Commercial data intelligent analysis method and system based on data driving

Abstract

The invention discloses a business data intelligent analysis method and system based on data driving, and relates to the field of business data analysis. The business data intelligent analysis method based on data driving comprises the following steps: acquiring and preprocessing commodity purchase data, commodity sales data and commodity market data; analyzing the preprocessed commodity purchase data based on an association rule learning algorithm to obtain a commodity association degree matrix; according to the method, the association rule learning algorithm is applied to the commodity purchase data, so that the hidden relation between different commodities can be revealed, merchants are helped to optimize product layout and sales promotion strategies, the opportunity and revenue of cross sales are improved, and the commodity demand prediction efficiency is improved. Through deep demand prediction of sales data, merchants can predict future inventory demands more accurately, occurrence of excessive inventory and stockout conditions is avoided, the operation cost is reduced, and the resource utilization efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of business data analysis, and in particular to a data-driven business data intelligent analysis method and system. Background Art

[0002] In today's business environment, data-driven decision-making has become the key to improving corporate competitiveness. By leveraging big data and advanced analytical tools, companies can extract valuable information from massive amounts of data to optimize business processes, improve customer satisfaction and increase revenue. The concept of business intelligence has undergone significant evolution over the past few decades. From the initial reports and dashboards to today's predictive analysis and machine learning, BI technology enables companies to gain in-depth insights in real time, allowing them to make faster and more accurate decisions. With the advancement of technology, analytical methods including artificial intelligence and machine learning have become increasingly advanced. These technologies are able to process and analyze large-scale data sets, providing companies with unprecedented analytical depth. Although data-driven analysis provides many opportunities, companies also face a series of challenges during implementation, such as data quality, data privacy protection, and how to integrate data from different sources. Therefore, a data-driven business data intelligence analysis method and system has emerged.

[0003] Prior art, such as the invention patent application with announcement number: CN115577991B, discloses a business intelligence data analysis system and analysis method based on big data. The present invention can first calculate the correlation between the marketing volume of related products and the marketing volume of front-end products, and then obtain the evaluation value of the related products based on the correlation, and then evaluate it in combination with the actual marketing volume of the related products. By setting multiple evaluation cycles, it can avoid the impact of short-term marketing volume on the marketing evaluation report, truly reflect the marketing information of related products within the marketing case cycle, and effectively provide decision makers with a systematic and comprehensive evaluation report, so that decision makers can get effective data support when making decisions.

[0004] Based on the above solution, it is found that the limitations of the existing technology include at least the following problems. First, the traditional business intelligence system conducts a comprehensive analysis of all product marketing information when applied, which may incorporate short-term market fluctuations into the long-term analysis, such as a sudden increase in sales during holidays, which may lead to misjudgment of decision-making. Second, the traditional business intelligence system fails to fully evaluate the marketing of other products related to the main products, which may lead to misunderstanding of market demand, thereby affecting inventory management and product promotion strategies. Finally, the traditional business intelligence system lacks sufficient flexibility and precision when processing data from different marketing cycles, which may lead to inaccurate evaluation results. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a data-driven business data intelligent analysis method and system, which solves the problem that traditional business intelligence systems conduct a comprehensive analysis of all commodity marketing information when applied, which may incorporate short-term market fluctuations into long-term analysis, such as a sudden increase in sales during holidays, which may lead to misjudgment of decisions; secondly, traditional business intelligence systems fail to fully evaluate the marketing of other commodities associated with the main commodities, which may lead to misunderstanding of market demand, thereby affecting inventory management and product promotion strategies; finally, traditional business intelligence systems lack sufficient flexibility and accuracy when processing data from different marketing cycles, which may lead to inaccurate evaluation results.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a data-driven commercial data intelligent analysis method, comprising the following steps: obtaining commodity purchase data, commodity sales data, and commodity market data and preprocessing them; analyzing the preprocessed commodity purchase data based on an association rule learning algorithm to obtain a commodity association matrix; analyzing and predicting the preprocessed commodity sales data to obtain commodity demand forecast data; conducting a comprehensive analysis based on the commodity demand forecast data and the preprocessed commodity market data to obtain commodity evaluation cycle data; integrating and organizing the commodity association matrix, commodity demand forecast data, and commodity evaluation cycle data, and displaying them.

[0007] Furthermore, the commodity purchase data specifically refers to the type information of the commodities in each purchase order, the commodity sales data specifically refers to the sales date, historical sales quantity and historical return quantity of each commodity, the commodity market data includes commodity sales date data, consumer behavior data and social media data, the commodity sales date data specifically refers to the sales date of each commodity, the consumer behavior data specifically refers to the purchase path and shopping cart data of each customer, the social media data specifically refers to the comment information of each commodity on social media, the commodity demand forecast data specifically refers to the quantity demand value of each commodity in a set future period, and the commodity evaluation cycle data specifically refers to the optimal evaluation cycle of each commodity.

[0008] Furthermore, the preprocessed commodity purchase data is analyzed based on the association rule learning algorithm to obtain the commodity association matrix. Specifically, the category information of the commodities in each purchase order after preprocessing is read; a commodity combination set is generated based on the Apriori algorithm, and the support and confidence of each commodity combination are analyzed respectively; the commodity combination set is screened and analyzed based on the preset minimum support, and all commodity combinations that meet the minimum support are obtained, and marked as a frequent commodity combination set; the frequent commodity combination set is screened based on the preset minimum confidence, and all commodity combinations that meet the minimum confidence are obtained, and a list of commodity association rules is output; a commodity association matrix is ​​established and initialized; the commodity association matrix is ​​filled based on the commodity association rule list; wherein, the calculation formula for the support of each commodity is as follows: ;in, For the The support of the product, is the total number of product combinations in the product combination set, The first item in all the product combinations in the product combination set appears The total number of product combinations of the product. , The number of product types in the product combination set.

[0009] Furthermore, the specific formula for calculating the confidence of a product is as follows: ;in, is an association rule, indicating that if a customer purchases a product , then they may also buy the product , For goods and products , indicating that the product is included and products Combination of goods, Indicates that you purchased the product at the same time and products The support of the product combination, For goods support.

[0010] Furthermore, the specific steps to obtain commodity demand forecast data are as follows: read the pre-processed sales date and historical sales quantity and historical return quantity of each commodity; analyze the sales time of each commodity based on the pre-processed sales date of each commodity; analyze based on the pre-processed historical sales quantity and historical return quantity of each commodity to obtain the actual sales quantity within the sales time of each commodity; conduct a comprehensive analysis based on the set future cycle, the sales time of each commodity and the actual sales quantity of each commodity within the sales time to obtain the quantity demand value of each commodity within the set future cycle.

[0011] Furthermore, the specific formula for calculating the quantity demand value of each commodity in the set future period is as follows: ;in, For the first The quantity demand value of a commodity, For the The sales time of the product, For the The historical sales volume of the product. For the The historical number of returns for this product. For the future period, , The number of product types.

[0012] Furthermore, a comprehensive analysis is performed based on the commodity demand forecast data and the pre-processed commodity market data to obtain the commodity evaluation cycle data, specifically: reading the historical sales quantity, historical return quantity, quantity demand value of each commodity in a set future period, each customer's purchase path, shopping cart data, and comment information of each commodity on social media for each commodity; performing demand identification analysis on the historical sales quantity, historical return quantity, and quantity demand value of each commodity in a set future period for each commodity based on the time series analysis method to obtain the seasonality and long-term trend of demand for each commodity; performing a comprehensive analysis on each customer's purchase path, shopping cart data, and comment information of each commodity on social media to obtain the market's reaction time for each commodity; determining the optimal evaluation cycle for each commodity based on the seasonality and long-term trend of demand for each commodity and the market's reaction time for each commodity.

[0013] Furthermore, the commodity correlation matrix, commodity demand forecast data, and commodity evaluation cycle data are integrated and organized, and then displayed as follows: a data table is established for each commodity, and each row in the data table corresponds to a commodity; the data table is filled based on the commodity correlation matrix, the quantity demand value of each commodity in the set future period, and the optimal evaluation cycle for each commodity; and the filled data table is displayed based on Tableau.

[0014] A data-driven business data intelligent analysis system comprises: a data collection and preprocessing subsystem, a data analysis subsystem, a data visualization subsystem, and a data management subsystem; the data collection and preprocessing subsystem is used to obtain commodity purchase data, commodity sales data, and commodity market data and perform preprocessing; the data analysis subsystem is used to analyze the preprocessed commodity purchase data based on an association rule learning algorithm to obtain a commodity association matrix, and to analyze and predict commodity sales data to obtain commodity demand forecast data, and to perform a comprehensive analysis based on the commodity demand forecast data and commodity market data to obtain commodity evaluation cycle data; the data visualization subsystem is used to integrate and organize based on the commodity association matrix, commodity demand forecast data, and commodity evaluation cycle data, and to display them; the data management subsystem is used to store and back up the commodity purchase data, commodity sales data, commodity market data, the commodity association matrix obtained by analyzing the commodity purchase data, the commodity demand forecast data obtained by analyzing the commodity sales data, and the commodity evaluation cycle data obtained by analyzing the commodity demand forecast data and the commodity market data.

[0015] Furthermore, the data analysis subsystem includes an association rule analysis module, a demand forecasting module, a trend analysis module, and an evaluation cycle determination module; the association rule analysis module is used to analyze commodity purchase data using the Apriori algorithm to generate a commodity association matrix; the demand forecasting module is used to predict commodity demand in future cycles based on historical sales and return data in combination with time series analysis technology; the trend analysis module is used to analyze historical sales data and market data to identify the seasonality and long-term trends of commodity demand; the evaluation cycle determination module is used to determine the optimal evaluation cycle of a commodity based on the demand trend and market response time of the commodity.

[0016] The present invention has the following beneficial effects: (1) This data-driven intelligent analysis method for commercial data can reveal the hidden relationships between different commodities by applying association rule learning algorithms to commodity purchase data, which helps merchants better understand consumers' purchasing behavior and optimize product layout and promotion strategies, thereby increasing cross-selling opportunities and revenue. At the same time, by conducting in-depth demand forecasts on sales data, merchants can more accurately predict future inventory needs, avoid over-inventory and out-of-stock situations, reduce operating costs, and improve resource utilization efficiency.

[0017] (2) This data-driven business data intelligent analysis method, by evaluating the optimal evaluation cycle of commodities and combining seasonal and long-term trend analysis, enables enterprises to respond quickly to market changes and respond to the market at the right time. This not only enhances the enterprise's ability to adapt to market dynamics, but also improves customer satisfaction and brand loyalty by meeting consumer needs on time.

[0018] (3) This data-driven business data intelligent analysis system can not only reveal the potential patterns of consumer purchasing behavior through the association rule analysis module, but also identify changes in market demand in advance through the demand forecasting module and trend analysis module, so that enterprises can quickly adjust market strategies, optimize inventory management, and dynamically adjust product supply according to market demand, thereby improving the speed and accuracy of responding to market changes.

[0019] (4) This data-driven business data intelligent analysis system, with the strong support of the data management subsystem, enables enterprises to ensure the security and privacy of all analysis data, while ensuring real-time updating and high availability of data. In addition, with the addition of the data visualization subsystem, complex data analysis results can be presented in the form of intuitive charts and reports, so that non-technical decision makers can also easily understand and use these analysis results, thereby greatly improving the scientific nature and accuracy of decision-making.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of a data-driven business data intelligent analysis method of the present invention.

[0022] Figure 2 The present invention is a flowchart of the specific steps of obtaining a commodity association matrix in a data-driven commercial data intelligent analysis method.

[0023] Figure 3 This is a block diagram of a data-driven business data intelligent analysis system of the present invention. DETAILED DESCRIPTION

[0024] The embodiment of the present application solves the problem that traditional business intelligence systems conduct a comprehensive analysis of all commodity marketing information when applied, such as a sudden increase in holiday sales, which may lead to misjudgment of decisions. Secondly, traditional business intelligence systems fail to fully evaluate the marketing of other commodities related to the main commodities, which may lead to misunderstanding of market demand and affect inventory management and product promotion strategies. Finally, traditional business intelligence systems lack sufficient flexibility and precision when processing data from different marketing cycles, which may lead to inaccurate evaluation results.

[0025] The overall idea of ​​the problem in the embodiment of this application is as follows: First, obtain necessary product data from different sources, including purchase data, sales data, and market data, perform data preprocessing, ensure data quality, and lay a solid foundation for subsequent analysis. Secondly, use association rule learning algorithms to process the preprocessed product purchase data, such as the Apriori algorithm, to generate a product association matrix. At the same time, perform demand forecasts on product sales data, and use statistical methods and models to analyze product sales trends and demand. Finally, conduct a comprehensive analysis of demand forecast results and market data, evaluate the product evaluation cycle, integrate the product association matrix, demand forecast data, and evaluation cycle data, and display them through the advanced visualization tool Tableau to support decision making.

[0026] See also Figure 1 The embodiment of the present invention provides a technical solution: a data-driven intelligent analysis method for commercial data, comprising the following steps: obtaining commodity purchase data, commodity sales data, and commodity market data and preprocessing them; analyzing the preprocessed commodity purchase data based on an association rule learning algorithm to obtain a commodity association matrix; analyzing and predicting the preprocessed commodity sales data to obtain commodity demand forecast data; performing a comprehensive analysis based on the commodity demand forecast data and the preprocessed commodity market data to obtain commodity evaluation cycle data; integrating and organizing the commodity association matrix, the commodity demand forecast data, and the commodity evaluation cycle data, and displaying them.

[0027] The commodity purchase data specifically refers to the type information of the commodities in each purchase order; the commodity sales data specifically refers to the sales date, historical sales quantity and historical return quantity of each commodity; the commodity market data includes commodity sales date data, consumer behavior data and social media data; the commodity sales date data specifically refers to the sales date of each commodity; the consumer behavior data specifically refers to the purchase path and shopping cart data of each customer; the social media data specifically refers to the comment information of each commodity on social media; the commodity demand forecast data specifically refers to the quantity demand value of each commodity in a set future period; the commodity evaluation cycle data specifically refers to the optimal evaluation cycle of each commodity.

[0028] Specifically, Figure 2 As shown in FIG, the commodity purchase data after preprocessing is analyzed based on the association rule learning algorithm to obtain the commodity association matrix. Specifically, the commodity category information in each purchase order after preprocessing is read; a commodity combination set is generated based on the Apriori algorithm, and the support and confidence of each commodity combination are analyzed respectively; the commodity combination set is screened and analyzed based on the preset minimum support, and all commodity combinations that meet the minimum support are obtained, and are marked as commodity combination frequent sets; the commodity combination frequent sets are screened based on the preset minimum confidence, and all commodity combinations that meet the minimum confidence are obtained, and a commodity association rule list is output, and each rule contains two parts: the rule itself, such as commodity X⇒commodity Y and the support and confidence of the rule; a commodity association matrix is ​​established and initialized, that is, the rows and columns of the matrix correspond to commodities; the commodity association matrix is ​​filled based on the commodity association rule list, in which each element, such as Display products and products The correlation degree, i.e., confidence degree, is calculated by the following formula: ;in, For the The support of the product, is the total number of product combinations in the product combination set, The first item in all the product combinations in the product combination set appears The total number of product combinations of the product. , The number of product types in the product combination set.

[0029] The specific formula for calculating the confidence of a product is as follows: ;in, is an association rule, indicating that if a customer purchases a product , then they may also buy the product , For goods and products , indicating that the product is included and products Combination of goods, Indicates that you purchased the product at the same time and products The support of the product combination, For goods support.

[0030] For example: It is known to be 0.5, If it is known to be 0.75, then the confidence The calculation is: , that is, the confidence level is about 0.67.

[0031] It is known to be 0.2, If it is known to be 0.5, then the confidence The calculation is: , that is, the confidence level is 0.4.

[0032] It is known to be 0.1, If it is known to be 0.25, then the confidence The calculation is: , that is, the confidence level is 0.4.

[0033] It is known to be 0.45, If it is known to be 0.9, then the confidence The calculation is: , that is, the confidence level is 0.5.

[0034] In this implementation scheme, the commodity category information in each purchase order is read, and necessary data cleaning and formatting are performed. This step ensures data quality, provides clean and consistent data input for accurate association analysis, improves data processing efficiency, and reduces the interference of erroneous or duplicate data on analysis results. The Apriori algorithm is applied to generate possible commodity combination sets, and the support of each combination is calculated. This step helps merchants understand which commodity combinations are most popular by identifying frequently purchased commodity combinations. Support analysis helps to confirm which commodity combinations are important and which can be ignored, thereby optimizing inventory and marketing strategies. Based on the preset minimum support, the qualified commodity combination frequent set is screened, and then the commodity combination that meets the minimum confidence is calculated. This step enables merchants to focus on those commodities that have a high probability of being purchased together to promote cross-selling and promotional activities. Confidence analysis further refines the support results and provides more operational market strategy suggestions. A commodity association matrix is ​​established, whose rows and columns correspond to different commodities, and then the matrix is ​​filled according to the commodity association rule list. This step provides an intuitive way to view the association strength between different commodities, which is convenient for quickly capturing market dynamics and consumer preferences. The association matrix becomes a powerful analysis tool to help merchants formulate specific action plans based on data.

[0035] Specifically, the specific steps to obtain commodity demand forecast data are as follows: read the pre-processed sales date and historical sales quantity and historical return quantity of each commodity; analyze the sales time of each commodity based on the pre-processed sales date of each commodity, and use days as the counting unit; analyze based on the pre-processed historical sales quantity and historical return quantity of each commodity, and obtain the actual sales quantity within the sales time of each commodity; conduct a comprehensive analysis based on the set future cycle, the sales time of each commodity, and the actual sales quantity within the sales time of each commodity, and obtain the quantity demand value of each commodity within the set future cycle, wherein the set future cycle is counted in days.

[0036] The specific formula for calculating the quantity demand value of each commodity in the set future period is as follows: ;in, For the first The quantity demand value of a commodity, For the The sales time of the product, measured in days. For the The historical sales volume of the product. For the The historical number of returns for this product. This item indicates the The actual sales quantity of the product during the sales period. For the set future cycle, the counting unit is day. , The number of product types.

[0037] The calculation data of the quantity demand value of each commodity in the set future period is shown in the following table: Table 1 Example of calculation data for the quantity demand value of each commodity in a given future period

[0038] In Table 1, For the The sales time of the product, with days as the unit of measurement, For the The historical sales volume of the product. For the The historical number of returns for this product. is the set future period, and the counting unit is day. For the first The quantity demanded of a commodity.

[0039] The first set of data: sales duration of product A: 10; historical sales quantity of product A: 100; historical return quantity of product A: 50; set future cycle: 2; quantity demand value of product A within the set future cycle: 10.

[0040] The second set of data: Sales duration of Product A: 5; Historical sales quantity of Product A: 125; Historical return quantity of Product A: 25; Set future cycle: 2; Quantity demand value of Product A within the set future cycle: 40.

[0041] The third set of data: sales duration of product A: 20; historical sales quantity of product A: 180; historical return quantity of product A: 160; set future cycle: 2; quantity demand value of product A within the set future cycle: 2.

[0042] The fourth set of data: sales duration of product A: 15; historical sales quantity of product A: 165; historical return quantity of product A: 105; set future cycle: 2; quantity demand value of product A within the set future cycle: 8.

[0043] The fifth set of data: sales duration of Product A: 8; historical sales quantity of Product A: 192; historical return quantity of Product A: 80; set future cycle: 2; quantity demand value of Product A within the set future cycle: 28.

[0044] In this implementation scheme, the sales date, historical sales quantity and historical return quantity of each commodity are extracted from the database. This step ensures that the basic data for analysis is complete and accurate, and provides necessary historical references for accurate prediction. The sales duration of the commodity is calculated based on the first and last sales dates of the commodity. This step understands the market duration of each commodity and helps analyze the life cycle and market performance of the commodity. The actual sales quantity is obtained by subtracting the return quantity from the historical sales quantity. This step provides accurate sales data to help enterprises evaluate the real market acceptance of the commodity. According to the sales duration and actual sales quantity of the commodity, combined with the future sales cycle, future demand is predicted. This step enables enterprises to formulate more accurate production and inventory strategies based on the predicted data, reduce capital backlog and out-of-stock risks, and calculate the specific formula for the quantity demand value of each commodity in the set future cycle. By combining historical sales data with future cycles, this formula provides enterprises with quantitative demand forecasts, helping enterprises make more scientific decisions in supply chain management. The forecast results enable enterprises to optimize production plans and inventory levels according to expected demand, improve resource utilization efficiency, and reduce waste. Accurate demand forecasts enable enterprises to respond quickly to market changes, especially during seasonal peaks with large demand fluctuations, to ensure that product supply is synchronized with market demand.

[0045] Specifically, based on the commodity demand forecast data and the pre-processed commodity market data, a comprehensive analysis is performed to obtain the commodity evaluation cycle data, which is specifically: reading the historical sales quantity, historical return quantity, quantity demand value of each commodity in the set future period, each customer's purchase path, shopping cart data, and comment information of each commodity on social media for each commodity; based on the time series analysis method, a demand identification analysis is performed on the historical sales quantity, historical return quantity, and quantity demand value of each commodity in the set future period for each commodity to obtain the seasonality and long-term trend of the demand for each commodity; a comprehensive analysis is performed on each customer's purchase path, shopping cart data, and comment information of each commodity on social media to obtain the market response to each commodity. The response time, such as the time from product release to consumers starting to actively discuss or increase purchases on social media, can reflect the market's response speed; determine the optimal evaluation cycle for each product based on the seasonal and long-term trends in demand for each product and the market's response time for each product. If the demand trend for a product shows significant seasonal fluctuations, a shorter evaluation cycle may be required, such as quarterly evaluation, so that marketing strategies and inventory can be adjusted in a timely manner. For products with fast market response speed, especially fashion or technology products, a shorter evaluation cycle is also required due to rapid changes in consumer preferences and technology. For products with stable demand and slow market response, a longer evaluation cycle can be set, such as evaluation every six months or every year. This is suitable for more traditional or stable demand products, such as basic daily necessities.

[0046] In this implementation plan, historical sales volume, return volume, future demand forecast, customer purchase path, shopping cart data and social media comments of each product are collected from different data sources. This step provides a comprehensive data view, supports in-depth analysis of the market performance and consumer behavior of each product, and conducts time series analysis on the sales data of each product to identify the seasonality and long-term trends of demand. This step enables enterprises to predict changes in product demand, arrange production and inventory management reasonably, and reduce excess or out-of-stock situations. It also analyzes the length of time from product release to positive market response, such as increased purchases and active social media discussions. This step understands the market's response speed to new products or marketing activities and helps enterprises adjust Marketing strategies are used to improve efficiency. The evaluation cycle of each commodity is determined based on the seasonality of commodity demand, long-term trends, and market response speed. This step enables companies to develop more effective evaluation and adjustment strategies for different categories of commodities, thereby optimizing resource allocation and market response. The steps to obtain commodity evaluation cycle data are as follows: through scientific data analysis methods, companies can flexibly adjust the evaluation cycle of commodities, respond to market changes in a timely manner, and reduce market opportunity losses or resource waste caused by fixed evaluation cycles. According to commodity demand fluctuations and market response times, companies can more accurately adjust marketing and inventory strategies for specific commodity groups, thereby improving the effectiveness of market strategies and reducing unnecessary marketing and storage expenses.

[0047] Specifically, the commodity association matrix, commodity demand forecast data, and commodity evaluation cycle data are integrated and organized, and then displayed as follows: a data table is established for each commodity, and each row in the data table corresponds to a commodity; and each row of the data table has a commodity name column: the identification name or code of the commodity, a demand forecast column: displaying the predicted value of the demand for the commodity in a certain period of time in the future, a support column, which contains several rows of small columns, each of which represents the frequency of associated purchases of the commodity with other commodities, a confidence column: the probability of purchasing this commodity when other commodities associated with this commodity are purchased, a trend analysis column: displaying the seasonality and long-term trend of the demand for each commodity, and the market's reaction time for each commodity, an evaluation cycle column: displaying the optimal evaluation cycle for each commodity; the data table is filled based on the commodity association matrix, the quantity demand value of each commodity in the set future period, and the optimal evaluation cycle for each commodity; and the filled data table is displayed based on Tableau.

[0048] In this implementation plan, a data row is created for each product, including multiple fields such as product name, demand forecast, support, confidence, trend analysis and evaluation cycle. This step can help management and analysts quickly grasp key information and facilitate tracking and comparison of the performance of different products. The product association matrix and demand forecast results obtained based on historical data analysis are input into the table. This step provides a comprehensive perspective that focuses on the market performance and expected demand of each product, which helps to optimize inventory and promotion strategies. Seasonal and long-term trend analysis data and market response time are added to each product in the data table. This step enables decision makers to understand market dynamics and adjust marketing activities to match consumer purchasing behavior and response. Tableau software converts the processed data tables into graphics and dashboards for intuitive display. This step uses visualization tools to simplify complex data into easy-to-understand charts to help make business decisions quickly. The step of integrating and sorting out the commodity correlation matrix, commodity demand forecast data, and commodity evaluation cycle data integrates key business indicators into a single view, such as demand forecast, support, confidence, etc., which strengthens the application of data in the decision-making process, overcomes the problems of decentralized data processing and information islands, improves decision-making efficiency and accuracy, and comprehensively analyzes commodity demand trends and market reactions, enabling companies to quickly adapt to market changes and adjust strategies in a timely manner. It solves the problem of slow response to market changes in the traditional decision-making process and improves the company's market competitiveness.

[0049] See also Figure 3 The embodiment of the present invention provides a technical solution: a data-driven business data intelligent analysis system, including: a data collection and preprocessing subsystem, a data analysis subsystem, a data visualization subsystem, and a data management subsystem; the data collection and preprocessing subsystem is used to obtain commodity purchase data, commodity sales data, and commodity market data and perform preprocessing; the data analysis subsystem is used to analyze the preprocessed commodity purchase data based on an association rule learning algorithm to obtain a commodity association matrix, and analyze and predict the commodity sales data to obtain commodity demand forecast data, and at the same time, perform a comprehensive analysis based on the commodity demand forecast data and the commodity market data to obtain commodity evaluation cycle data; the data visualization subsystem is used to integrate and organize based on the commodity association matrix, the commodity demand forecast data, and the commodity evaluation cycle data, and display them; the data management subsystem is used to store and back up the commodity purchase data, commodity sales data, commodity market data, the commodity association matrix obtained by analyzing the commodity purchase data, the commodity demand forecast data obtained by analyzing the commodity sales data, and the commodity evaluation cycle data obtained by analyzing the commodity demand forecast data and the commodity market data.

[0050] Specifically, the data analysis subsystem includes an association rule analysis module, a demand forecasting module, a trend analysis module, and an evaluation cycle determination module; the association rule analysis module is used to analyze commodity purchase data using the Apriori algorithm to generate a commodity association matrix; the demand forecasting module is used to predict commodity demand in future cycles based on historical sales and return data combined with time series analysis technology; the trend analysis module is used to analyze historical sales data and market data to identify the seasonality and long-term trends of commodity demand; the evaluation cycle determination module is used to determine the optimal evaluation cycle of a commodity based on the commodity demand trend and market response time.

[0051] In this implementation scheme, the association rule analysis module extracts all commodity purchase data from the database, processes the data using the Apriori algorithm, searches for frequent item sets and high-confidence association rules, and generates a commodity association matrix, where each element represents the probability of two commodities being purchased at the same time. This module provides a means to gain in-depth understanding of consumer purchasing behavior, and can help companies identify which commodities are often purchased together. By discovering potential associations between commodities, it helps companies develop effective marketing strategies, such as bundling sales and cross-promotion, to increase sales and customer satisfaction. The demand forecasting module collects historical sales data and return data, and uses time series analysis methods, such as the ARIMA model, to forecast future sales of commodities, and generates a demand forecasting report that indicates the expected sales volume of each commodity in the future cycle. This module enables companies to more accurately forecast future inventory demand, avoid oversupply or out-of-stock situations, and improve the efficiency of inventory management. The module improves efficiency and accuracy, reduces operating costs, and ensures that market demand can be met and customer satisfaction is enhanced. The trend analysis module analyzes and organizes historical sales data and market data, identifies seasonal fluctuations and long-term trends in commodity demand, and outputs trend reports for decision makers' reference. This module enables enterprises to understand the cyclical changes in commodity demand and provides timing guidance for product promotion and inventory management. By gaining insight into market trends, enterprises can optimize product launch time and marketing activities, and improve market response speed and flexibility. The evaluation cycle determination module analyzes commodity demand trends and market response time data. Based on the analysis results, a reasonable evaluation cycle is set for each commodity, and an evaluation cycle strategy report is generated. This module allows enterprises to dynamically adjust the evaluation cycle according to market changes, making product evaluation more timely and effective. For a rapidly changing market environment, this module helps enterprises maintain their competitiveness and better meet consumer needs through timely product iterations and market strategy adjustments.

[0052] In summary, this application has at least the following effects: By applying association rule learning algorithms to product purchase data, the hidden relationships between different products can be revealed, which helps merchants better understand consumer purchasing behavior and optimize product layout and promotion strategies, thereby increasing cross-selling opportunities and revenue. At the same time, through in-depth demand forecasting of sales data, merchants can more accurately predict future inventory needs, avoid over-inventory and out-of-stock situations, reduce operating costs, and improve resource utilization efficiency.

[0053] By evaluating the optimal evaluation cycle of goods and combining seasonal and long-term trend analysis, companies can respond quickly to market changes and react to the market at the right time. This not only enhances the company's ability to adapt to market dynamics, but also improves customer satisfaction and brand loyalty by meeting consumer needs on time.

[0054] The association rule analysis module can not only reveal the potential patterns of consumer purchasing behavior, but also identify changes in market demand in advance through the demand forecasting module and trend analysis module, so that companies can quickly adjust market strategies, optimize inventory management, and dynamically adjust product supply according to market demand, thereby improving the speed and accuracy of responding to market changes.

[0055] With the powerful support of the data management subsystem, enterprises can ensure the security and privacy of all analytical data while ensuring real-time updating and high availability of data. In addition, with the addition of the data visualization subsystem, complex data analysis results can be presented in the form of intuitive charts and reports, allowing non-technical decision makers to easily understand and use these analysis results, thereby greatly improving the scientific nature and accuracy of decision-making.

[0056] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0057] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A data-driven business data intelligence analysis method. It is characterized in that The following steps are involved: Obtain commodity purchase data, commodity sales data, commodity market data and perform pre-processing; The pre-processed commodity purchase data is analyzed based on the association rule learning algorithm to obtain the commodity association matrix; Analyze and forecast the pre-processed commodity sales data to obtain commodity demand forecast data; Based on the commodity demand forecast data and pre-processed commodity market data, a comprehensive analysis is performed to obtain commodity evaluation cycle data; The product correlation matrix, product demand forecast data, and product evaluation cycle data are integrated and organized for display.

2. The data-driven business data intelligent analysis method according to claim 1, It is characterized in that The commodity purchase data specifically refers to the type information of the commodities in each purchase order; the commodity sales data specifically refers to the sales date, historical sales quantity and historical return quantity of each commodity; the commodity market data includes commodity sales date data, consumer behavior data and social media data; the commodity sales date data specifically refers to the sales date of each commodity; the consumer behavior data specifically refers to the purchase path and shopping cart data of each customer; the social media data specifically refers to the comment information of each commodity on social media; the commodity demand forecast data specifically refers to the quantity demand value of each commodity in a set future period; and the commodity evaluation cycle data specifically refers to the optimal evaluation cycle of each commodity.

3. The data-driven business data intelligent analysis method according to claim 2, It is characterized in that Based on the association rule learning algorithm, the preprocessed commodity purchase data is analyzed to obtain the commodity association matrix: Read the category information of the goods in each purchase order after preprocessing; Generate a set of product combinations based on the Apriori algorithm, and analyze the support and confidence of each product combination; Based on the preset minimum support, the product combination set is screened and analyzed to obtain all product combinations that meet the minimum support and mark them as frequent product combination sets; Based on the preset minimum confidence, the frequent set of product combinations is screened to obtain all product combinations that meet the minimum confidence, and a list of product association rules is output; Establish a commodity correlation matrix and perform initialization processing; Filling the commodity association matrix based on the commodity association rule list; The calculation formula for the support of each product is as follows: ； in, For the The support of the product, is the total number of product combinations in the product combination set, The first item in all the product combinations in the product combination set appears The total number of product combinations of the product. , The number of product types in the product combination set.

4. The data-driven business data intelligent analysis method according to claim 3, It is characterized in that The specific formula for calculating the confidence of a product is as follows: ； in, is an association rule, indicating that if a customer purchases a product , then they may also buy the product , For goods and products , indicating that the product is included and products Combination of goods, Indicates that you purchased the product at the same time and products The support of the product combination, For goods support.

5. The data-driven business data intelligent analysis method according to claim 2, It is characterized in that The specific steps to obtain commodity demand forecast data are as follows: Read the pre-processed sales date, historical sales quantity, and historical return quantity of each product; Analyze the sales duration of each product based on the pre-processed sales date of each product; Based on the pre-processed historical sales quantity and historical return quantity of each commodity, the actual sales quantity of each commodity within the sales period is obtained; A comprehensive analysis is performed based on the set future cycle, the sales duration of each commodity, and the actual sales quantity of each commodity within the sales duration to obtain the quantity demand value of each commodity in the set future cycle.

6. The data-driven business data intelligent analysis method according to claim 5, It is characterized in that The specific formula for calculating the quantity demand value of each commodity in the set future period is as follows: ； in, For the first The quantity demand value of a commodity, For the The sales time of the product, For the The historical sales volume of the product. For the The historical number of returns for this product. For the future period, , The number of product types.

7. The data-driven business data intelligent analysis method according to claim 2, It is characterized in that Based on the comprehensive analysis of commodity demand forecast data and pre-processed commodity market data, the commodity evaluation cycle data is obtained as follows: Read the historical sales volume and return volume of each product, the quantity demand value of each product in the set future period, the purchase path of each customer, shopping cart data, and the comments of each product on social media; Based on the time series analysis method, the historical sales volume, historical return volume, and quantity demand value of each commodity in the set future period are analyzed to identify the demand, and the seasonality and long-term trend of each commodity demand are obtained; Comprehensively analyze each customer's purchase path, shopping cart data, and comments on each product on social media to obtain the market's reaction time for each product; Determine the optimal evaluation period for each commodity based on the seasonal and long-term trends in demand for each commodity and the market's response time to each commodity.

8. The data-driven business data intelligent analysis method according to claim 2, It is characterized in that Based on the commodity correlation matrix, commodity demand forecast data, and commodity evaluation cycle data, the data is integrated and organized, and displayed as follows: Create a data table for each product, and each row in the data table corresponds to a product; Fill the data table based on the commodity association matrix, the quantity demand value of each commodity in the set future period, and the optimal evaluation period for each commodity; The filled and processed data table is displayed based on Tableau.

9. A data-driven business data intelligent analysis system, applying the data-driven business data intelligent analysis method according to any one of claims 1 to 8, It is characterized in that include: Data collection and preprocessing subsystem, data analysis subsystem, data visualization subsystem, data management subsystem; The data collection and preprocessing subsystem is used to obtain commodity purchase data, commodity sales data, commodity market data and perform preprocessing; The data analysis subsystem is used to analyze the pre-processed commodity purchase data based on the association rule learning algorithm to obtain the commodity association matrix, and to analyze and predict the commodity sales data to obtain commodity demand forecast data, and to perform a comprehensive analysis based on the commodity demand forecast data and commodity market data to obtain commodity evaluation cycle data; The data visualization subsystem is used to integrate and organize the commodity association matrix, commodity demand forecast data, and commodity evaluation cycle data, and to display them; The data management subsystem is used to store and back up commodity purchase data, commodity sales data, commodity market data, commodity association matrix obtained by analyzing commodity purchase data, commodity demand forecast data obtained by analyzing commodity sales data, and commodity evaluation cycle data obtained by analyzing commodity demand forecast data and commodity market data.

10. The data-driven business data intelligent analysis system according to claim 9, It is characterized in that The data analysis subsystem includes an association rule analysis module, a demand forecasting module, a trend analysis module, and an evaluation cycle determination module; The association rule analysis module is used to analyze the commodity purchase data using the Apriori algorithm to generate a commodity association matrix; The demand forecasting module is used to forecast the demand for goods in future periods based on historical sales and return data combined with time series analysis technology; The trend analysis module is used to analyze historical sales data and market data to identify seasonality and long-term trends in commodity demand; The evaluation cycle determination module is used to determine the optimal evaluation cycle of the commodity based on the demand trend and market response time of the commodity.

Images