Household industry intelligent drainage service method and system based on big data

Abstract

The invention provides a household industry intelligent drainage service method and system based on big data, and the method comprises the steps: carrying out the target element extraction of a household product market statistical data set collected by a target analysis group, and obtaining a product conversion distribution state; according to the product conversion distribution state, determining product market data of which the conversion rate meets a set requirement, obtaining a target product market data set, and based on the conversion rate of the product market data in the target product market data set, determining a conversion characterization value of the household product market statistical data set; comparing the conversion representation numerical value of the household product market statistical data set with a market tendency reference value to obtain a market tendency analysis result of the household product market statistical data set; the market tendency reference value is determined through the conversion characterization numerical value of the market statistical data set under different product market tendencies. The method can improve the recognition precision, facilitates the determination of the market tendency, formulates an accurate drainage strategy, and improves the product conversion rate.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and system for intelligent traffic diversion services in the home furnishing industry based on big data. Background Art

[0002] In the marketing promotion of the home furnishing industry, there are many technical problems that need to be solved urgently. On the one hand, it is difficult for traditional drainage methods to accurately determine market trends. Enterprises lack reliable methods to analyze the status of the home furnishing product market, and it is difficult to know whether the market is positive (such as consumers are interested in products and purchase conversion rates are high) or negative (such as products are not popular and conversion rates are low). Due to the lack of effective means to accurately judge market trends, the drainage strategies formulated by enterprises are often blind, and resources cannot be reasonably allocated according to the actual status of the market, resulting in a waste of marketing resources and an inability to efficiently promote products to target customers. On the other hand, the existing drainage technology is not accurate enough in identifying market trends. Due to the lack of accurate analysis methods, when faced with complex home furnishing product market statistics, enterprises cannot accurately distinguish the data characteristics under different product market trends, which makes it difficult for enterprises to formulate drainage strategies in a targeted manner, which is not conducive to improving product conversion rates and difficult to gain advantages in fierce market competition. In addition, traditional drainage methods lack reliable decision-making methods based on data. Without an effective method for analysis based on a sample set of market statistical data, companies cannot know exactly how to direct and push data to the target analysis group based on the actual market situation, making it difficult to attract potential customers in a targeted manner, thus affecting the marketing effectiveness of home furnishing products. Summary of the invention

[0003] In view of this, the present application provides a method and system for intelligent traffic diversion services in the home furnishing industry based on big data. The technical solution of the present application is implemented as follows: On the one hand, the present application provides a smart drainage service method for the home furnishing industry based on big data, the method comprising: extracting target elements from a set of home furnishing product market statistical data collected by a target analysis group to obtain a product conversion distribution state of product market data in the home furnishing product market statistical data set; based on the product conversion distribution state, determining product market data whose conversion rate reaches a set requirement to obtain a target product market data set; based on the conversion rate of the product market data in the target product market data set, determining a conversion characterization value of the home furnishing product market statistical data set; comparing the conversion characterization value of the home furnishing product market statistical data set with a market tendency reference value to obtain a market tendency analysis result of the home furnishing product market statistical data set; wherein the market tendency reference value is determined by the conversion characterization value of the market statistical data set under different product market tendencies; based on the market tendency analysis result, performing data drainage and pushing to the target analysis group; On the other hand, the present application provides a computer system, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the steps in the above method when executing the program.

[0004] The beneficial effects of the present application include: the present application extracts target elements from the household product market statistical data set collected by the target analysis group, and obtains the product conversion distribution status in the household product market statistical data set. Based on the above product conversion distribution status, product market data whose conversion rate reaches the set requirements is obtained, and the target product market data set is obtained. Then, the conversion rate of the product market data in the target product market data set is quantified to obtain the conversion characterization value corresponding to the household product market statistical data set. The present application determines the market tendency analysis result for the household product market statistical data set based on the method of comparing the conversion characterization value of the household product market statistical data set with the market tendency reference value, so as to push data drainage to the target analysis group based on the market tendency analysis result. Because the market tendency reference value is determined based on the conversion characterization value corresponding to the market statistical data set sample, the reliability of the above comparison method is high. In addition, the above market tendency reference value is determined based on the conversion characterization value corresponding to the market statistical data set sample under multiple different product market tendencies, so the above comparison process can improve the recognition accuracy, which is conducive to determining the market tendency, formulating an accurate drainage strategy, and improving the product conversion rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 A schematic diagram of the implementation process of a smart traffic diversion service method for the home furnishing industry based on big data provided in an embodiment of the present application; Figure 2A hardware entity diagram of a computer system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0006] The embodiment of the present application provides a smart traffic diversion service method for the home furnishing industry based on big data, which can be executed by a processor of a computer system. The computer system can refer to a device with data processing capabilities such as a server, a laptop, a tablet computer, a desktop computer, a mobile device, etc. Figure 1 A schematic diagram of the implementation process of a smart traffic diversion service method for the home furnishing industry based on big data provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method comprises the following steps: Step S100: extracting target elements from the household product market statistical data set collected by the target analysis group to obtain product conversion distribution status of product market data in the household product market statistical data set.

[0007] The household products market statistical data set is an integration of household products market-related data. These data cover various types of information on household products in the market, such as sales volume and sales revenue of different household products, relevant data of different sales channels, sales data in different time periods, consumer evaluation data on different household products, and many other data related to the household products market situation.

[0008] The target analysis group is a predetermined group of people with specific characteristics or in a specific range in the entire home furnishing product market. For example, groups divided by age, such as young consumers aged 20-30; or groups divided by region, such as consumers in a certain city or region; or groups divided by spending power, such as mid- to high-end consumer groups, etc. The computer system collects relevant home furnishing product market statistics for these specific target analysis groups to form a collection of home furnishing product market statistics.

[0009] Target element extraction is the process of extracting elements related to product conversion from a set of household product market statistics by a computer system. Product conversion here refers to situations related to further operations of household products by consumers in the market, including but not limited to clicks (consumers click on the network platform to view detailed information of household products), favorites (adding household products of interest to favorites for subsequent viewing), purchases (consumers actually purchase household products), and sharing (consumers share household product information with others, such as sharing product links through social media).

[0010] The data mining techniques that can be used by computer systems include association rule mining algorithms. For example, the Apriori algorithm finds the association between data by mining frequent item sets. For another example, the decision tree algorithm can also be used for target element extraction. The decision tree is an algorithm for making decisions based on a tree structure. Each internal node is a test on an attribute, the branch is the test output, and the leaf node is a category or value. The computer system can use the different attributes of home products (such as price, brand, function, etc.) and the attributes of consumers (such as age, gender, consumption ability, etc.) as nodes of the decision tree, and build a decision tree model by learning the known product conversion situation (such as whether to buy), and then use this model to extract factors related to product conversion. For example, the decision tree model can conclude that home products with a price range of a certain brand, a specific brand, and functions that meet certain needs are more likely to be clicked and collected, thereby extracting these factors. Through the above target element extraction process, the product conversion distribution status of the product market data in the home product market statistical data set is finally obtained. This product conversion distribution status is a presentation of various conversion operations of home products in the target analysis group. For example, the following results may be obtained: in the target analysis group, clicks on a certain type of home products account for 30% of the total views, collections account for 10% of the total views, purchases account for 5% of the total views, shares account for 2% of the total views, etc.

[0011] Step S200: Based on the product conversion distribution state, determine the product market data whose conversion rate reaches the set requirement, and obtain the target product market data set.

[0012] The conversion rate is the ratio between the number of actual conversions and the total number of possible conversions in a specific conversion operation. For example, for the purchase conversion rate, if there are 1,000 views of home products and 50 purchases occur, then the purchase conversion rate is 50 / 1,000 = 0.05 or 5%. For different conversion operations (click, favorite, purchase, share, etc.), the corresponding conversion rate can be calculated in this way. The setting requirement is a standard value or standard range determined in advance by relevant business personnel or based on market strategies. This setting requirement can be determined based on a variety of factors, such as the company's profit target, market competition, industry average level, etc. The computer system can use database query technology for screening. If the home product market statistics are stored in a database, such as a relational database (such as MySQL, etc.), the computer system can write SQL query statements to achieve screening. Another technical means can be based on conditional judgment statements in the program. If the data is a data structure stored in memory (such as an array, list, etc.), the computer system can use a programming language (such as Python) to screen.

[0013] Step S300: determining a conversion representation value of the household product market statistical data set based on the conversion rate of the product market data in the target product market data set.

[0014] In step S300, the computer system determines a conversion representation value of the household product market statistical data set based on the conversion rate of the product market data in the target product market data set.

[0015] The conversion rate of product market data refers to the conversion ratio of each product in the target product market data set in different conversion operations (such as clicks, favorites, purchases, shares, etc.). Taking the purchase conversion rate as an example, it reflects the proportion of actual purchase behavior under a certain number of views or potential conversion opportunities. Assuming that a certain type of furniture product in the target product market data set has 1,000 views in a certain period of time, and the actual purchase volume is 100 times, then the purchase conversion rate of the furniture product is 100 / 1000 = 0.1 or 10%. The computer system determines the conversion characterization value of the home product market statistical data set in order to use a value to characterize the conversion characteristics of the home product market statistical data set as a whole. This conversion characterization value is calculated based on the conversion rate of the product market data in the target product market data set. One technical means is the weighted average method, which assigns different weights to different conversion operations (clicks, favorites, purchases, shares, etc.), and then calculates the conversion characterization value based on the conversion rate of each product in the target product market data set and the corresponding weight. Another technical means can be a comprehensive calculation based on data standardization. First, the computer system normalizes the different conversion rates of each product. For example, for click-through conversion rate, the click-through conversion rates of all products can be sorted and then normalized according to their positions in the sorting, so that all normalized conversion rate values ​​are between 0 and 1.

[0016] The determination of this conversion representation value provides a unified value that can comprehensively reflect the conversion characteristics of the household products market statistical data set for subsequent steps. This value can help companies or analysts more intuitively understand the overall situation of the entire household products market in terms of conversion, rather than just focusing on the conversion rate of a single product or a single conversion operation. For example, when a company wants to evaluate the household products market situation in different time periods or different regions, this conversion representation value can be used as an important comparison indicator.

[0017] Step S400: Compare the conversion representation value of the household product market statistical data set with the market tendency reference value to obtain the market tendency analysis result of the household product market statistical data set; wherein the market tendency reference value is determined by the conversion representation value of the market statistical data set under different product market tendencies.

[0018] The market tendency reference value is a predetermined standard value, which is determined by the conversion representation value of the market statistical data set under different product market tendencies. The product market tendency here is a description of the overall trend or status of the home product market, for example, it can be divided into positive market tendencies and negative market tendencies. Under different market tendencies, the conversion of home products will be different, and this difference is reflected in the conversion representation value.

[0019] In order to determine the reference value of market tendency, the computer system may adopt a variety of methods. For example, a large amount of historical data is collected, which covers a set of household product market statistical data in different periods and different market conditions. Then, according to the method of determining the conversion characterization value previously, the conversion characterization value corresponding to each historical data is calculated. Then, according to the distribution of these conversion characterization values ​​and the pre-classification of market tendencies (for example, positive market tendencies are determined based on market sales growth, positive consumer feedback, etc.; negative market tendencies are determined based on sales decline, low consumer interest, etc.), reference values ​​for distinguishing different market tendencies are determined.

[0020] Assume that after analyzing a large amount of historical data, the computer system determines that the conversion characterization values ​​under positive market trends are generally higher, and the conversion characterization values ​​under negative market trends are generally lower. For example, the conversion characterization values ​​under positive market trends range from 0.15 to 0.25, and the conversion characterization values ​​under negative market trends range from 0.05 to 0.1. Then, the computer system may determine a market trend reference value of 0.13. The determination of this reference value requires comprehensive consideration of factors such as data distribution, market trend characteristics, and actual business needs.

[0021] When the computer system compares the conversion representation value of the household product market statistical data set with the market tendency reference value, it is actually judging which predefined market tendency the market state represented by the current household product market statistical data set is more inclined to. For example, when the conversion representation value of the household product market statistical data set is 0.12 and the market tendency reference value is 0.13, since 0.12 is less than 0.13, the computer system preliminarily judges that the household product market statistical data set is more inclined to a negative market tendency. The result of this comparison, that is, the market tendency analysis result, can help enterprises understand the overall situation of the household product market and formulate corresponding market strategies. If the market tendency analysis result is a positive market tendency, the enterprise can consider further expanding production, increasing marketing investment, launching new household product series and other strategies to make full use of the positive market situation. For example, the enterprise can increase investment in advertising to improve brand awareness and attract more consumers to buy household products. On the contrary, if the market tendency analysis result is a negative market tendency, the enterprise may need to take measures such as adjusting product prices, optimizing product functions, and improving after-sales service to improve the market situation. For example, the enterprise can reduce the price of some unsalable household products for promotion, or improve the design of the product to improve the competitiveness of the product. At the same time, companies can also enhance consumers' confidence in home furnishing products by improving the quality of after-sales service, such as extending the warranty period and increasing after-sales service outlets, thereby increasing product conversion rates and prompting market trends to shift in a positive direction.

[0022] Step S500: Pushing data to the target analysis group based on the market trend analysis result.

[0023] The target analysis group is a specific group of people determined before step S100, and they have certain characteristics related to the home furnishing product market, such as a specific age range, geographical range, or consumption capacity range. For example, when the market tendency analysis result is a positive market tendency, it means that the conversion of home furnishing products in the market is good, and consumers show a high interest and willingness to buy home furnishing products. In this case, the computer system will adopt an active drainage strategy to further expand market share and increase product awareness and sales. For example, suppose the target analysis group is young consumers aged between 25 and 35, living in first-tier cities, and with mid-to-high-end consumption capabilities. The computer system can use the following technical means to perform data drainage and push.

[0024] One way is to deliver targeted advertisements. The computer system can cooperate with the advertising platform and use big data analysis to accurately push advertisements for home products to the target analysis group. For the target analysis group with positive market tendencies, the advertising content can emphasize the innovative features, high quality and fit with fashion trends of the product. At the same time, the computer system can use personalized recommendation algorithms to push data. Recommend relevant home products to the target analysis group based on their previous consumption records, browsing history and interest preferences. The personalized recommendation algorithm can be based on the collaborative filtering algorithm, the basic principle of which is to make recommendations by analyzing the behavioral similarity between users or the similarity between items. Assuming that user A and user B have a high similarity in purchasing home products (such as both have purchased high-end furniture and kitchen supplies), when user A purchases a new home product (such as a smart curtain), the computer system can recommend this smart curtain to user B. From the perspective of mathematical formulas, for the user-item rating matrix R, the cosine similarity formula can be used to calculate the similarity sim(u, v) between users: ; where u and v are two users, I is a collection of items, and are the ratings of user u and user v for item i respectively (in the absence of a clear rating, a quantitative rating can be made based on the user's purchase, click and other behaviors, such as 1 for purchase and 0.5 for click).

[0025] When the market tendency analysis result is a negative market tendency, it indicates that the conversion of home furnishing products in the market is not ideal, and consumers' interest and willingness to buy are low. The computer system needs to take corresponding measures to improve this situation and conduct targeted data diversion and push. By conducting data diversion and push to the target analysis group based on the market tendency analysis results, the computer system can more accurately meet market demand and improve the promotion effect and market competitiveness of home furnishing products.

[0026] As an embodiment, the method further includes: Step S101: obtaining X market statistical data set samples under market tendency modes, where X≥2; Step S102: determining a uth conversion representation value corresponding to the uth market tendency mode based on a sample set of market statistical data under the uth market tendency mode, wherein 1≤u<X; Step S103: Determine the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value; wherein the uth market tendency reference value is used to compare with the conversion characterization value corresponding to the household product market statistical data set to determine whether the household product market statistical data set belongs to the u+1th market tendency pattern or the uth market tendency pattern.

[0027] In step S101, the market trend model is a classification description of the overall situation of the home furnishing product market, which reflects the characteristics of the market in different states, such as the market activity, consumers' purchasing intention, product sales trends, etc. Under different market trend models, the market performance of home furnishing products will be significantly different, and these differences are reflected in various market statistics.

[0028] The market statistical data set sample is a data set collected from the actual home furnishing product market, which is used to represent the market situation under a specific market trend model. These data sets contain rich information, such as the sales volume, sales price, sales ratio of different channels, geographical distribution and age distribution of consumers, and data related to product conversion (such as clicks, favorites, purchases, shares, etc.).

[0029] Assume that X=3, that is, the computer system needs to obtain market statistical data set samples under three market tendency modes. The three market tendency modes can be a growth market tendency mode, a stable market tendency mode, and a decline market tendency mode.

[0030] For the sample of market statistical data set under the growth market tendency model, the computer system can collect data from some emerging home furnishing markets or home furnishing product categories that are in a rapid development stage. For example, the smart home product market has developed rapidly in recent years, and the computer system can obtain relevant data from multiple smart home product suppliers, e-commerce platforms, and offline stores. These data may include the growth trend of sales of various smart home devices (such as smart door locks, smart cameras, smart lamps, etc.) in the past few months or years, the changes in the sales proportion of products in different price ranges, and the age and geographical distribution of consumers who purchase these products. For example, the data shows that in a certain city, the purchase of smart door locks by young consumers aged 25-35 has increased by 50% in the past year, and is mainly concentrated in the mid-to-high-end price range (1,000-3,000 yuan), and the sales proportion of online channels accounts for 70% of the total sales.

[0031] For the sample set of market statistics data under the stable market tendency model, the computer system can select some traditional household products with high market saturation, such as ordinary furniture (such as sofas, dining tables, etc.). The computer system collects data from furniture manufacturers, dealers, and furniture stores, and finds that the sales volume of these products remains relatively stable for a long time, the price fluctuation is small, and the consumer's purchasing behavior is relatively regular. For example, the annual sales volume of a certain brand of sofa fluctuates between 1,000 and 1,200 sets, the average price is between 5,000 and 6,000 yuan, the age distribution of buyers is relatively wide, but mainly concentrated between 30 and 50 years old, and offline sales channels still dominate, accounting for about 80% of total sales. For the sample set of market statistics data under the recessionary market tendency model, the computer system can focus on household products that are gradually being eliminated by the market or whose demand has dropped significantly, such as old incandescent light bulbs (with the popularization of energy-saving lamps). Data is collected from some small hardware stores, grocery stores near old communities, and related sales records, and it is found that the sales volume of this type of incandescent light bulbs has continued to decline in the past few years. Although the price has decreased due to the reduction in market demand, the decline in sales volume is more obvious. For example, the sales volume of incandescent bulbs in a certain region has decreased by 30% each year in the past three years. The buyers are mainly elderly consumers who are sensitive to prices and not very concerned about energy saving. They are mainly concentrated in some remote rural areas or old urban areas. Offline sales channels are still the main sales method, but the sales share has dropped from 80% to 60%, mainly due to the impact of online sales channels and the shrinking market demand.

[0032] The technical means by which a computer system obtains these market statistics collection samples may include data collection software, web crawler technology, and data connection with the information system of an enterprise or merchant. Data collection software can directly obtain structured data from the sales management system, inventory management system, etc. within the enterprise, for example, by connecting to the database through the ODBC (OpenDatabase Connectivity) or JDBC (Java Database Connectivity) interface, and then executing SQL query statements to obtain the required data. Web crawler technology is mainly used to obtain public, unstructured data from e-commerce platforms, home product information websites, etc.

[0033] In step S102, the market statistical data set sample under the u-th market trend mode includes various data information of the home furnishing product market under the mode, such as the sales data, consumer data and product conversion data mentioned above. The conversion representation value is a value that comprehensively reflects the product conversion situation under the market trend mode, which is obtained by performing specific calculations on the relevant data in the market statistical data set sample.

[0034] Continuing with the above example, assuming u=1, that is, for the growth market tendency mode, the computer system needs to determine the conversion representation value corresponding to this mode.

[0035] First, the computer system extracts data related to product conversion from a sample set of market statistics data under the growth market tendency model. Then, the computer system can use the weighted average method to calculate the conversion representation value. Different weights are assigned to different conversion operations. For example, the weight of clicks is 0.2, the weight of favorites is 0.15, the weight of purchases is 0.4, and the weight of shares is 0.25.

[0036] For smart door locks: .

[0037] For smart cameras: .

[0038] If there are only these two products in this sample set of market statistics, then the average conversion performance value can be calculated (assuming that both products are equally important): This 2305 is the first conversion representation value corresponding to the growth market tendency model (u=1).

[0039] In addition to the weighted average method, the computer system can also use other technical means, such as principal component analysis (PCA), to transform multiple related variables into a few unrelated comprehensive variables (principal components). The computer system can treat variables such as clicks, favorites, purchases, and shares as original variables, and use the PCA method to obtain a comprehensive principal component score as a conversion representation value.

[0040] In step S103, the uth conversion characterization value and the u+1th conversion characterization value respectively represent the comprehensive conversion conditions under two adjacent market tendency modes. The market tendency reference value is a critical value that distinguishes two adjacent market tendency modes.

[0041] For example, continuing with the previous example, when u=1 (growth market tendency model), the first conversion representation value is 2305; assuming u+1=2 (stable market tendency model), the computer system calculates the second conversion representation value corresponding to the stable market tendency model as 1500 according to the method of step S102.

[0042] The computer system can use the interval division method to determine the first market tendency reference value. One feasible way is to take the middle value of the two conversion characterization values, that is: (2305+1500) / 2=1902.5. This 1902.5 is the first market tendency reference value. When the conversion characterization value corresponding to the household product market statistical data set is greater than 1902.5, the computer system can judge that the household product market is more inclined to the growth market tendency mode; when this value is less than 1902.5, it is more inclined to the stable market tendency mode. Another method is to consider the distribution characteristics of the data to determine the market tendency reference value. If a large amount of historical data is analyzed and it is found that the distribution of conversion characterization values ​​under the growth market tendency mode is relatively discrete, while the distribution of conversion characterization values ​​under the stable market tendency mode is relatively concentrated, the computer system can use the value that is more inclined to the stable market tendency mode as the market tendency reference value. For example, the first market tendency reference value can be determined as 1800.

[0043] The technical means by which the computer system determines the market tendency reference value can also be based on the cluster analysis method. Cluster analysis is the process of dividing data objects into different classes or clusters according to similarity. The computer system can regard the conversion representation values ​​under different market tendency modes as data objects, and divide them into different clusters through clustering algorithms (such as K-Means clustering algorithm). The market tendency reference value is then determined based on the boundary or center position of the cluster. For example, the K-Means clustering algorithm divides the conversion representation values ​​under the growth and stable market tendency modes into two clusters, and determines the market tendency reference value by calculating the center distance of the two clusters and according to a certain ratio (such as taking the value at 60% of the distance).

[0044] As an implementation manner, the u-th market tendency pattern includes Y sub-market tendency patterns; the market statistical data set sample under the u-th market tendency pattern includes Y sub-market statistical data set sample corresponding to the Y sub-market tendency patterns, where Y≥2. Based on this, the step S102, based on the market statistical data set sample under the u-th market tendency pattern, determines the u-th conversion representation value corresponding to the u-th market tendency pattern, which may include: Step S1021: determining the sub-conversion representation values ​​corresponding to the sub-market statistical data set samples under the u-th market tendency mode, and obtaining Y sub-conversion representation values; Step S1022: determining the vth influence parameters corresponding to the Y sub-market tendency patterns in the uth market tendency pattern, where 1≤v<Y; Step S1023: performing a weight adjustment operation on the sub-conversion representation value corresponding to the vth sub-market tendency mode according to the vth influence parameter; Step S1024: determining the conversion representation value corresponding to the u-th market tendency pattern based on the sub-conversion representation value after the weight adjustment operation.

[0045] In step S1021, the market statistical data set sample under the u-th market tendency mode contains a variety of information and may be further subdivided. The sub-market statistical data set sample here is the result of subdividing the market statistical data set sample under the u-th market tendency mode according to a certain rule. For example, in the aforementioned growth market tendency mode, if the home products are subdivided according to their functions, sub-market statistical data set samples such as smart home security products (such as smart door locks, smart cameras, etc.) and smart life home products (such as smart lamps, smart sockets, etc.) can be obtained. The sub-conversion characterization value is a value that reflects the product conversion status in each sub-market statistical data set sample. In order to obtain these sub-conversion characterization values, the computer system needs to extract data related to product conversion from each sub-market statistical data set sample, and then use a suitable calculation method.

[0046] Take the example of the sub-market statistical data set of smart home security products, assuming that it contains data on two types of products: smart door locks and smart cameras. For smart door locks, assume that in a certain period of time, the number of clicks on a certain e-commerce platform is 10,000, the number of favorites is 2,000, the number of purchases is 500, and the number of shares is 300; for smart cameras, the number of clicks is 8,000, the number of favorites is 1,500, the number of purchases is 400, and the number of shares is 200. The computer system can use the weighted average method to calculate the sub-conversion representation value, assuming that the weight of clicks is 0.2, the weight of favorites is 0.15, the weight of purchases is 0.4, and the weight of shares is 0.25.

[0047] For smart door locks: .

[0048] For smart cameras: .

[0049] In this way, the sub-conversion representation values ​​corresponding to the two products in the sample of the sub-market statistical data set of smart home security products are obtained, which are 2575 and 2035. In the same way, if there are other sub-market statistical data set samples, such as smart home products, the computer system can also calculate their corresponding sub-conversion representation values, and finally obtain Y sub-conversion representation values.

[0050] When calculating the sub-conversion representation value, the computer system can use other technical means in addition to the weighted average method. For example, a method based on regression analysis. If the computer system finds that variables such as the number of clicks, favorites, purchases and shares of a product have a linear relationship with a potential factor (such as marketing strength, product launch time, etc.), then the sub-conversion representation value can be calculated by establishing a regression equation. Assuming that there is a linear relationship between the product conversion volume (Y) and the marketing strength (X1) and the product launch time (X2), the regression equation is Y=aX1+bX2+c (where a, b, and c are coefficients obtained by fitting historical data). The computer system can calculate Y based on the values ​​of X1 and X2 for each product, and use it as a way to calculate the sub-conversion representation value.

[0051] In step S1022, the vth impact parameter is a value used to measure the degree of influence of each sub-market trend pattern on the overall market trend pattern. The determination of this value needs to consider multiple factors, such as the size of the sub-market (which can be measured by indicators such as sales volume, sales volume or market share of the sub-market), the growth potential of the sub-market (derived from historical data and market forecast analysis), the correlation between the sub-market and other related markets (such as the correlation between smart home security products and smart home control systems, the higher the correlation, the greater the impact), etc.

[0052] The technical means by which the computer system determines these influencing parameters can be based on a combination of data analysis and expert experience. From the perspective of data analysis, the computer system can perform statistical analysis on a large amount of historical market data, such as calculating the sales volume of sub-markets, the trend of sales volume changes in different time periods, and the correlation coefficient with other markets. At the same time, the computer system can also refer to the opinions of industry experts, who adjust and determine the influencing parameters based on their in-depth understanding of the market and their judgment of industry development trends. For example, experts may revise the influencing parameters initially determined by the computer system based on new technological development trends (such as a new technology that can greatly enhance the competitiveness of smart health home products, thereby changing their influencing parameters).

[0053] In step S1023, the purpose of the weight adjustment operation is to adjust the corresponding sub-conversion representation value according to the degree of influence of each sub-market tendency pattern on the overall market tendency pattern, so as to more accurately reflect the product conversion situation under the overall market tendency pattern.

[0054] Continuing with the above example, in step S1021, the sub-conversion characterization value of the smart home security product (assuming it is 2575) is obtained, and its corresponding first influence parameter is 0.4; the sub-conversion characterization value of the smart life home product (assuming it is 2035), and its corresponding second influence parameter is 0.3; the sub-conversion characterization value of the smart health home product (assuming it is 1800), and its corresponding third influence parameter is 0.2. The computer system performs a weight adjustment operation on the sub-conversion characterization value of the smart home security product: 2575×0.4=1030; the sub-conversion characterization value of the smart life home product is weighted: 2035×0.3=610.5; the sub-conversion characterization value of the smart health home product is weighted: 1800×0.2=360. Through such a weight adjustment operation, the sub-conversion characterization values ​​corresponding to different sub-market tendency patterns are readjusted according to their influence on the overall market tendency pattern, so that these values ​​can more reasonably reflect their respective contributions when the overall conversion characterization value is calculated later.

[0055] In step S1024, after the weight adjustment operation of step S1023, a new set of values ​​is obtained, which comprehensively consider the degree of influence of each sub-market tendency mode on the overall market tendency mode and the respective product conversion conditions. The computer system obtains the conversion characterization value corresponding to the u-th market tendency mode by summing up and calculating the sub-conversion characterization values ​​after these weight adjustment operations. Continuing the above example, the values ​​obtained after the weight adjustment operation are 1030, 610.5 and 360 respectively. The computer system can determine the conversion characterization value corresponding to the u-th market tendency mode by summing: 1030+610.5+360=2000.5, that is, the conversion characterization value corresponding to the u-th market tendency mode (such as the growth market tendency mode). In addition to summing, other calculation methods can also be used. For example, if based on the complexity of the market or the accuracy requirements of the data, the computer system can perform weighted average calculation on the sub-conversion characterization values ​​after these weight adjustment operations.

[0056] As an implementation manner, the step S103, determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value, may include: Step S1031: determining a u-th numerical range, wherein the maximum value of the u-th numerical range is the conversion representation value corresponding to the u+1-th market tendency mode, and the minimum value of the u-th numerical range is the conversion representation value corresponding to the u-th market tendency mode; Step S1032: Determine a uth market tendency reference value based on the uth numerical range, wherein the uth market tendency reference value belongs to the uth numerical range.

[0057] In step S1031, the u-th market tendency mode and the u+1-th market tendency mode represent different market states respectively, and the conversion characterization values ​​corresponding to them are values ​​that comprehensively reflect the product conversion conditions under the respective market tendency modes. The computer system determines this numerical range in order to further find a suitable u-th market tendency reference value, which will help distinguish the two market tendency modes. For example, assuming that the u-th market tendency mode is a stable market tendency mode, and the u+1-th market tendency mode is a growing market tendency mode. The computer system has determined through the previous steps (such as step S102 and its sub-steps) that the conversion characterization value corresponding to the stable market tendency mode is 1500, and the conversion characterization value corresponding to the growing market tendency mode is 2305. Then, the u-th numerical range is the interval from 1500 (inclusive) to 2305 (inclusive). This numerical range covers all possible situations of the conversion characterization values ​​under the two market tendency modes. The values ​​within this range may represent the intermediate state of transition from the stable market tendency mode to the growing market tendency mode. The technical means by which the computer system determines this numerical range can be based on simple data comparison and storage operations. In the database or data structure of the computer system, the conversion representation values ​​corresponding to the u-th market tendency pattern and the u+1-th market tendency pattern have been stored. The computer system queries these stored values ​​and determines the value range according to the set rules.

[0058] In step S1032, the uth market tendency reference value is a critical value, which is used to distinguish the uth market tendency mode from the u+1th market tendency mode. When determining this reference value, the computer system needs to comprehensively consider the uth value range and the actual market situation.

[0059] One feasible determination method is to take the middle value of the u-th numerical range. For example, in the case where the u-th numerical range mentioned above is [1500, 2305], the u-th market tendency reference value can be (1500+230) / 2=1902.5. This method of taking the middle value is simple and intuitive. Assuming that in this market, the conversion representation values ​​of the two market tendency patterns are relatively evenly distributed, then the middle value is a more reasonable reference value. When the conversion representation value corresponding to the household product market statistical data set is greater than 1902.5, the computer system can determine that the household product market is more inclined to a growth market tendency pattern; when this value is less than 1902.5, it is more inclined to a stable market tendency pattern.

[0060] Market conditions are often complex, and the distribution of conversion characterization values ​​may not be uniform. For example, through the analysis of a large amount of historical data, the computer system found that the conversion characterization values ​​under the stable market tendency mode are mostly concentrated near the lower limit of the value range, while the conversion characterization values ​​under the growth market tendency mode are mostly concentrated near the upper limit of the value range. In this case, the computer system can use a value that is more inclined to the stable market tendency mode as the u-th market tendency reference value. For example, the computer system can determine the u-th market tendency reference value as 1800. In addition to considering the imbalance of value distribution, the computer system can also determine the u-th market tendency reference value in combination with other market characteristics. For example, if in the stable market tendency mode, the market competition is fierce and the product homogeneity is serious, resulting in a small fluctuation of the conversion characterization value and a tendency to a lower level; while in the growth market tendency mode, the market innovation is active and the product differentiation is obvious, the conversion characterization value fluctuates greatly and tends to a higher level. The computer system can adjust the u-th market tendency reference value based on these market characteristics and combined with expert experience or market research results.

[0061] As an implementation manner, in the step S103, after determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value, the method may further include: Step S103a: obtaining a u-th market tendency mode change upper critical value corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode; Step S103b: Obtain the uth market tendency mode change lower critical value corresponding to the change from the u+1th market tendency mode to the uth market tendency mode; wherein the uth market tendency mode change upper critical value is greater than the uth market tendency mode change lower critical value, and the uth market tendency reference value is greater than the uth market tendency mode change upper critical value.

[0062] In step S103a, the computer system obtains the uth market tendency mode change upper critical value corresponding to the change from the uth market tendency mode to the u+1th market tendency mode. This critical value is used to determine under what circumstances the conversion representation value of the household product market statistical data set can be determined as a change from the uth market tendency mode to the u+1th market tendency mode.

[0063] The transformation of the market tendency mode is a complex process. The existence of the u-th market tendency mode change high critical value is to avoid misjudgment due to slight fluctuations in data. For example, the u-th market tendency mode is a stable market tendency mode, and the u+1-th market tendency mode is a growing market tendency mode. Assume that the market tendency reference value determined in step S103 is 1902.5. However, it may be too hasty to judge that the market has changed from a stable type to a growing type just because the conversion representation value is slightly higher than 1902.5 (such as 1905), because this may only be a short-term, accidental data fluctuation.

[0064] The technical means by which the computer system obtains this high critical value of change can be based on a large amount of historical data and statistical analysis methods. For example, the computer system has collected household product market statistics from many years and regions in the past. These data cover various situations in the process of transformation from a stable market tendency model to a growth market tendency model. The computer system can group these data, group those market data that have successfully transformed from a stable model to a growth model, and then conduct in-depth analysis of the transformation representation values ​​in this group of data.

[0065] One analysis method is to calculate the mean and standard deviation of these transformation characterization values. Suppose that after analysis, the mean of this set of data is 2000 and the standard deviation is 50. In this case, the computer system can determine the upper critical value of the u-th market tendency pattern change according to certain rules. A feasible rule is to add a certain multiple (such as 1) of the standard deviation to the mean as the upper critical value of the change. Then, in this example, the upper critical value of the u-th market tendency pattern change is 2000+50=2050. This means that when the transformation characterization value of the household product market statistical data set reaches 2050, the computer system can more confidently determine that the market has changed from a stable type to a growing type, rather than just based on the market tendency reference value of 1902.5.

[0066] In step S103b, the computer system obtains the uth market tendency mode change lower critical value corresponding to the change from the u+1th market tendency mode to the uth market tendency mode. This critical value corresponds to the change upper critical value in step S103a and is used to determine the transition from the growth market tendency mode to the stable market tendency mode.

[0067] Taking the stable market tendency model (the uth market tendency model) and the growth market tendency model (the u+1th market tendency model) as examples, when the market is in the growth mode, it can begin to shift to the stable mode due to various factors (such as intensified market competition, saturated consumer demand, etc.). Similar to changing the high critical value, changing the low critical value exists to prevent misjudging the shift in the market tendency model due to small fluctuations in the data.

[0068] The technical means by which the computer system obtains the lower critical value of the u-th market tendency pattern change also relies on the analysis of a large amount of historical data. Suppose the computer system analyzes the market data that has changed from a growth type to a stable type and finds that the mean of this set of data is 1800 and the standard deviation is 40. According to similar rules, the computer system can use the mean minus a certain multiple (such as 1 times) of the standard deviation as the lower critical value of the change. Then, the lower critical value of the u-th market tendency pattern change is 1800-40=1760. This means that when the transformation representation value of the household product market statistical data set is reduced to 1760, the computer system is more inclined to judge that the market has changed from a growth type to a stable type.

[0069] In the process of market tendency mode transformation, the transformation from a lower tendency mode to a higher tendency mode usually requires more significant changes to confirm, while the transformation from a higher tendency mode to a lower tendency mode is relatively more susceptible to some adverse factors. At the same time, the relationship that the u-th market tendency reference value is greater than the u-th market tendency mode change upper critical value is also to ensure the accuracy and logic in judging the transformation of the market tendency mode. For example, in the above example, the market tendency reference value is 1902.5, and the change upper critical value is 2050, which means that under normal circumstances, only when the conversion representation value exceeds 2050, it can be more certain that it is a transformation from a stable type to a growth type, and 1902.5 is only a preliminary reference value for distinction. Such a setting helps to avoid misjudgments caused by data instability or local fluctuations, thereby providing companies or analysts with more reliable market tendency analysis results so that they can formulate corresponding marketing strategies or decisions more accurately.

[0070] As an implementation manner, the step S103a, obtaining the u-th market tendency mode change upper critical value corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode, may include: Step S103a1: Acquire a plurality of first market statistical data set samples corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode; Step S103a2: obtaining a first sample conversion representation value corresponding to the first market statistical data set sample; Step S103a3: Determine the upper critical value of the u-th market tendency mode change based on the typical value of the first sample conversion characterization value.

[0071] In step S103a1, the computer system obtains a plurality of first market statistical data set samples corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode.

[0072] The first market statistical data set sample is selected from the actual home furnishing product market data, and is specifically used to analyze the transition process from the u-th market tendency mode to the u+1-th market tendency mode. These data sets contain rich information related to the home furnishing product market, such as the sales volume, sales price, sales ratio of different channels, geographical distribution and age distribution of consumers, and data related to product conversion (such as clicks, collections, purchases, shares, etc.), and these data sets are collected during the transition from the u-th to the u+1-th market tendency mode.

[0073] For example, assuming that the u-th market tendency pattern is a stable market tendency pattern, and the u+1-th market tendency pattern is a growing market tendency pattern, the computer system will filter out the home furnishing product market data from the historical database that are in the process of market transformation from stable to growing. This may involve analyzing data from different time periods to determine which time periods show such transformation characteristics.

[0074] The computer system can obtain these data set samples through database query operations. If the home product market data is stored in a relational database (such as MySQL), the computer system can use SQL query statements to filter out qualified data. The computer system can also pre-process the data to ensure the quality of the acquired data. For example, check the integrity of the data, handle missing values ​​(mean filling, median filling or other appropriate filling methods can be used), and identify and handle outliers (for example, determine outliers through the box plot method, and modify or delete them according to the specific situation). This helps to improve the accuracy of subsequent analysis.

[0075] In step S103a2, the first sample conversion characterization value is calculated based on the first market statistical data set sample obtained in step S103a1, and is used to comprehensively reflect the product conversion of these data set samples during the market trend pattern transformation process. The process of calculating this value is similar to the previous method of calculating the conversion characterization value, that is, it is necessary to extract data related to product conversion from the first market statistical data set sample, and use a suitable calculation method for comprehensive calculation.

[0076] Continuing with the above-mentioned example of the transformation from a stable to a growing market tendency pattern, for each sample set of first market statistical data, the computer system extracts data such as the number of clicks, collections, purchases, and shares of household products. Assume that in a sample set of first market statistical data, a certain type of household product has 8,000 clicks, 1,500 collections, 400 purchases, and 200 shares in a specific time period. The computer system uses the weighted average method to calculate the conversion representation value of the first sample, assuming that the weight of the click volume is 0.2, the weight of the collection volume is 0.15, the weight of the purchase volume is 0.4, and the weight of the share volume is 0.25. Then: . This is the first sample conversion characterization value corresponding to the first market statistical data set sample. The computer system calculates the corresponding first sample conversion characterization value for each acquired first market statistical data set sample in the same way. In addition to the weighted average method, the computer system can also use other methods to calculate the first sample conversion characterization value. For example, a comprehensive evaluation method based on the entropy weight method is used to first calculate the information entropy of each indicator (such as click volume, favorite volume, purchase volume, sharing volume, etc.), and then calculate the weight of each indicator based on the information entropy, and finally calculate the first sample conversion characterization value based on these weights.

[0077] In step S103a3, the computer system determines a higher critical value of the u-th market tendency pattern change based on a typical value (such as a statistic) of the first sample conversion characterizing the numerical value.

[0078] The typical value of the first sample transformation characterization value is a representative value obtained after statistical analysis of all the first sample transformation characterization values. The feasible statistics include mean, median, mode, etc. These statistics can reflect the central tendency or central position of the first sample transformation characterization values.

[0079] For example, the computer system obtains multiple first example conversion characterization values, which are 2035, 2100, 2250, 1950, 2300, etc. If the mean is used as the typical value, the computer system calculates the average value of these values ​​2127 as the basis for determining the upper critical value of the uth market tendency pattern change.

[0080] In addition to the mean, the median can also be used as a typical value. Sort these values ​​from small to large as: 1950, 2035, 2100, 2250, 2300, and the middle value 2100 is the median.

[0081] The method for the computer system to determine the upper threshold value of the u-th market tendency pattern change can be adjusted based on the typical value. One feasible method is to add or subtract a certain adjustment value based on the typical value. This adjustment value can be determined based on the fluctuation of market data or actual experience.

[0082] Assume that the computer system determines the adjustment value as 100 based on the fluctuation of market data, and takes the mean as the typical value. Then the upper critical value of the u-th market tendency mode change is 2127+100=2227.

[0083] Another way to determine the change-up threshold is to consider the distribution of the values ​​of the first sample transformation representation. If the distribution of these values ​​shows a certain skewness (for example, right skewness, that is, a few large values ​​pull up the overall mean), the computer system can use the median as the typical value and add an adjustment value to the median based on a certain proportion (for example, 10% determined based on historical data). Assuming the median is 2100, the adjustment value is increased by 10%, and the adjustment value is 2100×0.1=210, then the change-up threshold of the u-th market tendency model is 2100+210=2310.

[0084] By determining the upper critical value of the u-th market tendency pattern change based on the typical value of the numerical representation converted based on the first sample, the computer system can more accurately grasp the critical state when the u-th market tendency pattern changes to the u+1-th market tendency pattern, avoiding misjudgment caused by simply using a single numerical value or not considering data distribution, thereby providing a more reliable basis for accurate judgment of market tendency patterns, and helping enterprises or analysts to better formulate strategies that adapt to market changes.

[0085] In one implementation, step S103b, obtaining the uth market tendency mode change lower critical value corresponding to the change from the u+1th market tendency mode to the uth market tendency mode, may include: Step S103b1: obtaining a plurality of second market statistical data set samples corresponding to the change from the u+1th market tendency mode to the uth market tendency mode; Step S103b2: obtaining a second sample conversion representation value corresponding to the second market statistical data set sample; Step S103b3: Based on the typical value of the numerical representation of the second sample conversion, determine the u-th market tendency mode change lower critical value; wherein, the error between the u-th market tendency mode change upper critical value and the u-th market tendency mode change lower critical value is greater than the set error.

[0086] In step S103b1, the second market statistical data set sample is a household product market data set specifically for the process of transforming from the u+1th market tendency mode (such as the growth market tendency mode) to the uth market tendency mode (such as the stable market tendency mode). These data sets also contain rich household product market related information, such as the sales quantity, sales price, sales ratio of different channels, geographical distribution and age distribution of consumers, and product conversion related data (such as clicks, favorites, purchases, shares, etc.).

[0087] In step S103b2, the second sample conversion characterization value is calculated based on the second market statistical data set sample obtained in step S103b1, which comprehensively reflects the product conversion of these data set samples in the process of changing from the u+1th market tendency mode to the uth market tendency mode. The process of calculating this value is similar to the method of calculating the conversion characterization value before, that is, it is necessary to extract data related to product conversion from the second market statistical data set sample and use a suitable calculation method for comprehensive calculation.

[0088] Taking the transition from a growth-oriented market tendency model to a stable market tendency model as an example, for each sample of the second market statistical data set, the computer system will extract data such as the number of clicks, favorites, purchases, and shares of household products. Assume that in a sample of the second market statistical data set, a certain type of household product has 6,000 clicks, 1,000 favorites, 300 purchases, and 150 shares in a specific time period.

[0089] The computer system uses the weighted average method to calculate the conversion representation value of the second example. Assume that the weight of clicks is 0.2, the weight of favorites is 0.15, the weight of purchases is 0.4, and the weight of shares is 0.25. Then: This is the second sample conversion characterization value corresponding to the second market statistical data set sample. The computer system calculates the corresponding second sample conversion characterization value for each acquired second market statistical data set sample in the same way.

[0090] In addition to the weighted average method, other methods may be used to calculate the second sample transformation representation value, for example, a derivative algorithm of the principal component analysis (PCA) method, which will not be described here.

[0091] In step S103b3, the typical value of the second sample conversion characterization value is a representative value obtained after statistical analysis of all the second sample conversion characterization values. Feasible typical values ​​include statistics such as mean, median, mode, etc., which can reflect the central tendency or central position of the second sample conversion characterization value. For example, the computer system obtains multiple second sample conversion characterization values, which are 1507.5, 1450, 1550, 1400, 1600, etc. If the mean is used as the typical value, the computer system calculates the average value of these values: ; This mean value of 1501.5 can be used as the basis for determining the lower critical value of the u-th market tendency mode change. In addition to the mean, the median can also be used as a typical value. These values ​​are sorted from small to large as follows: 1400, 1450, 1507.5, 1550, 1600, and the middle value of 1507.5 is the median.

[0092] The method for the computer system to determine the lower critical value of the u-th market trend pattern change can be adjusted based on the typical value. One feasible method is to add or subtract a certain adjustment value based on the typical value. This adjustment value can be determined based on the fluctuation of market data or actual experience.

[0093] Assuming that the computer system determines the adjustment value to be 50 based on the fluctuation of market data, and takes the mean as the typical value, then the lower critical value of the u-th market tendency mode change is 1501.5-50=1451.5.

[0094] Another way to determine the lower critical value of the change is to consider the distribution of the values ​​of the second sample transformation representation. If the distribution of these values ​​shows a certain skewness (for example, left skewness, that is, a few small values ​​pull down the overall mean), the computer system can use the median as the typical value and subtract an adjustment value from the median according to a certain proportion (for example, 10% determined based on historical data). Assuming that the median is 1507.5, the adjustment value determined according to the proportion of 10% is 1507.5×0.1=150.75, then the lower critical value of the u-th market tendency model change is 1507.5-150.75=1356.75.

[0095] By determining the lower critical value of the u-th market tendency mode change based on the typical value of the second sample conversion representation value, the computer system can more accurately grasp the critical state when the u+1-th market tendency mode changes to the u-th market tendency mode, avoiding misjudgment caused by simply using a single value or not considering data distribution, thereby providing a more reliable basis for accurate judgment of the market tendency mode, helping enterprises or analysts to better formulate strategies to adapt to market changes. For example, when the market conversion representation value is close to this lower critical value of change, the enterprise can prepare in advance to cope with the market's transformation from growth to stability, such as adjusting production plans, optimizing marketing strategies, etc.

[0096] As an implementation mode, the X market tendency patterns include a negative market tendency pattern and a positive market tendency pattern; the market statistical data set samples under the negative market tendency pattern belong to a first market statistical data set set composed of sub-market statistical data set samples corresponding to R negative sub-market tendency patterns, and the market statistical data set samples under the positive market tendency pattern belong to a second market statistical data set set composed of sub-market statistical data set samples corresponding to S positive sub-market tendency patterns, R≥1, S≥1.

[0097] On this basis, the step S102, based on the market statistical data set sample under the u-th market tendency mode, determines the u-th conversion representation value corresponding to the u-th market tendency mode, which may include: Step S102a: Determine R sub-conversion representation values ​​corresponding to the R sub-market statistical data set samples in the first market statistical data set, and determine the conversion representation value corresponding to the negative market tendency pattern based on the R sub-conversion representation values.

[0098] The first market statistical data set is a set related to the market statistical data set sample under the negative market tendency model. The sub-market statistical data set sample here is obtained by further subdividing the market statistical data under the negative market tendency model according to a certain classification method. For example, under the negative market tendency model of the home product market, the market statistical data can be subdivided into furniture (such as sofas, tables and chairs, etc.), home accessories (such as decorative paintings, ornaments, etc.) and kitchen supplies (such as pots, tableware, etc.) according to product type. Here, it is assumed that R=3.

[0099] For each sub-market statistical data set sample, the computer system determines the corresponding sub-conversion characterization value. This sub-conversion characterization value is a value that comprehensively reflects the product conversion situation in the sub-market. The method for calculating the sub-conversion characterization value is similar to the previous method for calculating the conversion characterization value. It is necessary to extract data related to product conversion from the sub-market statistical data set sample, such as click volume, favorite volume, purchase volume, and share volume.

[0100] The step S103, determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value, may include: Step S1031': determining a first numerical range, wherein the maximum value of the first numerical range is a conversion characterization value corresponding to a positive market tendency mode, and the minimum value of the first numerical range is a conversion characterization value corresponding to a negative market tendency mode; Step S1032': determining a positive or negative market tendency reference value based on the first numerical range, wherein the positive or negative market tendency reference value belongs to the first numerical range.

[0101] The positive market tendency mode and the negative market tendency mode are two different market states, and their corresponding conversion representation values ​​respectively reflect the overall conversion of the home furnishing product market under their respective modes. For example, assuming that the conversion representation value corresponding to the negative market tendency mode calculated by the previous steps is 1125 (such as the calculation result in the above step S102a), the conversion representation value corresponding to the positive market tendency mode is 2000 (assuming that it is obtained through a similar calculation process).

[0102] Then the first numerical range is from 1125 (inclusive) to 2000 (inclusive). This numerical range covers all possible situations of transforming the representation value from the negative market tendency mode to the positive market tendency mode. The computer system determines this numerical range in order to further determine the positive and negative market tendency reference value, which will help distinguish whether the home furnishing product market is more inclined to the positive market tendency mode or the negative market tendency mode.

[0103] The computer system can determine the first numerical range by simple numerical comparison and storage operations. In the memory or database of the computer system, the conversion characterization values ​​corresponding to the negative market tendency mode and the positive market tendency mode have been stored. The computer system can read these values ​​and determine the first numerical range.

[0104] In step S1032', the positive and negative market tendency reference value is a critical value, which is used to distinguish whether the home furnishing product market tends to be in a positive market tendency mode or a negative market tendency mode. When determining this reference value, the computer system needs to comprehensively consider the first value range and the actual market situation.

[0105] One feasible determination method is to take the middle value of the first numerical range. However, market conditions are often complex, and the distribution of conversion characterization values ​​may not be uniform. For example, through the analysis of a large amount of historical data, the computer system finds that the conversion characterization values ​​under the negative market tendency mode are mostly concentrated near the lower limit of the numerical range, while the conversion characterization values ​​under the positive market tendency mode are mostly concentrated near the upper limit of the numerical range. In this case, the computer system can use the value that is more biased towards the negative market tendency mode as the positive and negative market tendency reference value. For example, the computer system can determine the positive and negative market tendency reference value as 1400. In addition to considering the imbalance of numerical distribution, the computer system can also determine the positive and negative market tendency reference value in combination with other characteristics of the market. For example, if in the negative market tendency mode, the market competition is fierce and the product homogeneity is serious, resulting in a small fluctuation of the conversion characterization value and a tendency to a lower level; while in the positive market tendency mode, the market innovation is active, the product differentiation is obvious, and the conversion characterization value fluctuates greatly and tends to a higher level. The computer system can adjust the positive and negative market tendency reference value based on these market characteristics, combined with expert experience or market research results.

[0106] When determining the reference value of the positive or negative market tendency, the computer system can calculate it in the program according to the pre-set algorithm and rules. If the method of taking the middle value is adopted, a simple mathematical operation formula can be used, such as the above (min+max) / 2 (where min is the minimum value of the first numerical range and max is the maximum value of the first numerical range).

[0107] As an implementation manner, the X market tendency modes include a negative market tendency mode and a positive market tendency mode; in the step S103, after determining the u-th market tendency reference value based on the u-th conversion characterization value and the u+1-th conversion characterization value, the method may further include: Step S103A: obtaining a first market tendency mode change threshold corresponding to the change from the negative market tendency mode to the positive market tendency mode; Step S103B: Obtain the second market tendency mode change critical value corresponding to the change from the positive market tendency mode to the negative market tendency mode; wherein the first market tendency mode change critical value is greater than the second market tendency mode change lower critical value, and the positive and negative market tendency reference value is greater than the first market tendency mode change upper critical value.

[0108] In step S103A, the computer system obtains a first market tendency mode change threshold corresponding to the change from the negative market tendency mode to the positive market tendency mode.

[0109] The negative market tendency model and the positive market tendency model represent two completely different states of the home furnishing product market. In the negative market tendency model, it may be manifested as slow sales growth of home furnishing products, low consumer willingness to buy, and low product conversion rate. For example, there is a serious backlog of home furnishing products in the market, and new products rarely receive consumer attention. The conversion data such as click volume and purchase volume are at a low level. The positive market tendency model is the opposite, with rapid sales growth, high consumer purchasing enthusiasm, and considerable product conversion rate. For example, new demands for home furnishing products continue to emerge, the purchase volume of popular products continues to rise, and more clicks and collections are driven through various channels (such as social media sharing, consumer recommendations, etc.).

[0110] The first market tendency mode change critical value is a boundary. When the relevant indicators of the household product market statistical data set (such as the conversion representation value) reach or exceed this critical value, the computer system can determine that the market has changed from a negative market tendency mode to a positive market tendency mode. The determination of this critical value requires the computer system to comprehensively consider multiple factors and analyze based on a large amount of historical data.

[0111] The technical means by which the computer system obtains this critical value may be through data mining of multiple cases in which the market tendency pattern has changed from a negative to a positive market tendency pattern in history. For example, the computer system selects all data that record the process of market tendency pattern change from the home furnishing product market database, and selects the cases that have changed from negative to positive to form a specific data set. Assume that this data set contains home furnishing product market data in multiple time periods, such as sales data and conversion data of home furnishing products in different regions and different types. For each case, the computer system calculates some key indicator values ​​before and after the moment of change. These values ​​can comprehensively reflect the market transformation, similar to the conversion representation values ​​mentioned earlier.

[0112] For example, in a case of a change from negative to positive, the calculated composite value is 1800 (this is just an assumption, the actual calculation may involve complex calculations such as weighting of multiple data indicators). The computer system performs statistical analysis on similar values ​​in all these cases, and may use statistics such as mean and median to determine the critical value of the first market tendency pattern change.

[0113] If the mean method is used, assuming that the mean of these values ​​is 1900, the computer system can preliminarily determine 1900 as the first market tendency model change critical value. This means that when the comprehensive value corresponding to the new set of household product market statistical data reaches or exceeds 1900, it is likely to indicate that the market is changing from a negative market tendency model to a positive market tendency model.

[0114] Market conditions are complex and changeable, and relying solely on the mean may not be accurate enough. The computer system can also consider the distribution of the data, such as standard deviation. If the data distribution is more dispersed and the standard deviation is large, the critical value may need to be appropriately adjusted. For example, if the standard deviation is 100, the computer system can determine the critical value based on experience or further data analysis by adding a certain proportion (such as 50%) of the standard deviation to the mean, that is, 1900+100×0.5=1950, and use 1950 as the final first market tendency mode change critical value.

[0115] In step S103B, the computer system obtains a second market tendency mode change threshold corresponding to the change from the positive market tendency mode to the negative market tendency mode.

[0116] In contrast to the transition from negative to positive, the change from a positive market tendency mode to a negative market tendency mode also requires a clear boundary to judge. In the positive market tendency mode, the market is in a prosperous state, and various conversion indicators of the product perform well, but when certain factors (such as intensified market competition, sudden changes in consumer preferences, deteriorating economic environment, etc.) appear, the market may begin to shift to a negative market tendency mode.

[0117] The computer system obtains the second market tendency mode change threshold value based on a large amount of historical data. For example, all historical cases of changes from positive market tendency mode to negative market tendency mode are found in the database to form a special data set.

[0118] For each case in this data set, the computer system analyzes the relevant indicator values ​​during the transition process. Assume that in one case, the calculated comprehensive value reflecting the market transformation is 1500 (this is also a simplified value assumed here). Perform statistical analysis on such values ​​in all these cases, such as calculating the mean, median, etc. Assume that the mean of these values ​​is 1400 after calculation, and 1400 is initially used as the reference value for the critical value of the second market tendency model change. However, due to the complexity of market data, further adjustments may be required. Considering that the transition from a positive market tendency model to a negative market tendency model may be relatively more susceptible to some sudden factors, and the data distribution may present different characteristics. For example, if the data shows that the value drops faster and fluctuates more when changing from positive to negative, the computer system can be adjusted according to the distribution characteristics of the data. Assuming the standard deviation is 80, the computer system subtracts a certain proportion (such as 30%) of the standard deviation from the mean to determine the final critical value, that is, 1400-80×0.3=1376, based on experience or further analysis, and uses 1376 as the critical value of the second market tendency model change.

[0119] It should be noted that the first market tendency mode change critical value is greater than the second market tendency mode change lower critical value. This is because the change from a negative market tendency mode to a positive market tendency mode usually requires more significant market positive signals and needs to overcome more negative factors. For example, the improvement of consumer confidence and the re-stimulation of market demand are not achieved overnight, so a higher critical value is needed to judge this change. The change from a positive market tendency mode to a negative market tendency mode may be more susceptible to some negative factors, such as the sudden low-price strategy of competitors and short-term fluctuations in the macroeconomic environment, so the critical value is relatively low. At the same time, the relationship that the positive and negative market tendency reference value is greater than the first market tendency mode change upper critical value is also reasonable. The positive and negative market tendency reference value is a reference standard for preliminarily distinguishing between positive and negative market tendency modes, while the first market tendency mode change upper critical value is a more rigorous boundary for judging the change from negative to positive. For example, the reference value of positive and negative market tendencies may be 1500, and the upper critical value of the first market tendency pattern change is 1950. When the corresponding value of the market data exceeds 1950, it can be more certain that the market has changed from negative to positive. 1500 is only a preliminary reference limit. Such a setting helps to improve the accuracy of the judgment of changes in market tendency patterns and avoid misjudgment, thereby providing a more reliable basis for enterprises to formulate market strategies.

[0120] As an implementation manner, the X market tendency patterns include a negative market tendency pattern and a positive market tendency pattern, and the u-th market tendency reference value is a positive and negative market tendency reference value; based on this, the step S400, comparing the transformed representation value of the household product market statistical data set with the market tendency reference value to obtain the market tendency analysis result of the household product market statistical data set, may include: Step S410: comparing the transformed representation value of the household product market statistical data set with the positive and negative market tendency reference value; Step S420: if the conversion representation value of the household product market statistical data set is greater than the positive and negative market tendency reference value, the market tendency analysis result is obtained that the household product market statistical data set belongs to a positive market tendency mode; Step S430: If the conversion representation value of the household product market statistical data set is less than or equal to the positive and negative market tendency reference value, the market tendency analysis result is obtained that the household product market statistical data set belongs to a negative market tendency mode.

[0121] In step S410, the computer system compares the converted representative value of the household product market statistical data set with the positive and negative market tendency reference values.

[0122] The conversion characterization value of the household product market statistical data set is a comprehensive value determined by the computer system in the previous step (such as step S300) based on the conversion rate of the product market data in the target product market data set. This value reflects the overall characteristics of the household product market statistical data set in terms of conversion. For example, it is assumed that through a series of calculations, the conversion characterization value of the household product market statistical data set is 1600. The positive and negative market tendency reference value is determined in the previous step (such as steps S1031'-S1032' in step S103). It is a critical value used to distinguish whether the household product market tends to be in a positive market tendency mode or a negative market tendency mode. For example, it is assumed that this positive and negative market tendency reference value is 1500. When the computer system performs a comparison, it is an operation based on numerical comparison. In the program implementation of the computer system, a simple conditional judgment statement can be used to implement this comparison. This comparison is a quantitative assessment of the current state of the household product market. By comparing the value reflecting the overall conversion situation with a predetermined reference value, a basis for judging the market tendency is provided.

[0123] In step S420, if the conversion representation value of the household product market statistical data set is greater than the positive or negative market tendency reference value, the computer system obtains a market tendency analysis result that the household product market statistical data set belongs to a positive market tendency mode.

[0124] Continuing with the above example, when the conversion representation value of the household product market statistical data set is 1600, and the positive and negative market tendency reference value is 1500, since 1600 > 1500, the computer system determines that this household product market statistical data set belongs to the positive market tendency model.

[0125] When the conversion representation value is greater than the positive and negative market tendency reference value, it means that the conversion of home furnishing products in the market is good, showing the characteristics of the positive market tendency model. For example, in the positive market tendency model, the number of clicks on home furnishing products may be high, and more consumers show interest in the product and click to view detailed information; the number of collections may also be at a high level, indicating that consumers have the potential to buy or are relatively satisfied with the product and hope to pay attention to it later; the number of purchases is relatively large, reflecting that the actual sales situation is good; the number of shares may also be large, which means that consumers are willing to recommend the product to others, and this word-of-mouth communication helps to further expand the market influence of the product.

[0126] In step S430, if the conversion representation value of the household product market statistical data set is less than or equal to the positive or negative market tendency reference value, the computer system obtains a market tendency analysis result that the household product market statistical data set belongs to a negative market tendency mode.

[0127] Assume that the conversion representation value of the household product market statistical data set is 1400, and the positive and negative market tendency reference value is 1500. Since 1400 < 1500, the computer system determines that this household product market statistical data set belongs to the negative market tendency mode.

[0128] In the negative market tendency mode, the market performance of home furnishing products is relatively poor. For example, the number of clicks on the product may be low, such as only 50,000 clicks per month (compared to more than 100,000 in the positive market tendency mode), the number of collections may be only 10,000 (compared to more than 20,000), the number of purchases may be only 3,000 sets (compared to more than 5,000 sets), and the number of shares may be only 1,500 times (compared to more than 3,000 times). This low conversion data combined makes the conversion representation value low, which is consistent with the characteristics of the negative market tendency mode.

[0129] As an implementation manner, before obtaining X market statistical data set samples under market tendency modes in step S101, the method may further include: Step S101a: Determine, according to the market analysis dimension, a market analysis indicator corresponding to the market analysis dimension.

[0130] Market analysis dimensions are different angles or aspects of analyzing the home furnishing product market. For example, the market can be analyzed from the regional dimension, which focuses on the performance of the home furnishing product market in different geographical regions; it can also be analyzed from the consumer age dimension to understand the demand and purchasing behavior of consumers of different age groups for home furnishing products; it can also be analyzed from the product type dimension, such as the market conditions of different types of home furnishing products such as furniture, home accessories, and kitchen supplies. Market analysis indicators are specific data or variables used to measure and describe the characteristics of these market analysis dimensions. Taking the regional dimension as an example, the corresponding market analysis indicators may include home furnishing product sales, sales volume, market growth rate, etc. in different regions. If the consumer age dimension is taken as an example, the market analysis indicators may be the purchase frequency, average purchase amount, and preference for different brands of consumers of different age groups. For the product type dimension, market analysis indicators can be the market share of various types of home furnishing products, the speed of new product launch, and the sales ratio of different price ranges.

[0131] Assuming that the computer system selects the regional dimension as the market analysis dimension, the market analysis indicators it determines may be calculated and obtained in this way. For example, the sales indicator can be obtained by summing up the household product sales records in each region. If there is a table called "sales_records" in the database, which contains the "region" field and the "amount" field, the computer system can use the SQL query statement "SELECT region, SUM(amount) AS total_sales FROM sales_records GROUP BY region;" to obtain the total sales of each region as a market analysis indicator. The sales volume indicator can also be similarly calculated by summarizing the sales quantities in the sales records.

[0132] By determining the market analysis dimensions and the corresponding market analysis indicators, the computer system lays the foundation for the subsequent acquisition of a sample set of market statistical data under the market tendency model, making the acquired data more targeted and valuable for analysis.

[0133] Step S101, obtaining a set of samples of market statistical data under X market tendency modes, may include: Step S1011: obtaining market statistical data set samples of the same category of market analysis indicators under X market tendency modes.

[0134] As mentioned above, the market tendency model here is a different classification description of the overall situation of the home furnishing product market, such as a positive market tendency model, a negative market tendency model or other more detailed models. After determining the market analysis indicators (as determined in the previous step S101a), the computer system obtains relevant market statistical data set samples for each market tendency model.

[0135] Assume that the market analysis indicators are determined as the sales and sales volume of household products in different regions, and there are three market tendency models: growth market tendency model, stable market tendency model and decline market tendency model (X=3). For the growth market tendency model, the computer system will obtain the sales and sales volume data of household products in different regions under this market tendency model from various data sources (such as the sales database of the enterprise, the data of the market research agency, etc.) to form a set of market statistical data samples. For example, under the growth market tendency model, it may be found that the sales of household products in a certain emerging urban area have increased by 30% and the sales volume has increased by 25% in the past year; while in some developed regions, the sales have increased by 15% and the sales volume has increased by 10%, etc. These data constitute the set of market statistical data samples under the growth market tendency model.

[0136] For the stable market tendency model, the data obtained by the computer system show that in some traditional regions with high market saturation, the sales volume and sales amount of household products have remained basically stable in the past few years, with a fluctuation range of ±5%. The data from these regions are aggregated to form a sample set of market statistical data under the stable market tendency model.

[0137] For the decline-type market tendency model, the computer system may find that in some old areas or areas that are more affected by emerging alternatives, the sales of home products have declined at an annual rate of 10%-20% in the past few years, and the sales volume has also decreased accordingly. The data from these areas constitute a sample set of market statistical data under the decline-type market tendency model.

[0138] Computer systems can use data integration tools to obtain this data. If the data source is multiple databases with different structures, ETL (Extract, Transform, Load) tools can be used. For example, for the internal sales database of the enterprise (such as the relational database Oracle) and the data stored in the form of CSV files provided by external market research institutions, open source ETL tools such as Pentaho Data Integration can be used. First, in the extraction (Extract) stage, SQL query statements are used to obtain relevant sales data from the Oracle database, and the data in the CSV file is read at the same time; in the transformation (Transform) stage, the data format is unified, and the data is cleaned (such as processing missing values, outliers) and other operations are performed; in the loading (Load) stage, the processed data is stored in a database or data warehouse specifically for analysis, forming a collection of market statistical data samples for each market tendency model.

[0139] In step S103, after determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value, the method may further include: Step S10301: Determine a first corresponding relationship between the u-th market tendency reference value and the corresponding market analysis dimension.

[0140] The uth market tendency reference value is a critical value determined in step S103 for distinguishing different market tendency patterns. The market analysis dimension is an angle of analyzing the home product market determined in step S101a, such as a regional dimension, a consumer age dimension, a product type dimension, and the like.

[0141] The determination of this first corresponding relationship helps to link the market tendency reference value with the specific market analysis dimension, so that in the subsequent market analysis, the corresponding market tendency reference value can be accurately used according to the specific market analysis dimension.

[0142] For example, suppose the market analysis dimension is the regional dimension, and there are different regional divisions such as the eastern region and the western region. For different regions, market performance may be different, resulting in different market tendency reference values ​​in different regions. By analyzing a large amount of historical data, the computer system found that due to the high level of economic development and newer consumption concepts in the eastern region, its market tendency reference value (used to distinguish different market tendency patterns, such as positive and negative market tendency patterns) may be relatively high. For example, in the eastern region, when the conversion representation value of the household product market statistical data set reaches 1800 (assuming that this 1800 is the positive and negative market tendency reference value of the eastern region), it may be judged as a positive market tendency pattern; while in the western region, due to differences in economic development level, consumption capacity and other factors, its positive and negative market tendency reference value may be 1600.

[0143] The technical means by which the computer system determines this correspondence can be based on the association rule mining algorithm in data mining. Association rule mining aims to discover interesting associations between different items in a data set. In this scenario, different values ​​of the market analysis dimension (such as the eastern region, the western region, etc.) are taken as one item set, and the market tendency reference value is taken as another item set. The computer system mines a large amount of historical data containing regional information and corresponding market tendency reference values ​​to find the association rules between them. For example, using the Apriori algorithm, the computer system can analyze which regional characteristics (such as population density, per capita income and other regional related factors) have a strong correlation with a specific market tendency reference value, thereby determining the first correspondence.

[0144] The determination of this correspondence makes the market analysis more refined and accurate. In actual market analysis, when conducting a market trend analysis of home furnishing products in a specific region, the computer system can use the market trend reference value corresponding to the region instead of using a unified reference value that does not consider regional differences, thereby improving the accuracy of the market trend analysis results.

[0145] In the step S400, before comparing the converted representative value of the household product market statistical data set with the market trend reference value, the method may further include: Step S401: Determine the current market analysis dimension; Step S402: Based on the current market analysis dimension, a market tendency reference value having a first corresponding relationship with the current market analysis dimension is obtained for comparison with the conversion representation value of the household product market statistical data set.

[0146] In step S401, the current market analysis dimension is determined based on the needs of the enterprise or analyst, the focus of market research, etc. For example, the enterprise may be concerned about the market tendency of consumers of different age groups for home furnishing products, and the market analysis dimension at this time is the consumer age dimension; or the enterprise wants to understand the tendency of different product types in the home furnishing product market, and the market analysis dimension at this time is the product type dimension.

[0147] Suppose a company has recently launched a series of smart home products and wants to know the market trends of these products under different product type dimensions (such as smart home furniture, smart kitchen appliances, etc.), the computer system will determine that the current market analysis dimension is the product type dimension.

[0148] In step S402, continuing to take the above product type dimension as an example, since the first corresponding relationship between the market tendency reference value and the market analysis dimension has been determined in step S10301, if the computer system determines that the current market analysis dimension is the product type dimension, it will obtain the market tendency reference value related to the product type dimension from the previously established corresponding relationship.

[0149] For example, for the smart home furniture product type, historical data analysis found that its market tendency reference value (used to distinguish between positive and negative market tendency patterns) is 1700; for the smart kitchen appliance product type, its market tendency reference value may be 1650. When the computer system analyzes the market tendency of smart home furniture products, it will obtain the market tendency reference value of 1700, and then compare the transformed representation value of the market statistical data set of smart home furniture products with 1700 to determine its market tendency pattern. The computer system can establish an association table in the database to store the corresponding relationship between the market analysis dimensions and the market tendency reference values.

[0150] As an implementation method, there is a second correspondence between the target market analysis dimension and the target market analysis indicator, wherein the market tendency reference value that has a first correspondence with the target market analysis dimension is determined by a sample set of market statistical data for the target market analysis indicator.

[0151] Based on this, before extracting target elements from the household product market statistical data set collected from the target analysis group in step S100, the method may further include: Step S1001: Determine the current market analysis dimension; Step S1002: Based on the current market analysis dimension, determine a current market analysis indicator having a second corresponding relationship with the current market analysis dimension; Step S1003: Determine whether the household product market statistical data set contains the current market analysis indicators; wherein, when the household product market statistical data set contains the current market analysis indicators, perform a process of extracting target elements from the household product market statistical data set collected from the target analysis group; when the household product market statistical data set does not contain the current market analysis indicators, obtain the household product market statistical data set again.

[0152] The market analysis dimension is a specific angle or scope for analyzing the home furnishing product market. The determination of this dimension depends on many factors, such as the company's market strategy goals, the focus of market research, or to respond to specific market issues.

[0153] For example, if a company wants to understand the acceptance and purchasing behavior differences of household products among consumers in different regions in order to carry out targeted regional market expansion, the market analysis dimension determined by the computer system is the geographical dimension. In this case, the region can be segmented according to different classification standards, such as by country, province, city, or by the city's economic development level (such as first-tier cities, second-tier cities, etc.), geographical location (such as coastal areas, inland areas, etc.), etc.

[0154] The computer system can obtain the setting information about the market analysis dimension by reading the analysis strategy file pre-set by the enterprise or from the decision support system (DSS) within the enterprise. If it is based on the database management system (DBMS) within the enterprise, such as a relational database (such as MySQL), there can be a special configuration table (such as "analysis_strategy_config"), which contains the field "analysis_dimension" to store the information related to the currently set market analysis dimension. The computer system obtains the current market analysis dimension by executing SQL query statements such as "SELECT analysis_dimension FROManalysis_strategy_config;".

[0155] In step S1002, once the current market analysis dimension is determined, the computer system needs to determine the market analysis indicators related thereto. This relationship is established based on the understanding of the market operation mechanism and data characteristics, that is, the second corresponding relationship.

[0156] Taking the regional dimension as an example, if it is the current market analysis dimension, then the market analysis indicators with a second corresponding relationship may include household product sales volume, sales revenue, market penetration rate, sales ratio of different channels (such as online and offline) in each region, etc. For example, when analyzing the market situation of a household product in different provinces, the sales volume can directly reflect the popularity of the product in various regions, the sales revenue can reflect the market value, the market penetration rate (the proportion of households that purchase the household product in a specific region to the total number of households) can measure the market coverage of the product in the region, and the sales ratio of different channels can help understand the sales model preferences of various regions.

[0157] If the current market analysis dimension is the consumer age dimension, the relevant market analysis indicators may include the average purchase frequency, average purchase amount, awareness of different brands, price sensitivity, etc. of consumers of different age groups. For example, for young consumers aged 20-30, they may have a relatively low purchase frequency of home products, but a high awareness of new brands and fashion styles, and price sensitivity may vary due to differences in consumption concepts. These indicators can help to gain a deeper understanding of the market behavior of consumers in this age group.

[0158] When the product type dimension is the current market analysis dimension, the market analysis indicators may include the market share of various home furnishing products, the speed of product replacement, the matching degree of product functional diversity and market demand, etc. For example, for furniture products, the market share reflects its position in the entire home furnishing product market, the speed of product replacement (such as the frequency of new furniture launches) affects its market competitiveness, and the matching degree of product functional diversity (such as multifunctional sofas, foldable furniture, etc.) and market demand determines the popularity of the product.

[0159] The technical means by which the computer system determines these corresponding relationships can be based on a combination of data dictionaries and expert knowledge. The data dictionary is a detailed description of the definition, structure, and interrelationships of data elements in the database, which can be stored in the database or in the form of a special file. Expert knowledge comes from the in-depth understanding of the market by market researchers, industry experts, etc.

[0160] In step S1003, the household product market statistical data set is a set of various household product market-related data, which may come from multiple data sources, such as the sales record system of the enterprise, the data report of the market research organization, the data of the e-commerce platform, etc. The computer system needs to check whether this data set contains data related to the current market analysis index determined in step S1002.

[0161] For example, if the current market analysis dimension is the region dimension, and the corresponding market analysis indicators are sales volume and sales amount, the computer system will search in the home product market statistical data set whether there is sales volume and sales amount data divided by region. Assume that the home product market statistical data set is stored in a table named "home_product_market_data" in a relational database, which contains fields such as "region", "product_type", "quantity_sold" and "total_amount". The computer system can check the existence of data by executing SQL query statements, such as "SELECT COUNT(*) FROM home_product_market_data WHERE quantity_sold IS NOT NULL AND total_amount IS NOT NULL;". If the query result is greater than 0, it means that there is sales volume and sales amount data in the data set, which meets the requirements of the current market analysis indicators; if the result is 0, it means that these necessary data are missing in the data set.

[0162] If the current market analysis indicator does not exist, the computer system obtains the household product market statistical data set again. The way of re-acquisition may vary depending on the data source. If the data source is the internal sales record system of the enterprise, the computer system can request the sales department to update the data or re-extract the data from the sales database. For example, if the sales volume and sales data are missing due to data synchronization problems, the sales department can re-run the data synchronization program to update the latest sales data to the database, and then the computer system can check again whether the data exists. If the data source is an external market research organization, the computer system can send a request to obtain the latest report containing the required indicator data.

[0163] When the household product market statistical data set contains the current market analysis index, the computer system can subsequently perform a process of extracting target elements from the household product market statistical data set collected from the target analysis group (ie, step S100 ).

[0164] As an implementation manner, the step S200, based on the product conversion distribution state, determines the product market data whose conversion rate reaches the set requirement, and obtains the target product market data set, may include: Step S210: determining a total numerical range of product market conversion rate data in the household product market statistical data set based on the product conversion distribution state;

[0165] Step S220: determining a target numerical interval in the total numerical range, wherein a conversion rate of product market data in the target numerical interval reaches product market data of set requirements, wherein the set required product market data conversion rate is greater than the conversion rate of other product market data in the total numerical range except the target numerical interval, or the set required product market data conversion rate is less than the conversion rate of other product market data in the total numerical range except the target numerical interval.

[0166] In step S210, the computer system obtains the product conversion distribution status, and determines a total value range of the product market conversion rate data based on these conversion data. The computer system obtains the respective conversion rates of all conversion types in a similar calculation method, and then determines the range of these conversion rate values. This range may be an interval between a minimum value and a maximum value.

[0167] For example, a computer system analyzes each product data in a set of household product market statistics data and calculates the conversion rate for each conversion operation. Assume that the product market data is stored in a table with columns such as product number, number of clicks, number of favorites, number of purchases, number of shares, and total number of products. For the calculation of purchase conversion rate, the computer system can use the formula: purchase conversion rate = number of purchases / total number of products. The same is true for other conversion types. Then, by comparing the various conversion rate values ​​of all products, find the minimum and maximum values, and thus determine the total value range. For example, after calculation, the minimum purchase conversion rate of all products is 0.05 (5%), and the maximum is 0.3 (30%), so the total value range of the purchase conversion rate data is [0.05, 0.3].

[0168] In step S220, after determining the total value range, the computer system finds the target value interval within the range according to the set requirements. The set requirements are based on business needs or market strategies, for example. For example, if the business wants to focus on product market data with high conversion rates, then the set requirement may be that the conversion rate of the product market data is greater than the conversion rate of the remaining product market data in the total value range except the target value interval. Assuming that the total value range is [0.05, 0.3], if the set requirement is to focus on high-conversion products, and the computer system determines that the product market data with a conversion rate greater than 0.2 (20%) meets the requirements, then the target value interval is [0.2, 0.3]. Conversely, if the business wants to focus on low-conversion products for improvement strategies, the set requirement may be that the conversion rate of the product market data is less than the conversion rate of the other product market data in the total value range except the target value interval. For example, if the set requirement is to focus on product market data with a conversion rate less than 0.1 (10%), then the target value interval is [0.05, 0.1]. For example, the computer system can compare each value with the set requirement by traversing each value in the total value range. If the set requirement is greater than other conversion rates, the computer system can start from the maximum value of the total value range and gradually compare downwards. When a value is found, the value below it meets the set requirement, the upper limit of the target value range is determined. The lower limit can be this value or obtained through further precise comparison. If the set requirement is less than other conversion rates, the computer system can start from the minimum value of the total value range and gradually compare upwards. When a value is found, the value above it meets the set requirement, the lower limit of the target value range is determined. The upper limit can be this value or obtained through further precise comparison.

[0169] Step S300, based on the conversion rate of the product market data in the target product market data set, determines the conversion representation value of the household product market statistical data set, which may include: Step S310: quantifying the conversion rate of the product market data in the target product market data set to obtain a quantified value of the product market data; Step S320: Determine a conversion representation value of the household product market statistical data set based on the quantitative value of the product market data.

[0170] In step S310, after obtaining the target product market data set (the product market data in this set is determined according to the conversion rate requirement in step S220), the computer system quantifies the conversion rate of the product market data therein. The purpose of quantification is to convert the conversion rate into a numerical form that can be uniformly calculated and compared. For example, the conversion rate may be a proportional value (such as 0.1 represents a 10% conversion rate), but for the convenience of subsequent calculations, it may be necessary to convert it into other forms of numerical values. One possible quantification method is to map the conversion rate to a new numerical range according to a pre-set quantification standard. For example, a quantization function is set, when the conversion rate is 0, the quantization value is 0; when the conversion rate is 1 (100%), the quantization value is 100. For the intermediate conversion rate, linear interpolation or other functional relationships can be used for quantification. Assume that a linear interpolation function is used, the conversion rate is x, and the quantization value y=100x. If the purchase conversion rate of a product is 0.2 (20%), then according to this quantization function, its quantization value y=100*0.2=20.

[0171] For example, the computer system may store a quantization function or a quantization table. For a linear quantization function, the formula y=kx+b may be used (in the above example, k=100, b=0). The computer system reads the conversion rate of each product market data in the target product market data set, and then substitutes it into the quantization function or searches for the corresponding quantization value in the quantization table. If a nonlinear quantization function is used, such as the quadratic function y=ax²+bx+c, the computer system also substitutes the conversion rate into the function to calculate the quantization value, where a, b, and c are predetermined coefficients.

[0172] In step S320, after obtaining the quantitative values ​​of the product market data in the target product market data set, the computer system needs to determine the conversion characterization value of the household product market statistical data set based on these quantitative values. This conversion characterization value is a value that can represent the conversion status of the entire household product market statistical data set. One possible method is to calculate the average value of the quantitative values. For example, there are n product market data in the target product market data set, and their quantitative values ​​are y1, y2, …, yn respectively. Then the conversion characterization value can be the average value of these quantitative values, that is, the conversion characterization value=(y1+y2+…+yn) / n. Another method may be to calculate the weighted average value based on the weight of the quantitative value. Assuming that different product market data have different weights w1, w2, …, wn according to factors such as their importance or market share, then the conversion characterization value=(w1*y1+w2*y2+…+wn*yn) / (w1+w2+…+wn).

[0173] Figure 2 A hardware entity diagram of a computer system provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the hardware entity of the computer system 1000 includes: a processor 1001 and a memory 1002, wherein the memory 1002 stores a computer program that can be run on the processor 1001, and the processor 1001 implements the steps in the method of any of the above embodiments when executing the program.

Claims

1. A smart traffic diversion service method for the home furnishing industry based on big data, characterized in that: The method comprises: extracting target elements from a household product market statistical data set collected from a target analysis group to obtain a product conversion distribution state of product market data in the household product market statistical data set; based on the product conversion distribution state, determining product market data whose conversion rate reaches a set requirement to obtain a target product market data set; based on the conversion rate of the product market data in the target product market data set, determining a conversion characterization value of the household product market statistical data set; comparing the conversion characterization value of the household product market statistical data set with a market tendency reference value to obtain a market tendency analysis result of the household product market statistical data set; wherein the market tendency reference value is determined by the conversion characterization value of the market statistical data set under different product market tendencies; and performing data drainage and pushing to the target analysis group based on the market tendency analysis result.

2. The method according to claim 1, characterized in that The method also includes: obtaining market statistical data set samples under X market tendency modes, where X≥2; determining the uth conversion characterization value corresponding to the uth market tendency mode based on the market statistical data set samples under the uth market tendency mode, where 1≤u<X; determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value; wherein the uth market tendency reference value is used to compare with the conversion characterization value corresponding to the household product market statistical data set to determine whether the household product market statistical data set belongs to the u+1th market tendency mode or the uth market tendency mode.

3. The method according to claim 2, characterized in that The u-th market tendency pattern includes Y sub-market tendency patterns; the market statistical data set samples under the u-th market tendency pattern include Y sub-market statistical data set samples corresponding to the Y sub-market tendency patterns, wherein Y≥2; the determining of the u-th conversion characterization value corresponding to the u-th market tendency pattern based on the market statistical data set samples under the u-th market tendency pattern includes: determining the sub-conversion characterization values ​​corresponding to the sub-market statistical data set samples under the u-th market tendency pattern respectively, and obtaining Y sub-conversion characterization values; determining the v-th influence parameter corresponding to the Y sub-market tendency patterns under the u-th market tendency pattern respectively, wherein 1≤v<Y; performing a weight adjustment operation on the sub-conversion characterization value corresponding to the v-th sub-market tendency pattern according to the v-th influence parameter; determining the conversion characterization value corresponding to the u-th market tendency pattern based on the sub-conversion characterization value after the weight adjustment operation; the determining of the u-th market tendency reference value based on the u-th conversion characterization value and the u+1-th conversion characterization value includes: The method comprises: determining a u-th numerical range, wherein the maximum value of the u-th numerical range is the conversion characterization value corresponding to the u+1-th market tendency mode, and the minimum value of the u-th numerical range is the conversion characterization value corresponding to the u-th market tendency mode; determining a u-th market tendency reference value based on the u-th numerical range, wherein the u-th market tendency reference value belongs to the u-th numerical range; after determining the u-th market tendency reference value based on the u-th conversion characterization value and the u+1-th conversion characterization value, the method further comprises: obtaining a u-th market tendency mode change upper critical value corresponding to when the u-th market tendency mode is changed to the u+1-th market tendency mode; obtaining a u-th market tendency mode change lower critical value corresponding to when the u+1-th market tendency mode is changed to the u-th market tendency mode; wherein the u-th market tendency mode change upper critical value is greater than the u-th market tendency mode change lower critical value, and the u-th market tendency reference value is greater than the u-th market tendency mode change upper critical value.

4. The method according to claim 3, characterized in that The method of obtaining the u-th market tendency mode change upper critical value corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode comprises: obtaining a plurality of first market statistical data set samples corresponding to the change from the u-th market tendency mode to the u+1-th market tendency mode; obtaining a first sample conversion characterization value corresponding to the first market statistical data set sample; and determining the u-th market tendency mode change upper critical value based on the typical value of the first sample conversion characterization value; the method of obtaining the u-th market tendency mode change lower critical value corresponding to the change from the u+1-th market tendency mode to the u-th market tendency mode comprises: obtaining a plurality of second market statistical data set samples corresponding to the change from the u+1-th market tendency mode to the u-th market tendency mode; obtaining a second sample conversion characterization value corresponding to the second market statistical data set sample; and determining the u-th market tendency mode change lower critical value based on the typical value of the second sample conversion characterization value; wherein, the error between the u-th market tendency mode change upper critical value being greater than the u-th market tendency mode change lower critical value is greater than the set error.

5. The method according to claim 2, characterized in that: The X market tendency patterns include a negative market tendency pattern and a positive market tendency pattern; the market statistical data set samples under the negative market tendency pattern belong to a first market statistical data set set composed of sub-market statistical data set samples corresponding to R negative sub-market tendency patterns, and the market statistical data set samples under the positive market tendency pattern belong to a second market statistical data set set composed of sub-market statistical data set samples corresponding to S positive sub-market tendency patterns, R≥1, S≥1; the determining of the uth conversion representation value corresponding to the uth market tendency pattern based on the market statistical data set samples under the uth market tendency pattern includes: determining the first market R sub-conversion characterization values ​​corresponding to R sub-market statistical data set samples in the statistical data set set, and determining the conversion characterization value corresponding to the negative market tendency pattern based on the R sub-conversion characterization values; determining the u-th market tendency reference value based on the u-th conversion characterization value and the u+1-th conversion characterization value, including: determining a first numerical range, wherein the maximum value of the first numerical range is the conversion characterization value corresponding to the positive market tendency pattern, and the minimum value of the first numerical range is the conversion characterization value corresponding to the negative market tendency pattern; determining positive and negative market tendency reference values ​​based on the first numerical range, wherein the positive and negative market tendency reference values ​​belong to the first numerical range.

6. The method according to claim 5, characterized in that The X market tendency patterns include a negative market tendency pattern and a positive market tendency pattern; after determining the uth market tendency reference value based on the uth conversion characterization value and the u+1th conversion characterization value, the method further includes: obtaining a first market tendency pattern change critical value corresponding to the change from the negative market tendency pattern to the positive market tendency pattern; obtaining a second market tendency pattern change critical value corresponding to the change from the positive market tendency pattern to the negative market tendency pattern; wherein, the first market tendency pattern change critical value is greater than the second market tendency pattern change lower critical value, and the positive-negative market tendency reference value is greater than the first market tendency pattern change upper critical value.

7. The method according to claim 2, characterized in that The X market tendency patterns include a negative market tendency pattern and a positive market tendency pattern, and the u-th market tendency reference value is a positive or negative market tendency reference value; the step of comparing the conversion characterization value of the household product market statistical data set with the market tendency reference value to obtain a market tendency analysis result of the household product market statistical data set includes: comparing the conversion characterization value of the household product market statistical data set with the positive or negative market tendency reference value; if the conversion characterization value of the household product market statistical data set is greater than the positive or negative market tendency reference value, obtaining a market tendency analysis result that the household product market statistical data set belongs to a positive market tendency pattern; if the conversion characterization value of the household product market statistical data set is less than or equal to the positive or negative market tendency reference value, obtaining a market tendency analysis result that the household product market statistical data set belongs to a negative market tendency pattern; in the step of obtaining the X market tendency patterns Before obtaining a sample of a set of market statistical data, the method further includes: determining a market analysis indicator corresponding to the market analysis dimension according to the market analysis dimension; obtaining the sample of a set of market statistical data under X market tendency modes includes: obtaining the sample of a set of market statistical data of the same category under X market tendency modes respectively; after determining the uth market tendency reference value based on the uth conversion representation value and the u+1th conversion representation value, the method further includes: determining a first corresponding relationship between the uth market tendency reference value and the corresponding market analysis dimension; before comparing the conversion representation value of the set of household product market statistical data with the market tendency reference value, the method further includes: determining the current market analysis dimension; based on the current market analysis dimension, obtaining a market tendency reference value having a first corresponding relationship with the current market analysis dimension for comparing with the conversion representation value of the set of household product market statistical data.

8. The method according to claim 7, characterized in that There is a second corresponding relationship between the target market analysis dimension and the target market analysis indicator, wherein the market tendency reference value that has a first corresponding relationship with the target market analysis dimension is determined by a sample of a market statistical data set for the target market analysis indicator; before extracting the target element from the household product market statistical data set collected from the target analysis group, the method also includes: determining the current market analysis dimension; based on the current market analysis dimension, determining the current market analysis indicator that has a second corresponding relationship with the current market analysis dimension; determining whether the household product market statistical data set contains the current market analysis indicator; wherein, when the household product market statistical data set contains the current market analysis indicator, a process of extracting the target element from the household product market statistical data set collected from the target analysis group is performed; when the household product market statistical data set does not contain the current market analysis indicator, the household product market statistical data set is obtained again.

9. The method according to any one of claims 1 to 8, characterized in that The method of determining product market data whose conversion rate reaches set requirements based on the product conversion distribution state and obtaining a target product market data set includes: determining a total numerical range of product market conversion rate data in the household product market statistical data set based on the product conversion distribution state; determining a target numerical interval in the total numerical range, wherein the conversion rate of product market data in the target numerical interval reaches set requirements for product market data, wherein the set requirement is that the conversion rate of product market data is greater than the conversion rate of other product market data in the total numerical range except the target numerical interval, or the set requirement is that the conversion rate of product market data is less than the conversion rate of other product market data in the total numerical range except the target numerical interval; the method of determining a conversion representation value of the household product market statistical data set based on the conversion rate of product market data in the target product market data set includes: quantifying the conversion rate of product market data in the target product market data set to obtain a quantified value of the product market data; and determining a conversion representation value of the household product market statistical data set based on the quantified value of the product market data.

10. A computer system comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, wherein: When the processor executes the program, the steps in the method according to any one of claims 1 to 9 are implemented.

Images