Business administration decision support method and system based on artificial intelligence

Abstract

The invention discloses an artificial intelligence-based business administration decision support method and system. The method comprises the steps of obtaining to-be-analyzed digital economic information of a preset enterprise; analyzing the digital economic information, and obtaining consumer behavior data, market trend data and industry development data; inputting the consumer behavior data into a preset deep learning model, and outputting consumer demand change information; inputting the market trend data and the industry development data into a preset time sequence analysis model, and outputting market trend information; inputting the consumer behavior data, the market trend data and the industry development data into a preset multi-factor model, and outputting risk assessment information corresponding to a preset enterprise; generating management decision information of the preset enterprise according to the consumer demand change information, the market trend information and the risk assessment information, so as to complete decision support for the preset enterprise according to the management decision information; the management decision information at least comprises a dynamic pricing decision, a supply chain optimization decision and a personalized marketing decision.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence, and in particular relates to a business administration decision support method and system based on artificial intelligence. Background Art

[0002] With the rapid development of the digital economy, business management decisions are facing a complex and changing market environment. Traditional decision support systems mainly rely on historical data and expert experience, and it is difficult to make accurate and timely predictions for rapidly changing market trends and consumer behaviors. In addition, the constant changes in the internal and external environment of the industry have also increased the difficulty of enterprise operation and management.

[0003] Therefore, there is an urgent need for an artificial intelligence-based business management decision support method to solve at least one of the above problems. Summary of the invention

[0004] This application provides a business management decision support method and system based on artificial intelligence, which aims to solve the problem that traditional decision support systems mainly rely on historical data and expert experience, and it is difficult to make accurate and timely predictions for rapidly changing market trends and consumer behaviors. In addition, the constant changes in the internal and external environment of the industry have also increased the difficulty of enterprise operation and management.

[0005] In a first aspect, the present application provides a business administration decision support method based on artificial intelligence, the method comprising: Obtain digital economic information of preset enterprises to be analyzed; Analyze the digital economic information to obtain consumer behavior data, market trend data, and industry development data; Input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand; Input the market trend data and industry development data into a preset time series analysis model to output market trend information; Input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output risk assessment information corresponding to the preset enterprise; Management decision information of the preset enterprise is generated according to the information on changes in consumer demand, market trends and risk assessment information, so as to provide decision support for the preset enterprise according to the management decision information; the management decision information includes at least one or more of dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

[0006] In some embodiments, obtaining the digital economic information of a preset enterprise to be analyzed includes: obtaining first original data in an internal database corresponding to the preset enterprise based on crawler technology; generating a search identifier according to the preset enterprise, and obtaining second original data in at least one preset web page according to the search identifier; inputting the first original data and the second original data into a natural language processing model, and outputting the first original data and the second original data after data cleaning; and generating the digital economic information based on the cleaned first original data and the second original data.

[0007] In some embodiments, the parsing of the digital economic information to obtain consumer behavior data, market trend data and industry development data includes: parsing the digital economic information to obtain multiple sub-economic information and label information corresponding to each sub-economic information; obtaining text embedding information corresponding to each sub-economic information; extracting relevant information from each text embedding information based on a self-attention mechanism; and generating the consumer behavior data, market trend data and industry development data based on multiple relevant information and the corresponding label information.

[0008] In some embodiments, the deep learning model is a long short-term memory network model; the consumer behavior data is input into a preset deep learning model and the consumer demand change information is output, including: extracting the time series features corresponding to the consumer behavior data; performing sentiment analysis on the consumer behavior data to obtain consumer emotion change information and satisfaction information; inputting the time series features, emotion change information and satisfaction information into the long short-term memory network model and outputting the consumer demand change information.

[0009] In some embodiments, the time series analysis model includes an autoregressive integral moving average model; the market trend data and industry development data are input into a preset time series analysis model to output market trend information, including: extracting the time series data and multiple external factor data corresponding to the market trend data and industry development data; inputting the time series data into the autoregressive integral moving average model to output preliminary trend information; calculating the influence weight of each external factor data on the market trend according to a regression analysis method; generating the market trend information according to the preliminary trend information, each external factor data and the corresponding influence weight.

[0010] In some embodiments, the inputting of the consumer behavior data, market trend data and industry development data into a preset multi-factor model and outputting of risk assessment information corresponding to the preset enterprise include: extracting multiple data factors from the consumer behavior data, market trend data and industry development data according to preset business requirements and data characteristics to obtain principal component analysis information corresponding to each of the data factors; obtaining market environment information and enterprise status information corresponding to the preset enterprise; updating the data factors according to the principal component analysis information, market environment information and enterprise status information; inputting the updated data factors into the multi-factor model, and the multi-factor model outputting risk assessment information corresponding to the preset enterprise.

[0011] In some embodiments, if the management decision information includes the dynamic pricing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: according to the consumer demand change information, obtaining the demand elasticity of each product of the preset enterprise, and determining the optimal pricing strategy corresponding to each time period according to the product demand elasticity; according to the market trend information, obtaining the opponent's pricing strategy information and market trend information; according to the risk assessment information, obtaining potential market risks; generating the dynamic pricing decision according to the opponent's pricing strategy information, market trend information, potential market risks, each of the optimal pricing strategies and the corresponding time periods.

[0012] In some embodiments, if the management decision information includes the supply chain optimization decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: obtaining future market demand information according to the consumer demand change information; obtaining potential supply chain risks of the preset enterprise according to the risk assessment information; the potential supply chain risks include at least supplier reliability and logistics delay risks; obtaining market trend information according to the market trend information; generating the supply chain optimization decision according to the future market demand information, potential supply chain risks and market demand change trends.

[0013] In some embodiments, if the management decision information includes the personalized marketing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: according to the consumer demand change information, obtaining purchasing habit information, purchasing preference information and life cycle value information; generating a user portrait according to the purchasing habit information, purchasing preference information and life cycle value information; according to the market trend information, obtaining market trend information and hot spot information; generating marketing content and recommendation strategies according to the market trend information and hot spot information; generating a risk threshold according to the risk assessment information; generating the personalized marketing decision according to the user portrait, market trend information, hot spot information, marketing content and recommendation strategy.

[0014] In a second aspect, the present application provides a business administration decision support system based on artificial intelligence, the system comprising: An information acquisition unit, used to acquire digital economic information of a preset enterprise to be analyzed; An information analysis unit, used to analyze the digital economic information and obtain consumer behavior data, market trend data, and industry development data; A demand output unit, used to input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand; A trend output unit, used to input the market trend data and industry development data into a preset time series analysis model and output market trend information; A risk output unit, used to input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output risk assessment information corresponding to the preset enterprise; A decision generation unit is used to generate management decision information of the preset enterprise based on the consumer demand change information, market trend information and risk assessment information, so as to complete decision support for the preset enterprise based on the management decision information; the management decision information at least includes dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

[0015] In a third aspect, the present application further provides a computer device, including: Memory and processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the steps of the business management decision support method based on artificial intelligence as described in the first aspect when executing the computer program.

[0016] In a fourth aspect, the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the steps of the artificial intelligence-based business management decision support method as described in the first aspect above.

[0017] The embodiment of the present application provides an artificial intelligence-based business management decision support method and system. First, by collecting digital economic information related to a preset enterprise, this includes data generated by the enterprise itself, and various types of data from the external market, such as consumer behavior data, market trend data, and industry development data. Then, specific consumer behavior data, market trend data, and industry development data are parsed from the acquired digital economic information. These data may come from a variety of sources, such as social media, transaction records, industry reports, etc. Then, the parsed consumer behavior data is input into a preset deep learning model. The deep learning model can learn consumer preferences, purchasing habits, and behavior patterns from a large amount of data, thereby outputting information on changes in consumer demand. This helps companies more accurately predict consumers' future needs. Further, the parsed market trend data and industry development data are input into a preset time series analysis model. The time series analysis model can identify patterns and trends in time series, thereby outputting market trend information. This helps companies understand the current and future market environment and adjust strategies in a timely manner. Furthermore, the consumer behavior data, market trend data, and industry development data are input into a preset multi-factor model. The multi-factor model comprehensively considers multiple factors to evaluate the operational risks of enterprises. This helps enterprises identify potential risk points and take measures to avoid them. Finally, based on the above analysis results, management decision information of the preset enterprise is generated. This information includes at least one or more of dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions. Enterprise managers can formulate more scientific and effective management strategies based on this decision information.

[0018] The provided method has the following beneficial effects: 1. Improve the accuracy and timeliness of decision-making: Through deep learning models and time series analysis models, this method can more accurately predict consumer demand and market trends, making corporate decision-making more scientific. At the same time, the risk assessment results of the multi-factor model help companies to promptly discover and respond to potential risks, thereby improving the timeliness of decision-making.

[0019] 2. Enhance market adaptability: Traditional decision support systems rely on historical data and expert experience, and it is difficult to make accurate predictions for rapidly changing market trends and consumer behaviors. This method can adapt to market changes more quickly and enhance the competitiveness of enterprises through real-time data analysis and model prediction.

[0020] 3. Optimize operating costs: Dynamic pricing decisions can help companies adjust pricing strategies under different market conditions to maximize profits. Supply chain optimization decisions can reduce inventory and logistics costs and improve overall operational efficiency. Personalized marketing decisions can improve marketing effectiveness and reduce ineffective advertising, thereby optimizing operating costs.

[0021] 4. Improve customer satisfaction: Through consumer behavior analysis, companies can understand consumer needs more accurately and provide more personalized services and products. This helps improve customer satisfaction and loyalty and enhance the company's market position.

[0022] 5. Risk management: The risk assessment results of the multi-factor model can help enterprises better identify and manage operational risks and reduce the negative impact caused by market uncertainty and changes in the internal and external environment. This helps enterprises maintain stable operations in an uncertain environment.

[0023] 6. Data-driven decision support: This method makes full use of big data and artificial intelligence technology to transform data into valuable decision-making information. Business managers can make more data-driven decisions based on this information and avoid mistakes caused by subjective judgment.

[0024] In summary, this method provides enterprises with a comprehensive, dynamic and accurate decision support system by comprehensively applying a variety of artificial intelligence technologies, which helps enterprises maintain competitiveness and achieve sustainable development in a rapidly changing market environment.

[0025] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0027] Figure 1 This is a schematic flowchart of the steps of an artificial intelligence-based business management decision support method provided in one embodiment of the present application; Figure 2 It is a structural diagram of an artificial intelligence-based business management decision support system provided in one embodiment of the present application; Figure 3 It is a schematic block diagram of the structure of a computer device provided in one embodiment of the present application.

[0028] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0030] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0031] It should be understood that, in order to facilitate the clear description of the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second", etc. are used to distinguish the same items or similar items with substantially the same functions and effects. Those skilled in the art can understand that the words "first", "second", etc. do not limit the quantity and execution order, and the words "first", "second", etc. do not necessarily limit the difference.

[0032] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0033] It should also be understood that the term “and / or” used in the specification and appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0034] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0035] With the rapid development of the digital economy, business management decisions are facing a complex and changing market environment. Traditional decision support systems mainly rely on historical data and expert experience, and it is difficult to make accurate and timely predictions for rapidly changing market trends and consumer behaviors. In addition, the constant changes in the internal and external environment of the industry have also increased the difficulty of enterprise operation and management.

[0036] Therefore, there is an urgent need for an artificial intelligence-based business management decision support method to solve at least one of the above problems.

[0037] To solve the above problems, please refer to Figure 1 , Figure 1 This is a schematic flow chart of an artificial intelligence-based business administration decision support method provided by an embodiment of the present application. The artificial intelligence-based business administration decision support method can be implemented by a computer device, which can be deployed on a single server or a server cluster. It can also be deployed on a handheld terminal, a laptop computer, a wearable device, or a robot.

[0038] It should be noted that the acquisition of any information corresponding to this application complies with the requirements of relevant laws and regulations, and is implemented with the user's consent and without infringing on relevant privacy.

[0039] To solve the above problems, please refer to Figure 1 Specifically, Figure 1 As shown, the provided business management decision support method based on artificial intelligence includes steps S101 to S106. The details are as follows: Step S101. Obtain digital economic information of a preset enterprise to be analyzed.

[0040] Specifically, by collecting and integrating digital economic information related to the preset enterprises from various data sources (such as e-commerce platforms, social media, industry reports, etc.), this information may include but is not limited to sales records, customer reviews, market share, competitor dynamics, etc. Use API interfaces or web crawler technology to obtain data from public channels or partners' private channels. For example, by calling the API interface of the e-commerce platform, obtain the sales data and customer feedback of the company's products. Use web crawler technology to collect discussions and evaluations about the company on social media. Then provide comprehensive data support to help decision makers understand the company's operating conditions and market performance from multiple perspectives. Real-time or quasi-real-time data acquisition helps to respond quickly to market changes.

[0041] Step S102. Analyze the digital economic information to obtain consumer behavior data, market trend data and industry development data.

[0042] Specifically, the obtained raw data is cleaned and preprocessed to remove noise and irrelevant information. For example, natural language processing (NLP) technology is used to parse text data and extract consumer opinions and preferences. Market trends and development patterns, including seasonal trends, consumer trends, and technological trends, are identified through data mining technology. For example, NLP technology is used to perform sentiment analysis on texts on social media to identify consumers' positive or negative comments on products. Time series analysis technology is used to identify the changing trend of sales in the industry over the past few years. This ensures the quality and accuracy of the data and improves the reliability of subsequent analysis. Valuable information is extracted from large amounts of data to help companies better understand the market and consumers.

[0043] Step S103. Input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand.

[0044] Specifically, by using deep learning models (such as LSTM, GRU, etc.) to model consumer behavior data, we can predict future changes in consumer demand. During model training, historical data can be used as a training set, and the validation set can be used to adjust model parameters to ensure the accuracy and generalization of the model. For example, by building an LSTM model, inputting sales data and consumer evaluation data from the past few quarters, we can predict changes in consumer demand in the next quarter. Through model training, we can identify the impact of specific factors (such as holidays, promotional activities, etc.) on consumer demand. By understanding the changing trends of consumer demand in advance, we can help companies develop more targeted market strategies. Through accurate demand forecasting, we can optimize inventory management and reduce waste and out-of-stock risks.

[0045] Step S104: Input the market trend data and industry development data into a preset time series analysis model, and output market trend information.

[0046] Specifically, by using time series analysis models (such as ARIMA, Prophet, etc.) to analyze market trend data and industry development data, the future market trend can be predicted. The model can be combined with external economic indicators (such as GDP, unemployment rate, etc.) for comprehensive prediction. For example, a Prophet model is established, and the industry sales data and related economic indicators of the past few years are input to predict the market trend in the next year. Through model training, the impact of cyclical fluctuations in the industry and economic cycles on the market can be identified. Then, by providing forward-looking predictions of market trends, it helps companies to plan and adjust strategies in advance. Through comprehensive analysis, the accuracy and reliability of predictions can be improved.

[0047] Step S105. Input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output the risk assessment information corresponding to the preset enterprise.

[0048] Specifically, by using multi-factor models (such as factor analysis, linear regression, etc.) to comprehensively analyze various data, the risks faced by the enterprise are assessed.

[0049] The model can consider multiple factors, such as market demand, supply chain stability, competitor strategies, etc., and output a comprehensive risk score. Build a linear regression model, input consumer behavior data, market trend data, and industry development data, and evaluate the financial risk, market risk, and technical risk of the enterprise. Through model training, identify key risk factors and help enterprises take corresponding risk management measures. Provide enterprises with comprehensive risk assessment to help identify and manage potential operating risks. Improve the scientificity and reliability of decision-making through data-driven risk assessment.

[0050] Step S106. Generate management decision information of the preset enterprise based on the consumer demand change information, market trend information and risk assessment information, so as to complete decision support for the preset enterprise based on the management decision information; the management decision information includes at least one or more of dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

[0051] Specifically, by combining the output information of the previous steps, the decision support system is used to generate specific management decision information. Management decision information can include dynamic pricing decisions, supply chain optimization decisions, and personalized marketing decisions. The system can provide a visual interface to help decision makers intuitively understand various information and suggestions.

[0052] For example, based on information on changes in consumer demand and market trends, we recommend that companies adjust their pricing strategies to cope with the upcoming seasonal demand peak. Combined with risk assessment information, we recommend that companies optimize their supply chains to ensure the stability and flexibility of their supply chains during market fluctuations. Through personalized marketing decisions, we launch customized marketing activities for different consumer groups to improve market response speed and customer satisfaction.

[0053] By providing data-driven management decision-making suggestions, we help enterprises respond to market changes and risks more scientifically. Through the visual interface, we improve the transparency and operability of decisions and optimize the operational efficiency and market competitiveness of enterprises.

[0054] In summary, the artificial intelligence-based business management decision support method provided in this application provides enterprises with more accurate and scientific decision support by comprehensively analyzing multiple data sources, helping enterprises to better cope with the rapidly changing market environment.

[0055] In some embodiments, obtaining the digital economic information of a preset enterprise to be analyzed includes: obtaining first original data in an internal database corresponding to the preset enterprise based on crawler technology; generating a search identifier according to the preset enterprise, and obtaining second original data in at least one preset web page according to the search identifier; inputting the first original data and the second original data into a natural language processing model, and outputting the first original data and the second original data after data cleaning; and generating the digital economic information based on the cleaned first original data and the second original data.

[0056] By using web crawler technology, relevant data is extracted from the internal database of the enterprise, which may include sales records, inventory data, customer feedback, etc. It is necessary to ensure that the crawler technology operates under the premise of legality and compliance, and comply with relevant laws and regulations on data privacy and security. Generate a search identifier based on the preset enterprise, and obtain the second raw data from at least one preset web page based on the search identifier to generate a search identifier related to the preset enterprise (such as enterprise name, product keywords, etc.). Use these search identifiers to crawl relevant data from preset web pages (such as news websites, industry forums, social media, etc.). Ensure the accuracy of the search identifier so that useful information related to the enterprise can be screened out from a large number of web pages. Input the first raw data and the second raw data into the natural language processing model, and output the first raw data and the second raw data after data cleaning. Use the natural language processing (NLP) model to pre-process the first raw data and the second raw data, including removing noise, standardizing text, extracting key information, etc. Ensure the accuracy and consistency of the data for subsequent analysis and modeling. Generate the digital economic information based on the cleaned first raw data and the second raw data. Integrate and analyze the cleaned data to generate digital economic information, which may include consumer behavior data, market trend data, and industry development data.

[0057] By combining internal databases and external network data, we ensure the diversity and comprehensiveness of information sources. We use NLP technology to clean and preprocess data to improve data accuracy and consistency. We integrate digital economic information generated by multiple data sources to provide high-quality input for subsequent deep learning models, time series analysis models, and multi-factor models, thereby improving the scientificity and reliability of the final decision support.

[0058] Exemplarily, the generating of the digital economic information based on the cleaned first original data and the second original data includes: obtaining first modal information corresponding to the first original data; obtaining second modal information corresponding to the second original data; performing data conversion on the first original data according to the first modal information to convert the modality of the first original data into a preset modality; performing data conversion on the second original data according to the second modal information to convert the modality of the second original data into the preset modality; generating the digital economic information based on the converted first original data and the second original data.

[0059] The process of generating digital economic information based on the cleaned first original data and the second original data is further refined, specifically including the following contents: Identify the modality of the first raw data, for example, the data may be structured (such as an Excel spreadsheet), semi-structured (such as JSON format), or unstructured (such as text).

[0060] Identify the modality of the second original data, which may also be structured, semi-structured or unstructured. Perform data conversion on the first original data according to the first modal information to convert the modality of the first original data into a preset modality. Use data conversion technology to convert the first original data from its original modality to a preset modality, such as converting unstructured text data into structured table data. Ensure that the format of the converted data is unified to facilitate subsequent integration and analysis. Perform data conversion on the second original data according to the second modal information to convert the modality of the second original data into the preset modality. Use the same data conversion technology to convert the second original data from its original modality to the preset modality. Ensure that the format of the converted data is consistent with that of the first original data. Generate the digital economic information based on the converted first original data and the second original data. Integrate the first original data and the second original data after modality conversion to generate digital economic information.

[0061] Ensure that the information generated is comprehensive, accurate, consistent, and able to reflect the company's operating conditions and market changes.

[0062] Through data conversion technology, data of different modalities are unified into a preset modality to facilitate subsequent integration and analysis. Modal conversion can reduce the complexity of data processing and improve the efficiency of data processing. Data of unified modality is easier to understand and analyze, which helps decision makers quickly obtain useful information. High-quality, modality-unified data provides more reliable input for subsequent deep learning models, time series analysis models, and multi-factor models, thereby improving the accuracy of final decision support.

[0063] Suppose a company needs to analyze the market performance and consumer behavior of its products: the first raw data is sales records and customer feedback obtained from the company's internal database, which may be structured Excel tables. The second raw data is discussions and comments about the company obtained from social media and news websites, which are unstructured text data. Identify the first raw data as structured data and the second raw data as unstructured data. Clean and standardize the first raw data (Excel table) to ensure consistent data format. Use NLP technology to perform sentiment analysis and keyword extraction on the second raw data (text data), and convert it into structured tabular data, including sentiment scores and keywords. Integrate the cleaned and converted data to generate digital economic information including sales records, customer feedback, and social media discussions.

[0064] Through this process, companies can gain a more comprehensive understanding of the market and consumer situation, and thus make more scientific management decisions. For example, they can adjust marketing strategies and product pricing based on positive feedback on social media and increased sales data; optimize supply chain management and reduce inventory pressure based on market trends and risk assessments.

[0065] In some embodiments, the parsing of the digital economic information to obtain consumer behavior data, market trend data and industry development data includes: parsing the digital economic information to obtain multiple sub-economic information and label information corresponding to each sub-economic information; obtaining text embedding information corresponding to each sub-economic information; extracting relevant information from each text embedding information based on a self-attention mechanism; and generating the consumer behavior data, market trend data and industry development data based on multiple relevant information and the corresponding label information.

[0066] A plurality of sub-economic information is extracted from the digital economic information. The sub-economic information may be specific fragments or paragraphs, and each fragment contains different topics or information.

[0067] Corresponding label information is assigned to each sub-economic information. The label information can be a category label (such as "consumer comments", "market news", "industry reports", etc.) or a sentiment label (such as "positive", "negative", "neutral", etc.).

[0068] Use natural language processing (NLP) technology to convert each sub-economic information into a text embedding (TextEmbedding), which is usually a high-dimensional vector that can represent the semantic information of the text. Common text embedding technologies include Word2Vec, GloVe, BERT, etc. Use the self-attention mechanism to analyze each text embedding information and extract the key information related to the specific label. The self-attention mechanism can help the model focus on the most important part of the text and ignore irrelevant or minor information. Generate the consumer behavior data, market trend data and industry development data based on multiple related information and corresponding label information. Integrate the extracted key information with the corresponding label information to generate consumer behavior data, market trend data and industry development data.

[0069] For example, the positive and negative sentiment information related to "consumer comments" can be integrated to generate consumer behavior data; the economic indicator information related to "market news" can be integrated to generate market trend data; and the technological development information related to "industry reports" can be integrated to generate industry development data.

[0070] Through the self-attention mechanism, key information related to specific tags can be extracted more accurately, improving the accuracy of information extraction.

[0071] For example, extracting positive and negative keywords from a large number of consumer reviews can more accurately reflect consumers' true emotions and needs. Text embedding technology can convert text data into numerical vectors, which is convenient for processing and analysis by machine learning models. The self-attention mechanism can quickly identify important parts of the text and reduce the time complexity of data processing.

[0072] Through label information, the extracted information can be classified to facilitate subsequent analysis and understanding. For example, consumer behavior data can be refined into "purchase behavior", "user experience", "after-sales service", etc., and market trend data can be refined into "economic indicators", "policy impact", etc.

[0073] Combining multiple sub-economic information and key information extracted by the self-attention mechanism, the generated consumer behavior data, market trend data, and industry development data are more comprehensive and scientific. Decision recommendations generated based on high-quality data are more reliable, helping companies make scientific decisions in a complex and changing market environment.

[0074] Suppose a retail company needs to analyze the market performance and consumer behavior of its newly launched smart speakers: extract multiple sub-economic information from internal databases and external network data, such as: Consumer review: "This smart speaker is powerful and has clear sound, but the price is a bit high." Tagged "Consumer review".

[0075] News article: "The smart home market has grown rapidly due to the pandemic and is expected to continue to expand in the coming years." Tagged "Market News".

[0076] Industry Report: "The development of artificial intelligence technology has promoted innovation in the smart speaker market, and new products continue to emerge." Tagged "Industry Report".

[0077] The BERT model is used to convert the above sub-economic information into a high-dimensional vector and generate text embedding information.

[0078] For example, the text embedding information of a consumer review may be a 512-dimensional vector. The self-attention mechanism is used to analyze each text embedding information and extract key information related to the label information.

[0079] For example, keywords such as “powerful functions”, “clear sound”, and “high price” are extracted from the text embedding information of “consumer reviews”, and keywords such as “market growth” and “next few years” are extracted from the text embedding information of “market news”.

[0080] Integrate the extracted key information with the corresponding label information to generate consumer behavior data, market trend data and industry development data.

[0081] For example, the generated consumer behavior data may include: consumers’ positive comments on product features and sound quality. Consumers’ negative comments on product prices. The generated market trend data may include: the rapid growth trend of the smart home market. Market forecasts for the next few years. The generated industry development data may include: the impact of the development of artificial intelligence technology on the smart speaker market. New product launches and market competitiveness analysis.

[0082] Based on consumers' positive comments on product functions and sound quality, companies can further optimize product features and enhance market competitiveness; based on consumers' negative comments on prices, companies can consider adjusting pricing strategies to improve product market acceptance. Through market trend data, companies can predict the growth rate of the smart home market, plan market layout and resource investment in advance, and seize market opportunities. Through industry development data, companies can understand the development direction of artificial intelligence technology and new product trends, so as to stay ahead in R&D and market promotion. Combining consumer behavior data, market trend data and industry development data, companies can more comprehensively assess potential risks and formulate corresponding risk management measures, such as responding to market competition and price fluctuations.

[0083] Through the detailed steps of the above embodiment, enterprises can more accurately analyze digital economic information, extract key consumer behavior, market trends and industry development data, and make more scientific and effective management decisions.

[0084] In some embodiments, the deep learning model is a long short-term memory network model; the consumer behavior data is input into a preset deep learning model and the consumer demand change information is output, including: extracting the time series features corresponding to the consumer behavior data; performing sentiment analysis on the consumer behavior data to obtain consumer emotion change information and satisfaction information; inputting the time series features, emotion change information and satisfaction information into the long short-term memory network model and outputting the consumer demand change information.

[0085] Identify time series features in consumer behavior data, such as the time point of data collection (such as date, timestamp). Time series features can help the model understand the time-varying patterns of consumer behavior, so as to better capture the trend of demand changes. Use sentiment analysis technology to analyze the text data in consumer behavior data and identify positive, negative, and neutral sentiments. Through sentiment analysis, obtain information about consumers' emotional changes, such as satisfaction and dissatisfaction with a product. Emotional change information and satisfaction information can be used as input features of the model to help the model better understand consumers' emotional changes. Combine the extracted time series features, emotional change information, and satisfaction information into an input vector and input it into the LSTM model. The LSTM model can process time series data, capture long-term and short-term dependencies, and output information about changes in consumer demand. Information about changes in consumer demand can include trends in increasing or decreasing demand, key time nodes for demand changes, etc.

[0086] The LSTM model can capture the long-term and short-term dependencies in time series data and improve the accuracy of demand forecasting. By combining time series features and sentiment analysis information, the model can more comprehensively understand the changing patterns of consumer behavior. Sentiment analysis technology can reveal consumers' true emotions and satisfaction, helping companies to adjust products and services in a timely manner. For example, if it is found that consumers are less satisfied with a certain function, companies can give priority to improving that function. Since the LSTM model can process continuous time series data, it can update information on changes in consumer demand in real time. Real-time information helps companies respond quickly to market changes and adjust marketing strategies and product planning. Based on the demand change information output by the LSTM model, companies can formulate more scientific market strategies and product improvement plans. Predicting future demand changes can help companies make more reasonable arrangements in production, inventory, and marketing.

[0087] Suppose a home appliance manufacturer needs to analyze changes in consumer demand for its latest air purifier: Extract time features from sales records and customer feedback, such as monthly sales volume, monthly number of customer feedback, and timestamps. Use sentiment analysis technology to analyze customer feedback and identify the sentiment (positive, negative, neutral) and satisfaction of each feedback. For example, identify "the air purification effect is very good, but the price is high" as a positive sentiment with high satisfaction; "frequent freezes during use" as a negative sentiment with low satisfaction. Combine the extracted time series features, sentiment change information, and satisfaction information into an input vector. For example, the input vector may include monthly sales volume, average sentiment score of customer feedback, and average satisfaction of customer feedback. Input these input vectors into the LSTM model to output information about changes in consumer demand. The information output by the LSTM model may include: monthly demand change trends, such as changes in demand volume, predictions of demand peaks, and trends in consumer satisfaction with product features and prices. Possible reasons for changes in market demand, such as seasonal factors and the impact of promotional activities.

[0088] In some embodiments, the time series analysis model includes an autoregressive integral moving average model; the market trend data and industry development data are input into a preset time series analysis model to output market trend information, including: extracting the time series data and multiple external factor data corresponding to the market trend data and industry development data; inputting the time series data into the autoregressive integral moving average model to output preliminary trend information; calculating the influence weight of each external factor data on the market trend according to a regression analysis method; generating the market trend information according to the preliminary trend information, each external factor data and the corresponding influence weight.

[0089] Identify time series data in market trend data and industry development data, such as monthly sales, quarterly industry reports, etc. Identify multiple external factor data, such as macroeconomic indicators (such as GDP growth rate), policy changes, competitors' market performance, etc. Use the ARIMA model to model the extracted time series data and output preliminary market trend information. The ARIMA model can capture trends, periodicities, and random fluctuations in time series data to predict future market trends. Use regression analysis methods (such as linear regression, ridge regression, etc.) to calculate the impact weight of each external factor data on market trends. For example, calculate the impact weight of GDP growth rate on market sales, the impact weight of policy changes on market share, etc. Generate the market trend information based on the preliminary trend information, each of the external factor data and the corresponding impact weight. Combine the preliminary trend information with the external factors data and their influence weights to generate the final market trend information. For example, combine the preliminary market forecast and the influence weights of external factors to generate the market sales forecast and influencing factors for the next few months.

[0090] The ARIMA model can effectively capture trends and periodicity in time series data and improve the accuracy of market forecasts. By combining the influence weights of external factor data, market trends can be more comprehensively evaluated and forecast errors can be reduced.

[0091] By calculating the impact weight of external factors through regression analysis, we can consider the impact of various external factors on the market. For example, policy changes, macroeconomic indicators, competitor dynamics, etc., the impact of these factors on market trends can be quantified.

[0092] Based on the weights and preliminary trend information of multiple external factors, the generated market trend information is more reliable and scientific. Enterprises can formulate detailed market strategies based on this information, such as adjusting production plans, increasing marketing investment, etc. By comprehensively analyzing market trends and the impact of external factors, enterprises can better identify potential market risks. For example, if it is found that a certain policy change may lead to a decline in market demand, enterprises can take measures in advance to deal with it.

[0093] Extract annual sales data from sales records, such as sales in 2020, 2021, and 2022. Identify multiple external factor data, such as annual GDP growth rate, government automobile subsidy policy, and new product launches by competitors. Input annual sales data into the ARIMA model and output preliminary market trend information. For example, it is initially predicted that sales will increase by 10% in 2024. Use linear regression method to calculate the impact weight of each external factor on market sales. For example, the impact weight of GDP growth rate is 0.6, the impact weight of automobile subsidy policy is 0.3, and the impact weight of competitor's new products is 0.1.

[0094] Combine the preliminary trend information with the external factor data and their influence weights to generate the final market trend information. For example, combining the preliminary forecast and the influence weights of external factors, it is generated that the market sales in 2024 will increase by 8% (the impact of GDP growth is greater, but the impact of new product releases by competitors weakens some of the growth). Through the detailed steps of the above embodiment, enterprises can more accurately predict market trends, formulate scientific market strategies and countermeasures, and thus stay ahead in the fierce market competition.

[0095] In some embodiments, the inputting of the consumer behavior data, market trend data and industry development data into a preset multi-factor model and outputting of risk assessment information corresponding to the preset enterprise include: extracting multiple data factors from the consumer behavior data, market trend data and industry development data according to preset business requirements and data characteristics to obtain principal component analysis information corresponding to each of the data factors; obtaining market environment information and enterprise status information corresponding to the preset enterprise; updating the data factors according to the principal component analysis information, market environment information and enterprise status information; inputting the updated data factors into the multi-factor model, and the multi-factor model outputting risk assessment information corresponding to the preset enterprise.

[0096] According to the specific business needs of the enterprise, multiple data factors are extracted from consumer behavior data, market trend data and industry development data. For example, factors such as purchase frequency and satisfaction are extracted from consumer behavior data; factors such as sales growth rate and market share are extracted from market trend data; factors such as technological innovation frequency and industry policy changes are extracted from industry development data. The principal component analysis (PCA) technology is used to reduce the dimension of the extracted data factors and extract the principal component information of each data factor. Principal component analysis can reduce the redundancy of data and improve the calculation efficiency and accuracy of the model. The current market environment information of the preset enterprise is collected, such as macroeconomic indicators, industry competition status, etc. The current status information of the preset enterprise is collected, such as financial status, operational efficiency, etc. The principal component analysis information, market environment information and enterprise status information are combined to update the extracted data factors. For example, the weight of the consumer behavior data factor is adjusted according to the market environment information, and the weight of the industry trend data factor is adjusted according to the enterprise status information. The updated data factors are input into the multi-factor model, which can be a linear regression model, a logistic regression model, a decision tree model, etc. The risk assessment information output by the multi-factor model can include financial risk, market risk, operational risk, etc. The multi-factor model takes into account multiple data factors and can comprehensively assess the risks of an enterprise. The principal component analysis reduces the redundancy of data and improves the accuracy and reliability of the model. The principal component analysis technology can effectively reduce the dimension, reduce the complexity of data processing and the calculation time. The updated data factors are more accurate, which helps to improve the performance of the model. The risk assessment information based on the multi-factor model can provide scientific decision-making support for enterprises. For example, the finance department can adjust the capital structure based on the risk assessment information, and the marketing department can adjust the marketing strategy based on the risk assessment information. Through systematic risk assessment, enterprises can better identify and respond to potential risks. For example, if the supply chain risk is high, the enterprise can look for alternative suppliers in advance; if the market risk is high, the enterprise can reduce the marketing budget.

[0097] Suppose an e-commerce company needs to assess its risk for the coming year: extract factors such as purchase frequency and satisfaction from consumer behavior data. Extract factors such as sales growth rate and market share from market trend data. Extract factors such as new technology adoption rate and industry policy changes from industry development data. Use PCA technology to reduce the dimension of the extracted data factors and generate principal component information. For example, the principal components extracted from consumer behavior data are purchase frequency and satisfaction; the principal components extracted from market trend data are sales growth rate and market share; the principal components extracted from industry development data are new technology adoption rate and policy changes. Collect current macroeconomic indicators, such as GDP growth rate and inflation rate. Collect the company's current financial status, such as profit margin and cash flow. Adjust the weight of data factors based on market environment information and company status information. For example, if the economic environment is unstable, increase the weight of market trend data factors; if the company's cash flow is tight, increase the weight of financial risk factors. Input the updated data factors into the multi-factor model and output risk assessment information. For example, the model may output financial risk of 30%, market risk of 40%, and operational risk of 30%.

[0098] In some embodiments, if the management decision information includes the dynamic pricing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: according to the consumer demand change information, obtaining the demand elasticity of each product of the preset enterprise, and determining the optimal pricing strategy corresponding to each time period according to the product demand elasticity; according to the market trend information, obtaining the opponent's pricing strategy information and market trend information; according to the risk assessment information, obtaining potential market risks; generating the dynamic pricing decision according to the opponent's pricing strategy information, market trend information, potential market risks, each of the optimal pricing strategies and the corresponding time periods.

[0099] According to the information on changes in consumer demand, the demand elasticity of each product of the preset enterprise is obtained. Using the information on changes in consumer demand, the impact of price changes on demand for each product is analyzed to calculate demand elasticity. Demand elasticity can reflect the consumer's sensitivity to price and help determine the optimal pricing strategy. According to demand elasticity, the optimal pricing strategy is formulated for each time period. For example, if consumers are sensitive to prices in a certain time period, a promotion strategy can be adopted; if consumers are not sensitive to prices, the price can be appropriately increased. Competitors' pricing strategy information is extracted from market trend information, such as competitors' price changes in different time periods. Market trend information is extracted from market trend information, such as economic environment, seasonal changes, etc. Potential market risks are extracted from risk assessment information, such as economic depression, policy adjustments, etc. The competitor's pricing strategy, market trend information and potential market risks are combined to generate the final dynamic pricing decision. For example, in a certain time period, the competing product reduces the price for promotion, and the market trend shows that consumer demand increases, but there is a potential risk of economic depression. The enterprise can adopt corresponding pricing strategies, such as partial promotion, slight price increase, etc.

[0100] Dynamic pricing decisions can be adjusted according to different market environments and time periods, making them more flexible and adaptable. For example, during holidays or promotional seasons, companies can adopt more aggressive pricing strategies. Through demand elasticity analysis, companies can formulate pricing strategies that are more in line with consumer demand and improve market competitiveness. For example, appropriate price adjustments can attract more consumers and increase market share. Through risk assessment information, companies can better identify and respond to potential market risks. For example, if there is a risk of economic depression, companies can adjust their pricing strategies in advance to reduce losses. Dynamic pricing decisions generated based on multi-faceted information (demand elasticity, market trends, rival strategies, potential risks) are more scientific and reliable. For example, companies can formulate more reasonable promotion plans and pricing strategies based on comprehensive information.

[0101] Suppose an electronics manufacturer needs to develop a dynamic pricing strategy for its latest smartwatch: analyze information about changes in consumer demand, calculate the impact of price changes on demand for smartwatches over different time periods, and derive demand elasticity. For example, the demand elasticity for smartwatches is -1.5, which means that for every 1% drop in price, demand increases by 1.5%.

[0102] According to the elasticity of demand, the optimal pricing strategy is formulated for each time period. For example, during holidays and promotional seasons, a price reduction promotion strategy is adopted; during the off-season, a strategy of maintaining the original price or slightly raising the price is adopted.

[0103] Extract competitor pricing strategy information from market trend information, such as a competitor reducing prices by 20% during holidays. Extract market trend information from market trend information, such as a sharp increase in demand for smart watches during holidays. Extract potential market risks from risk assessment information, such as a decline in overall spending power due to an economic recession.

[0104] The final dynamic pricing decision is generated by integrating competitor pricing strategies, market trend information and potential market risks. For example, during holidays, companies can reduce prices by 15% to attract consumers, while closely monitoring market dynamics and competitor responses and flexibly adjusting strategies.

[0105] The expression of the dynamic pricing strategy includes: ; ; represents the price adjustment amount, Indicates at time Pricing, . It is the demand elasticity coefficient, which indicates the weight of the impact of demand elasticity on price adjustment. If the demand elasticity is high, the impact of price changes on demand will be greater, and companies need to adjust prices more cautiously. is the demand elasticity value, In time The price adjustment amount. It is the price reduction factor of competitors during holidays. is the influence weight of competitor pricing strategy. If competitors reduce prices, the company may need to consider following the price reduction or adopt other strategies to remain competitive. It is the market trend that shows the coefficient of demand increase. It is the impact weight of market trends. If the market trend shows an increase in demand, the company can appropriately increase prices; if the market trend shows a decrease in demand, the company may need to reduce prices for promotions. is the impact weight of risk assessment. If there is a higher market risk, the company needs to adopt a more conservative pricing strategy; if the risk is lower, the company can adjust the price more aggressively. It is the potential market risk factor shown by risk assessment.

[0106] This can be determined based on historical data and corporate strategy. If the company wants to be more responsive to changes in demand, Can be set higher; otherwise, Can be set lower. The value of can be determined based on the market influence of competitors. If the market influence of competitors is large, It can be set higher; conversely, it can be set lower. The value of can be determined based on the clarity of market trends and the market sensitivity of the enterprise. If the market trend is clear and the enterprise responds quickly to market changes, It can be set higher; conversely, it can be set lower. The value of can be determined based on the enterprise's risk tolerance and market environment. If the enterprise's risk tolerance is low or the market environment is unstable, It can be set higher; conversely, it can be set lower.

[0107] In some embodiments, if the management decision information includes the supply chain optimization decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: obtaining future market demand information according to the consumer demand change information; obtaining potential supply chain risks of the preset enterprise according to the risk assessment information; the potential supply chain risks include at least supplier reliability and logistics delay risks; obtaining market trend information according to the market trend information; generating the supply chain optimization decision according to the future market demand information, potential supply chain risks and market demand change trends.

[0108] By extracting market demand information in future time periods from information on changes in consumer demand, such as sales forecasts for the next few months. Based on the risk assessment information, obtain the potential supply chain risks of the preset enterprise. Extract potential supply chain risks from the risk assessment information, such as supplier reliability, logistics delay risks, etc. Supplier reliability: the supplier's ability to deliver on time and quality stability. Logistics delay risk: delays that may occur in the logistics process, such as traffic congestion, weather impacts, etc. Extract information such as future market growth trends and policy changes from market trend information. Combine future market demand information, potential supply chain risks, and market trend information to generate supply chain optimization decisions. For example, increase inventory preparation to cope with future demand peaks, find reliable backup suppliers in advance, and optimize logistics routes to reduce delay risks.

[0109] With information on future market demand and market trends, companies can proactively manage their supply chains and avoid imbalances between supply and demand. For example, they can increase inventory preparations in advance to cope with peak demand. The assessment of potential supply chain risks can help companies identify and respond to various risks in the supply chain. For example, they can find alternative suppliers in advance to reduce single reliance on suppliers; optimize logistics routes to reduce logistics delays. Supply chain optimization decisions can improve the operational efficiency of companies and reduce inventory and logistics costs. For example, they can plan inventory reasonably to avoid capital occupation caused by excessive inventory; optimize logistics routes to reduce transportation time and costs. Supply chain optimization decisions generated based on multiple aspects of information (future demand, market trends, supply chain risks) are more scientific and reliable. For example, companies can formulate more reasonable production plans and inventory management strategies based on market trends and supply chain risks.

[0110] Suppose a food company needs to optimize its supply chain to cope with future changes in demand and potential risks: Extract sales forecasts for the next few months from information on changes in consumer demand. For example, sales are expected to increase by 15% in the next three months. Extract potential supply chain risks such as supplier reliability and logistics delay risk from risk assessment information. For example, the delivery reliability of supplier A is 80% and the risk of logistics delay is 20%. Extract future market growth trends and policy changes from market trend information. For example, the market is expected to continue to grow and the government supports green logistics. Combine future market demand information, supply chain potential risks, and market trend information to generate supply chain optimization decisions. For example, increase inventory preparation to cope with future demand peaks, find reliable backup suppliers in advance, optimize logistics routes to reduce delay risks, and adopt green logistics to meet policy requirements.

[0111] In some embodiments, if the management decision information includes the personalized marketing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: according to the consumer demand change information, obtaining purchasing habit information, purchasing preference information and life cycle value information; generating a user portrait according to the purchasing habit information, purchasing preference information and life cycle value information; according to the market trend information, obtaining market trend information and hot spot information; generating marketing content and recommendation strategies according to the market trend information and hot spot information; generating a risk threshold according to the risk assessment information; generating the personalized marketing decision according to the user portrait, market trend information, hot spot information, marketing content and recommendation strategy.

[0112] By extracting consumers' purchasing habits (such as purchase frequency and purchase time), purchasing preferences (such as favorite product types and functional requirements), and life cycle value (such as long-term purchase value and consumption potential) from information on changes in consumer demand. Integrate the extracted purchasing habit information, purchasing preference information, and life cycle value information to generate a detailed user portrait. For example, user A often buys smart home appliances on holidays, likes new products and cost-effective products, and has a high long-term purchase value. Extract future market growth trends, competitor dynamics, etc. from market trend information. Extract current market hot spots and consumer focus from market hot spot information. For example, market trends show an increase in demand for smart home appliances, and hot spot information shows that consumers are concerned about health functions. Extract potential market risks and internal corporate risks from risk assessment information to generate risk management thresholds.

[0113] For example, if the potential market risk is high, set the risk threshold at 30%.

[0114] Generate the personalized marketing decision based on the user portrait, market trend information, hot information, marketing content and recommendation strategy Combine user profiles, market trend information, hot spot information and risk thresholds to generate personalized marketing content and recommendation strategies. For example, for user A, push holiday promotion information and focus on recommending smart home appliances with health functions, while monitoring market dynamics and potential risks to ensure the sustainability of marketing strategies.

[0115] Personalized marketing based on user portraits can reach target consumers more accurately and improve marketing effectiveness. For example, marketing content is more in line with consumers' interests and needs, increasing click-through rates and conversion rates. Through market trend information and hot spot information, companies can quickly respond to market changes and adjust marketing strategies. For example, if the market hot spot is health functions, companies can focus on promoting related products. Through risk assessment information, companies can set risk thresholds to ensure that marketing strategies are implemented within a controllable risk range. For example, if the potential market risk is high, companies can appropriately reduce high-risk marketing activities. Personalized marketing decisions generated based on multi-faceted information (user portraits, market trends, hot spot information, risk assessment) are more scientific and reliable. For example, companies can formulate more reasonable promotion plans and marketing budgets based on comprehensive information.

[0116] Suppose an Internet company needs to develop a personalized marketing strategy for its newly launched health app: extract purchasing habits, purchasing preferences, and lifetime value information from information on changes in consumer demand. For example, user B uses health apps more than 3 times a week, prefers apps with comprehensive functions, and has a higher lifetime value.

[0117] Extract the growth trend of the smart health application market and the dynamics of competitors from the market trend information. For example, the smart health application market is expected to grow rapidly in the next few years, and competitor A has launched a new health monitoring function. Extract the current focus of consumers from the market hot spot information, for example, consumers generally pay attention to mental health and weight management functions. Extract potential market risks and internal risks of enterprises from risk assessment information and generate risk thresholds. For example, the potential market risk is the change of consumer privacy protection policy, and the risk threshold is set to 20%.

[0118] Generate personalized marketing content and recommendation strategies based on the user portrait of user B. For example, push user B's special articles on mental health and recommend health applications that include mental health monitoring functions. Design marketing points that are different from competitors based on market trend information, such as emphasizing the comprehensiveness of the application and personalized recommendation functions. According to market hot information, push success stories and user reviews about weight management. Monitor risk thresholds to ensure that marketing activities are implemented within a controllable risk range. For example, strengthen user data protection measures to avoid user trust issues caused by changes in privacy policies. Integrate the above information to generate the final personalized marketing decision. For example, for user B, push weekly mental health monitoring and weight management content, provide personalized health application recommendations, and at the same time strengthen the publicity of user data protection to reduce potential market risks.

[0119] Personalized marketing content and recommendation strategies can better meet the needs and interests of users and improve user satisfaction and loyalty. For example, user B trusts the app more because he receives personalized health advice, which increases the frequency of use and the possibility of recommending it to other friends. Personalized marketing based on user portraits can increase the click-through rate and conversion rate of marketing content. For example, the pushed mental health content is more likely to be clicked and shared by user B. Through market trend information and hot information, companies can quickly adjust marketing strategies and seize market opportunities. For example, when the market focus turns to weight management, companies can adjust the focus of publicity in time to attract more users. By setting risk thresholds, companies can ensure that marketing activities are implemented within a controllable risk range and avoid negative impacts caused by market changes. For example, strengthening user data protection measures can reduce user churn caused by changes in privacy policies. Personalized marketing decisions generated based on multi-faceted information (user portraits, market trends, hot information, risk assessment) are more comprehensive and scientific. For example, companies can formulate more reasonable marketing budgets and promotion plans based on comprehensive information to achieve optimal resource allocation. Through the above steps, companies can reach target consumers more accurately and improve marketing effectiveness, while effectively responding to market changes and potential risks and achieving scientific management and decision-making.

[0120] See also Figure 2 As shown, Figure 2 1 is a schematic diagram of the structure of the business administration decision support system 200 based on artificial intelligence provided in an embodiment of the present application. The business administration decision support system 200 based on artificial intelligence is used to execute the steps of the business administration decision support method based on artificial intelligence shown in the above embodiments. The business administration decision support system 200 based on artificial intelligence can be a single server or a server cluster, or the business administration decision support system 200 based on artificial intelligence can be a terminal, which can be a handheld terminal, a laptop computer, a wearable device or a robot, etc.

[0121] like Figure 2 As shown, the business management decision support system 200 based on artificial intelligence includes: The information acquisition unit 201 is used to acquire the digital economic information of a preset enterprise to be analyzed; An information parsing unit 202 is used to parse the digital economic information to obtain consumer behavior data, market trend data, and industry development data; The demand output unit 203 is used to input the consumer behavior data into a preset deep learning model and output consumer demand change information; The trend output unit 204 is used to input the market trend data and industry development data into a preset time series analysis model and output market trend information; The risk output unit 205 is used to input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output the risk assessment information corresponding to the preset enterprise; The decision generation unit 206 is used to generate the management decision information of the preset enterprise according to the consumer demand change information, market trend information and risk assessment information, so as to complete the decision support for the preset enterprise according to the management decision information; the management decision information at least includes dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

[0122] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working processes of the artificial intelligence-based business administration decision support system and each module described above can refer to the corresponding processes in the artificial intelligence-based business administration decision support method embodiments described in the above embodiments, and will not be repeated here.

[0123] The above-mentioned business management decision support method based on artificial intelligence can be implemented in the form of a computer program, which can be used in Figure 2 Run on the system shown.

[0124] See also Figure 3 , Figure 31 is a schematic block diagram of the structure of a computer device provided in an embodiment of the present application. The computer device includes a processor, a memory and a network interface connected via a device bus, wherein the memory may include a storage medium and an internal memory.

[0125] The storage medium can store an operating device and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any one of the business management decision support methods based on artificial intelligence.

[0126] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0127] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any business management decision support method based on artificial intelligence.

[0128] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the terminal to which the scheme of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0129] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0130] In one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps: Obtain digital economic information of preset enterprises to be analyzed; Analyze the digital economic information to obtain consumer behavior data, market trend data, and industry development data; Input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand; Input the market trend data and industry development data into a preset time series analysis model to output market trend information; Input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output risk assessment information corresponding to the preset enterprise; Management decision information of the preset enterprise is generated according to the information on changes in consumer demand, market trends and risk assessment information, so as to provide decision support for the preset enterprise according to the management decision information; the management decision information includes at least one or more of dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

[0131] The present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the steps of the business management decision support method based on artificial intelligence as described in the first aspect above.

[0132] The computer-readable storage medium may be an internal storage unit of the computer device described in the above embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device.

[0133] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A business management decision support method based on artificial intelligence, characterized in that: The method comprises: Obtain digital economic information of preset enterprises to be analyzed; Analyze the digital economic information to obtain consumer behavior data, market trend data, and industry development data; Input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand; Input the market trend data and industry development data into a preset time series analysis model to output market trend information; Input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output risk assessment information corresponding to the preset enterprise; Management decision information of the preset enterprise is generated according to the information on changes in consumer demand, market trends and risk assessment information, so as to provide decision support for the preset enterprise according to the management decision information; the management decision information includes at least one or more of dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

2. The method according to claim 1, characterized in that The obtaining of the digital economic information of the preset enterprise to be analyzed includes: Acquire first original data in an internal database corresponding to the preset enterprise based on crawler technology; Generate a search identifier according to the preset enterprise, and obtain second original data from at least one preset web page according to the search identifier; Inputting the first original data and the second original data into a natural language processing model, and outputting the first original data and the second original data after data cleaning; The digital economic information is generated according to the cleaned first original data and the second original data.

3. The method according to claim 1, characterized in that The analyzing of the digital economic information to obtain consumer behavior data, market trend data and industry development data includes: Parsing the digital economic information to obtain a plurality of sub-economic information and label information corresponding to each of the sub-economic information; Obtaining text embedding information corresponding to each sub-economic information; Extract relevant information from each of the text embedding information based on a self-attention mechanism; The consumer behavior data, market trend data and industry development data are generated based on the multiple related information and the corresponding tag information.

4. The method according to claim 1, characterized in that: The deep learning model is a long short-term memory network model; the consumer behavior data is input into a preset deep learning model, and the consumer demand change information is output, including: Extracting time series features corresponding to the consumer behavior data; Performing sentiment analysis on the consumer behavior data to obtain consumer sentiment change information and satisfaction information; The time series characteristics, emotion change information and satisfaction information are input into the long short-term memory network model, and the consumer demand change information is output.

5. The method according to claim 1, characterized in that: The time series analysis model includes an autoregressive integrated moving average model; the market trend data and industry development data are input into a preset time series analysis model, and the market trend information is output, including: Extracting time series data and multiple external factor data corresponding to the market trend data and industry development data; Inputting the time series data into an autoregressive integrated moving average model to output preliminary trend information; Calculate the influence weight of each external factor data on the market trend according to the regression analysis method; The market trend information is generated according to the preliminary trend information, each of the external factor data and the corresponding impact weight.

6. The method according to claim 1, characterized in that The step of inputting the consumer behavior data, market trend data and industry development data into a preset multi-factor model and outputting risk assessment information corresponding to the preset enterprise includes: Extract multiple data factors from the consumer behavior data, market trend data, and industry development data according to preset business requirements and data characteristics; Obtaining principal component analysis information corresponding to each of the data factors; Obtaining market environment information and enterprise status information corresponding to the preset enterprise; Updating the data factors according to the main component analysis information, market environment information and enterprise status information; The updated data factors are input into the multi-factor model, and the multi-factor model outputs the risk assessment information corresponding to the preset enterprise.

7. The method according to claim 1, characterized in that If the management decision information includes the dynamic pricing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: According to the consumer demand change information, the demand elasticity of each product of the preset enterprise is obtained, and the optimal pricing strategy corresponding to each time period is determined according to the product demand elasticity; Obtain competitor pricing strategy information and market trend information based on the market trend information; Obtaining potential market risks based on the risk assessment information; The dynamic pricing decision is generated based on the competitor pricing strategy information, market trend information, potential market risks, each of the optimal pricing strategies and the corresponding time period.

8. The method according to claim 1, characterized in that: If the management decision information includes the supply chain optimization decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: Obtaining future market demand information based on the consumer demand change information; According to the risk assessment information, the potential supply chain risks of the preset enterprise are obtained; the potential supply chain risks at least include supplier reliability and logistics delay risks; Obtaining market trend information based on the market trend information; The supply chain optimization decision is generated based on the future market demand information, potential supply chain risks and market demand change trends.

9. The method according to claim 1, characterized in that: If the management decision information includes the personalized marketing decision; the management decision information of the preset enterprise is generated according to the consumer demand change information, market trend information and risk assessment information, including: According to the consumer demand change information, obtain purchasing habit information, purchasing preference information and life cycle value information; Generate a user profile based on the purchasing habit information, purchasing preference information and life cycle value information; According to the market trend information, obtain market trend information and hot spot information; Generate marketing content and recommendation strategies based on the market trend information and hot spot information; generating a risk threshold based on the risk assessment information; The personalized marketing decision is generated based on the user portrait, market trend information, hot spot information, marketing content and recommendation strategy.

10. An artificial intelligence-based business management decision support system, characterized in that: The system comprises: An information acquisition unit, used to acquire digital economic information of a preset enterprise to be analyzed; An information analysis unit, used to analyze the digital economic information and obtain consumer behavior data, market trend data, and industry development data; A demand output unit, used to input the consumer behavior data into a preset deep learning model and output information on changes in consumer demand; A trend output unit, used to input the market trend data and industry development data into a preset time series analysis model and output market trend information; A risk output unit, used to input the consumer behavior data, market trend data and industry development data into a preset multi-factor model, and output risk assessment information corresponding to the preset enterprise; A decision generation unit is used to generate management decision information of the preset enterprise based on the consumer demand change information, market trend information and risk assessment information, so as to complete decision support for the preset enterprise based on the management decision information; the management decision information at least includes dynamic pricing decisions, supply chain optimization decisions and personalized marketing decisions.

Images