purchasing.

[0056] It can be concluded from the above that this embodiment sets different N values ​​according to differences in regional traffic flow, so that item recommendations are more in line with the actual needs of consumers in different regions.

[0057] In one embodiment of the present disclosure, the item recommendation method based on big data analysis further includes: Determine the fifth item based on the e-commerce platform demand data of the first region; the first region is the region to which the target region belongs, and the fifth item is a new product with the demand ranking in the top M; The fifth item is recommended to users in the target area.

[0058] In this embodiment, the first area is a larger area where the target area is located. For example, if the target area is a school, the first area may be the city where the school is located. By collecting demand data for various commodities on the e-commerce platform in the first area, it is possible to take advantage of the large amount of data, wide coverage, and fast update speed of the e-commerce platform to obtain rich market information and fully understand the purchase of various commodities by consumers in the first area, especially the acceptance of new products and purchase trends.

[0059] Based on the demand data of the e-commerce platform in the first region, we can screen out new products with demand ranking in the top M. The demand situation of new products reflects the emerging demands and consumption trends in the current market. New products with high demand ranking have gained a certain degree of market recognition on the e-commerce platform and have high potential appeal. Therefore, new products with high demand ranking are recommended as the fifth item to users in the target area. This can introduce emerging products on the market into the target area in a timely manner, satisfy consumers' pursuit of new things, and thus improve user experience.

[0060] From the above, it can be concluded that this embodiment determines the new products with the top M demands by analyzing the demand data of the e-commerce platform in the first region, and recommends the new products with the top M demands to users, which helps to optimize the product portfolio. This dynamic product adjustment mechanism can improve the operating efficiency of the vending machine, reduce inventory costs, and achieve optimal allocation of resources.

[0061] Corresponding to the above embodiment, a method for recommending items based on big data analysis is described. Figure 2This is a structural block diagram of an item recommendation device based on big data analysis provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 2 The item recommendation device 20 based on big data analysis includes: a first processing module 21, a second processing module 22, a data classification module 23, a data prediction module 24 and an item recommendation module 25.

[0062] The first processing module 21 is used to obtain multiple item classifications in the target area based on the historical demand data of the target area; The second processing module 22 is used to determine the first item with the top N demand in each item category; The data classification module 23 is used to classify each item, and divide the N first items into second items and third items based on the historical demand fluctuation value of the first items; the second item is a first item whose historical demand fluctuation value is less than or equal to a first threshold, and the third item is a first item whose historical demand fluctuation value is greater than the first threshold; A data prediction module 24, configured to predict a current demand for the third item based on a historical demand for the third item, and select a fourth item from the third items based on the prediction result of the current demand; The item recommendation module 25 is used to recommend the second item and the fourth item to users in the target area.

[0063] In one embodiment of the present disclosure, the data prediction module 24 is specifically used to: Determining the difference order based on the historical demand fluctuation value of the third item; randomly generating a plurality of first combinations of autoregressive orders and moving average orders based on a first limit value; Determine multiple first combinations of autoregressive order and moving average order as initial population, use genetic algorithm to perform multiple iterations to determine the optimal order combination; the optimal order combination is the optimal combination of autoregressive order, moving average order and difference order; Construct ARIMA model based on optimal order combination; Predict the current demand for the third item based on the ARIMA model.

[0064] In one embodiment of the present disclosure, the data prediction module 24 is further configured to: Calculate the average value of the historical demand for the third item in each time period based on the set sliding window; The average values ​​of each time period are clustered and analyzed to obtain the first cluster area and the second cluster area; the average value corresponding to the first cluster area is greater than the average value corresponding to the second cluster area; Sort the average values ​​in the first clustering area by time period, and calculate the first time interval between adjacent time periods; If the relative standard deviation of the first time interval is less than or equal to the second threshold, setting the first limit value to the first value; If the relative standard deviation of the first time interval is greater than the second threshold, the first limit value is set to the second value; the first value is greater than the second value.

[0065] In one embodiment of the present disclosure, the data prediction module 24 is further configured to: If the number of samples of the historical demand for the third item is less than or equal to a third threshold, determining a fitness function of the genetic algorithm based on the Akaike information criterion; If the number of samples of the historical demand for the third item is greater than a third threshold, a fitness function of the genetic algorithm is determined based on the Bayesian information criterion.

[0066] In one embodiment of the present disclosure, the data prediction module 24 is further configured to: Starting from the difference order equal to 0, the difference order adjustment operation is performed multiple times until the stop condition is met; The differential order adjustment operation includes: performing a differential operation on the historical demand of the third item based on the differential order, and calculating a fluctuation value of the historical demand of the third item after the differential operation; If the historical demand fluctuation value of the third item after the difference operation is greater than the first threshold, the difference order is accumulated by 1; The stopping condition is: after the difference operation, the historical demand fluctuation value of the third item is less than or equal to the first threshold.

[0067] In one embodiment of the present disclosure, the second processing module 22 is specifically used for: If the target area belongs to the first type of area, N is set to the third value; If the target area belongs to the second type of area, N is set to a fourth value; Among them, the flow of people in the first type of area is smaller than the flow of people in the second type of area, and the third value is smaller than the fourth value.

[0068] In one embodiment of the present disclosure, the item recommendation module 25 is specifically used to: Determine the fifth item based on the e-commerce platform demand data of the first region; the first region is the region to which the target region belongs, and the fifth item is a new product with the demand ranking in the top M; The fifth item is recommended to users in the target area.

[0069] See also Figure 3 , Figure 3 This is a schematic block diagram of an item recommendation system based on big data analysis provided by an embodiment of the present disclosure. Figure 3The item recommendation system 300 based on big data analysis in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, and the computer programs include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of each module / unit in the above-mentioned device embodiments, such as Figure 2 The functions of modules 21 to 25 are shown.

[0070] It should be understood that in the embodiment of the present disclosure, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0071] The input device 302 may include a touch panel, a fingerprint collection sensor (for collecting the user's fingerprint information and fingerprint direction information), a microphone, etc., and the output device 303 may include a display (LCD, etc.), a speaker, etc.

[0072] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0073] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure can execute the implementation methods described in the first and second embodiments of a method for recommending items based on big data analysis provided in the embodiments of the present disclosure, and can also execute the implementation methods of the item recommendation system based on big data analysis described in the embodiments of the present disclosure, which will not be repeated here.

[0074] In another embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by the processor, all or part of the processes in the above-mentioned embodiment method are implemented, and the computer program can also be completed by instructing the relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0075] The computer-readable storage medium may be an internal storage unit of the item recommendation system based on big data analysis of any of the aforementioned embodiments, such as a hard disk or memory of the item recommendation system based on big data analysis. The computer-readable storage medium may also be an external storage device of the item recommendation system based on big data analysis, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the item recommendation system based on big data analysis. Furthermore, the computer-readable storage medium may also include both an internal storage unit of the item recommendation system based on big data analysis and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the item recommendation system based on big data analysis. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0076] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0077] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the item recommendation system and unit based on big data analysis described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0078] In the several embodiments provided in the present application, it should be understood that the disclosed system and method for recommending items based on big data analysis can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0079] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present disclosure.

[0080] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0081] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. An item recommendation method based on big data analysis, characterized in that: include: Obtain multiple item classifications in the target area based on historical demand data of the target area; Determine the first item with the top N demand in each item category; For each item classification, N first items are divided into second items and third items based on the historical demand fluctuation value of the first items; the second items are first items whose historical demand fluctuation value is less than or equal to a first threshold, and the third items are first items whose historical demand fluctuation value is greater than the first threshold; Predicting a current demand for the third item based on a historical demand for the third item, and selecting a fourth item from the third items based on the predicted result of the current demand; The second item and the fourth item are recommended to users in a target area.

2. The method for recommending items based on big data analysis according to claim 1, characterized in that: The predicting the current demand for the third item based on the historical demand for the third item includes: Determining the difference order based on the historical demand fluctuation value of the third item; randomly generating a plurality of first combinations of autoregressive orders and moving average orders based on a first limit value; Determine multiple first combinations of the autoregressive order and the moving average order as initial populations, perform multiple iterations using a genetic algorithm, and determine an optimal order combination; the optimal order combination is an optimal combination of the autoregressive order, the moving average order, and the difference order; Constructing an ARIMA model based on the optimal order combination; The current demand for the third item is predicted based on the ARIMA model.

3. The method for recommending items based on big data analysis according to claim 2, characterized in that: Also includes: Calculate the average value of the historical demand for the third item in each time period based on the set sliding window; Perform cluster analysis on the average values ​​of each time period to obtain a first cluster area and a second cluster area; the average value corresponding to the first cluster area is greater than the average value corresponding to the second cluster area; Sort the average values ​​in the first clustering area by time period, and calculate the first time interval between adjacent time periods; If the relative standard deviation of the first time interval is less than or equal to a second threshold, setting the first limit value to a first value; If the relative standard deviation of the first time interval is greater than a second threshold, the first limit value is set to a second value; and the first value is greater than the second value.

4. The method for recommending items based on big data analysis according to claim 2, characterized in that: Also includes: If the number of samples of the historical demand for the third item is less than or equal to a third threshold, determining a fitness function of the genetic algorithm based on the Akaike Information Criterion; If the number of samples of the historical demand for the third item is greater than a third threshold, the fitness function of the genetic algorithm is determined based on the Bayesian information criterion.

5. The method for recommending items based on big data analysis according to claim 2, characterized in that: The difference order is determined based on the historical demand fluctuation value of the third item, including: Starting from the difference order equal to 0, the difference order adjustment operation is performed multiple times until the stop condition is met; The differential order adjustment operation includes: performing a differential operation on the historical demand of the third item based on the differential order, and calculating a historical demand fluctuation value of the third item after the differential operation; If the historical demand fluctuation value of the third item after the difference operation is greater than the first threshold, the difference order is accumulated by 1; The stopping condition is: the historical demand fluctuation value of the third item after the difference operation is less than or equal to the first threshold.

6. The method for recommending items based on big data analysis according to claim 1, characterized in that: Also includes: If the target area belongs to the first type of area, N is set to a third value; If the target area belongs to the second type of area, N is set to a fourth value; The flow of people in the first type of area is smaller than the flow of people in the second type of area, and the third value is smaller than the fourth value.

7. The method for recommending items based on big data analysis according to claim 1, characterized in that: Also includes: Determine the fifth item based on the e-commerce platform demand data of the first region; the first region is the region to which the target region belongs, and the fifth item is a new product with the demand ranking in the top M; The fifth item is recommended to users in the target area.

8. An item recommendation device based on big data analysis, characterized in that: include: A first processing module, configured to obtain a plurality of item classifications in a target area based on historical demand data of the target area; The second processing module is used to determine the first item with the top N demand in each item category; a data classification module, configured to classify each item, and divide N first items into second items and third items based on the historical demand fluctuation value of the first items; the second items are first items whose historical demand fluctuation value is less than or equal to a first threshold, and the third items are first items whose historical demand fluctuation value is greater than the first threshold; A data prediction module, used for predicting the current demand of the third item based on the historical demand of the third item, and selecting the fourth item from the third items based on the prediction result of the current demand; The item recommendation module is used to recommend the second item and the fourth item to users in a target area.

9. An item recommendation system based on big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images