AI-based enterprise data asset digital service method for industrial ecological chain management

Abstract

The invention provides an enterprise data asset digital service method for industrial ecological chain management based on AI, and the method comprises the steps: carrying out the two-dimensional value evaluation of enterprise data asset data based on AI driving, generating the data evaluation value of the enterprise data asset data, and analyzing the industrial characteristic indexes of the enterprise data asset data. And finally, according to the demand analysis of the enterprise target user on the data assets, finding the enterprise data assets of which the data evaluation values and the industrial characteristic indexes meet the demands, and pushing the enterprise data assets to the enterprise target user. Therefore, the enterprise data assets are not in a static data storage management and calling mode any more, but are pushed based on user demands, AI-driven analysis is supported, and asset digital circulation is carried out according to a mode that data evaluation values and industrial characteristic indexes of the enterprise data assets are matched with the user demands. Dynamic management and operation of the data assets are realized, a service system of a data asset industry ecological chain can be created, and the processing and circulation efficiency and the operation value of the data assets are improved.

Description

Technical Field

[0001] The present invention relates to the field of data asset application technology, and in particular to an enterprise data asset digitization service method based on AI-based industrial ecosystem management, and an enterprise data asset digitization service device, electronic device, and computer-readable storage medium based on AI-based industrial ecosystem management. Background Art

[0002] The core significance of enterprise data asset digitization services lies in:

[0003] 1. Asset appreciation and value release

[0004] Convert raw data into standardized digital assets, activate the value potential of data evaluation, and achieve the transformation from "cost center" to "profit center" (IDC research shows that data-driven enterprises increase efficiency by 40%).

[0005] 2. Intelligent decision support

[0006] Real-time analysis of customer behavior, supply chain dynamics, and other data drives precision marketing (for example, a retail giant achieved a 28% increase in inventory turnover through data assetization).

[0007] 3. Business model innovation

[0008] Support new business models such as data product development (such as industrial equipment operation data packages) and data API services to expand corporate revenue sources.

[0009] 4. Compliance and Security

[0010] Establish a data classification, encryption, and permission system to meet GDPR / Data Security Law requirements and avoid the risk of tens of millions of yuan in fines (the average global cost of data breaches in 2023 was US$4.35 million).

[0011] Traditional methods of providing digital services for enterprise data assets usually revolve around data collection, storage, processing, analysis, and application, and follow certain standards and processes. The main methods and specific means are:

[0012] 1. Data standardization and governance

[0013] Metadata management: Create a data dictionary and business glossary to record data definitions, sources, formats, owners, and other information (such as using data catalog tools).

[0014] Data model design: Design conceptual, logical, and physical data models to ensure that data is structured and consistent (e.g., relational database design).

[0015] Data quality rule definition and execution: Develop rules to check data accuracy, completeness, consistency, uniqueness, and timeliness (e.g., data constraints, validation codes, and specialized tools).

[0016] Master Data Management: Establish an authoritative, single version of the “golden record” of core business entity data (e.g., customers, products, suppliers) and ensure consistency across systems (using MDM tools).

[0017] Data policy and process development: clarify data ownership, access rights, data lifecycle management, data security and compliance requirements.

[0018] Standardized dictionary and code mapping: unify the meaning of codes between different systems to ensure consistency in data exchange.

[0019] 2. Data integration and warehouse construction: Aggregate data scattered across multiple heterogeneous systems onto a unified platform for centralized storage and analysis. Specific methods are as follows:

[0020] ETL / ELT:

[0021] Extraction: Reading data from source systems (such as ERP, CRM, business systems).

[0022] Transformation: Apply cleansing, integration, and standardization rules to process data.

[0023] Loading: Loading the processed data into the target system (such as a data warehouse). Sometimes the order changes to ELT (first loading the raw data into the target platform, then performing the transformation).

[0024] Enterprise Data Warehouse: Build centralized, integrated, subject-oriented, historical data sets to support analysis (such as the Kimball / Inmon model).

[0025] Data mart: A targeted subset of data built for a specific department or application (such as sales analysis or financial reporting).

[0026] Data lake (early form): Stores relatively raw, diversely structured data, offers high flexibility but is complex to manage (relies on strong governance to deliver value and is typically incorporated into a more mature management framework later).

[0027] 3. Data analysis and business intelligence: transforming data into insights to support decision-making. Specific methods are as follows:

[0028] OLAP / Multidimensional Analysis: Rapidly analyze multidimensional data through slicing, dicing, drilling, and rotation (such as Cube deployed in a data warehouse).

[0029] Report generation: Generate structured reports regularly (such as weekly sales reports and monthly financial reports).

[0030] Dashboard: Monitor key performance indicators and business status in real time through a graphical interface.

[0031] Ad hoc query: Allows users to freely query data as needed (requires self-service tools).

[0032] Basic Data Mining and Statistical Analysis: Use statistical methods (such as regression analysis, clustering, and classification) to discover patterns and relationships in data (using tools such as SAS and SPSS).

[0033] However, in the process of digitizing enterprise data assets, enterprises often use disorganized and non-standardized methods to apply these services. Enterprise data assets are typically stored in a generalized manner in enterprise databases, with subsequent retrieval and access required. Consequently, a robust ecosystem for managing enterprise data assets has not been established. This traditional digital service approach fails to ensure orderly storage of enterprise data assets, nor allows for applications based on their tangible value. Furthermore, it fails to digitally manage enterprise data assets from the perspectives of enterprise cost and market transformation. Consequently, the digital service applications of traditional enterprise data assets are largely managed and operated internally, lacking dynamic industry ecosystem management.

[0034] In addition, the digital service methods of traditional enterprise data assets are mostly static data storage and operation management methods, which do not support AI-driven digital services and have low data processing efficiency. Summary of the Invention

[0035] In order to solve the technical problems existing in the prior art, the present invention provides the following technical solutions:

[0036] On the one hand, a method for digitalizing enterprise data assets for industrial ecosystem management based on AI is provided. The method is implemented by an electronic device and includes:

[0037] S1. Build an asset evaluation system that integrates cost-oriented evaluation, market reference evaluation, and income forecast evaluation;

[0038] S2. Based on the LLM large language model training, the evaluation mechanism of the asset organic integration evaluation system is learned to generate an LLM-asset organic evaluation model;

[0039] S3. Collect enterprise data asset data generated by the enterprise in real time, and input the enterprise data asset data into the LLM-asset organic valuation model to perform asset organic valuation, thereby generating a data valuation value for the enterprise data asset data;

[0040] S4. Analyze the industry characteristic indicators of the enterprise data asset data based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix;

[0041] S5. Push the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data.

[0042] Preferably, the evaluation model of the cost-oriented evaluation is as follows:

[0043] ,

[0044] in:

[0045] : The cost of collecting and maintaining enterprise data assets;

[0046] TC‌: total acquisition cost;

[0047] ‌MC‌: annual maintenance cost;

[0048] ‌ROIC‌: rate of cost of capital;

[0049] ‌U‌: Use premium factor.

[0050] Preferably, the evaluation model of the market reference evaluation is as follows:

[0051] ,

[0052] in:

[0053] : The current market reference price of enterprise data assets;

[0054] ‌: The transaction price of the reference asset;

[0055] ‌: Adjustment coefficient of the i-th dimension (5 dimensions: technology maturity, scenario adaptability, value density, technology correction coefficient and timeliness index).

[0056] Preferably, the evaluation model for the revenue forecast evaluation is as follows:

[0057] ,

[0058] in:

[0059] NPV is the expected return on an enterprise’s data assets;

[0060] R t ‌: expected return in year t;

[0061] ‌λ t ‌: risk attenuation factor;

[0062] ‌r‌: industry discount rate;

[0063] NPV uses LSTM to predict the economic life cycle and expected returns of enterprise data assets.

[0064] Preferably, the IVI-BVI-PVI three-dimensional indicator evaluation matrix includes:

[0065] The IVI matrix unit includes the following internal industry characteristic indicators:

[0066] The field missing rate used to characterize data completeness,

[0067] The error rate used to characterize the accuracy of the data,

[0068] Cross-system conflict rate used to characterize data consistency;

[0069] The BVI Matrix unit includes the following business industry characteristic indicators:

[0070] The Spearman correlation coefficient with the core KPI used to characterize the strategic support,

[0071] The proportion of covered business process nodes used to characterize the scenario penetration rate,

[0072] Data update delay time used to characterize timeliness;

[0073] The PVI matrix unit includes the following performance industry characteristic indicators:

[0074] The annual revenue per TB of data used to represent the ROI conversion rate,

[0075] The proportion of new products driven by innovation is used to characterize the contribution of innovation.

[0076] The size of the dominant data alliance (number of nodes ≥ 50) used to characterize ecological control power.

[0077] Preferably, S4, based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix, analyzing the industry characteristic indicators of the enterprise data asset data, including:

[0078] Parsing the enterprise data asset data and classifying it into internal asset data, business asset data, and performance asset data;

[0079] Invoking the LLM large language model to calculate and analyze the internal value, business value, and performance value of the internal asset data, business asset data, and performance asset data based on the matrix cells of the IVI-BVI-PVI three-dimensional indicator evaluation matrix;

[0080] The internal value, business value and performance value are weighted and summed to generate the industry characteristic indicators of the enterprise data asset data.

[0081] Preferably, S5, pushing the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data, includes:

[0082] The front-end accesses the enterprise target users' demand information on the required enterprise data assets;

[0083] Calling the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including: cost-oriented keywords, market reference keywords, revenue forecast keywords and / or indicator keywords;

[0084] Constructing data retrieval logic, and generating corresponding asset demand prompt words based on the demand keywords, and inputting the asset demand prompt words into the LLM large language model;

[0085] The LLM large language model retrieves the enterprise data asset data whose data evaluation value and industry characteristic indicators match the demand information from the database according to the prompt word;

[0086] Output the enterprise data asset data and feed it back to the front end.

[0087] In another aspect, an enterprise data asset digitization service device for industrial ecosystem management based on AI is provided. The enterprise data asset digitization service device for industrial ecosystem management based on AI is used to implement the above-mentioned enterprise data asset digitization service method for industrial ecosystem management based on AI. The device includes:

[0088] The collection end is used to collect the demand information of enterprise target users for the required enterprise data assets through the front end;

[0089] The backend server is configured to call the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including cost-oriented keywords, market reference keywords, revenue forecast keywords, and / or indicator keywords; construct data retrieval logic, and generate corresponding asset demand prompt words based on the demand keywords, and input the asset demand prompt words into the LLM large language model; and the LLM large language model retrieves enterprise data assets from a database based on the prompt words, whose data evaluation values ​​and industry characteristic indicators match the demand information;

[0090] Message middleware is used to push the enterprise data asset data to the enterprise target users.

[0091] On the other hand, an electronic device is provided, comprising: a processor; and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, any one of the above-mentioned AI-based industrial ecosystem management enterprise data asset digitalization service methods is implemented.

[0092] On the other hand, a computer-readable storage medium is provided, in which at least one instruction is stored. The at least one instruction is loaded and executed by a processor to implement any one of the above-mentioned AI-based industrial ecosystem management enterprise data asset digitalization service methods.

[0093] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0094] The present invention proposes an AI-driven system that can perform a two-dimensional value assessment on enterprise data asset data. It can perform an organic asset assessment based on the LLM-asset organic assessment model, generate the data assessment value of the enterprise data asset data, and analyze the industry characteristic indicators of the enterprise data asset data. Finally, based on the demand analysis of the enterprise target users for data assets, the enterprise data assets whose data assessment value and industry characteristic indicators meet their needs are found and pushed to the enterprise target users. Therefore, enterprise data assets are no longer a static data storage management and call method, but are pushed based on user needs, and support AI-driven analysis. Assets are digitized and circulated in a way that matches the data assessment value and industry characteristic indicators of enterprise data assets with user needs, realizing dynamic management and operation of data assets, and being able to create a service system for the data asset industry ecosystem chain, improving the processing and circulation efficiency and operational value of data assets. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0096] Figure 1 This is a flow chart of an enterprise data asset digitization service method for AI-based industrial ecosystem management provided by an embodiment of the present invention;

[0097] Figure 2 Schematic diagram of a data evaluation value evaluation mechanism provided by an embodiment of the present invention;

[0098] Figure 3 This is a block diagram of an enterprise data asset digitization service device for AI-based industrial ecosystem management provided by an embodiment of the present invention;

[0099] Figure 4 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0100] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0101] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0102] In the embodiments of the present invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same. The terms "of," "corresponding," and "corresponding" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same.

[0103] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript form such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0104] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0105] The embodiment of the present invention provides an enterprise data asset digital service method based on AI-based industrial ecological chain management, which can be implemented by an electronic device, which can be a terminal or a server. Figure 1 The flowchart of the enterprise data asset digital service method based on AI-based industrial ecosystem management is shown. The processing flow of this method may include the following steps:

[0106] S1. Build an asset evaluation system that integrates cost-oriented evaluation, market reference evaluation, and income forecast evaluation;

[0107] S2. Based on the LLM large language model training, the evaluation mechanism of the asset organic integration evaluation system is learned to generate an LLM-asset organic evaluation model;

[0108] S3. Collect enterprise data asset data generated by the enterprise in real time, and input the enterprise data asset data into the LLM-asset organic valuation model to perform asset organic valuation, thereby generating a data valuation value for the enterprise data asset data;

[0109] S4. Analyze the industry characteristic indicators of the enterprise data asset data based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix;

[0110] S5. Push the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data.

[0111] The present invention combines the LLM large language model to provide AI-driven computing services. By constructing an asset organic integration evaluation system that includes cost-oriented evaluation, market reference evaluation, and revenue forecast evaluation, as well as an IVI-BVI-PVI three-dimensional indicator evaluation matrix, the LLM large language model realizes data value evaluation and industry characteristic indicator analysis of corresponding enterprise data asset data, thereby providing adapted enterprise data asset data to target users of the enterprise, accurately operating enterprise data asset data, and realizing the maximum circulation, operational efficiency and performance of enterprise data asset data.

[0112] In step S1, it is necessary to build in advance an asset organic integration evaluation system that includes cost-oriented evaluation, market reference evaluation, and income forecast evaluation.

[0113] The asset organic integration evaluation system adopts a multi-model fusion architecture, integrating the three basic evaluation dimensions of cost approach, income approach and market approach to build a composite evaluation system. The evaluation framework consists of three core levels:

[0114] Basic evaluation layer: Deploy a cost-oriented algorithm model group, including the data acquisition cost model (TC), maintenance cost model (MC), and capital cost rate model (ROIC).

[0115] Market reference layer: Build a horizontal comparison model based on similarity matching, and introduce five adjustment parameters such as technology maturity, scenario adaptability, value density, technology correction coefficient and timeliness index.

[0116] ‌Return Prediction Layer‌: Uses LSTM time series networks to predict the economic life cycle returns of data assets, combined with Monte Carlo simulation to quantify risk factors.

[0117] Therefore, it is possible to analyze the dynamic management and operation of data assets, build a service system for the data asset industry ecosystem, and improve the processing and circulation efficiency and operational value of data assets.

[0118] The evaluation models of each layer are as follows:

[0119] Preferably, the evaluation model of the cost-oriented evaluation is as follows:

[0120] ,

[0121] in:

[0122] : The cost of collecting and maintaining enterprise data assets;

[0123] TC‌: total acquisition cost (including ETL, cleaning, annotation, etc.);

[0124] ‌MC‌: annual maintenance cost (storage + computing + manpower);

[0125] ROIC: Cost of capital (adjusted for the firm's WACC);

[0126] ‌U‌: Uses a premium coefficient (between 1.2 and 3.5). This allows for precise calculation of the cost contribution of each item in a single table based on a hierarchical backtracking algorithm based on the data lineage graph.

[0127] The collection and maintenance of enterprise data assets requires costs. For example, the cost structure during the collection phase is shown in Table 1 below:

[0128] Cost Type Proportion illustrate Data collection 35-45% Including ETL tools and manual annotation Storage Processing 25-30% Cloud storage and computing resource consumption Quality Management 15-20% Cleaning, deduplication, standardization, etc. Safety and Compliance 10-15% Encryption, auditing, permission management, etc.

[0129] Table 1

[0130] Preferably, the evaluation model of the market reference evaluation is as follows:

[0131] ,

[0132] in:

[0133] : The current market reference price of enterprise data assets;

[0134] ‌: The transaction price of the reference asset;

[0135] ‌: Adjustment coefficient for the i-th dimension (five dimensions: technology maturity (enterprise scoring), scenario adaptability (proportion of business nodes covered), value density (field-level entropy calculation), technology correction coefficient (0.8-1.5), and timeliness index (T + update delay penalty). A dynamic case library can be used to integrate over 2,000 industry transaction cases, with each benchmark coefficient updated quarterly.

[0136] Preferably, the evaluation model for the revenue forecast evaluation is as follows:

[0137] ,

[0138] in:

[0139] NPV is the expected return on an enterprise’s data assets;

[0140] R t ‌: Expected return in year t (using LSTM to predict expected return in year t based on historical big data);

[0141] ‌λ t ‌ : risk attenuation factor (Monte Carlo simulation);

[0142] ‌r‌: industry discount rate (determined based on the market);

[0143] NPV uses LSTM to predict the economic life cycle and expected returns of enterprise data assets. The specific steps of LSTM prediction of the economic life cycle and returns of enterprise data assets can be referred to the following scheme:

[0144] 1. Problem definition and indicator quantification

[0145] Economic life cycle: defined as the critical year T from when data generates value to when the benefits continue to decay to the maintenance cost;

[0146] Expected return: Quantitative formula:

[0147] Annual income = incremental income from data application x data maintenance cost x data depreciation cost.

[0148] Key outputs:

[0149] Sequence forecast: future annual return values ​​(time series).

[0150] End of life: In the first year t, the benefit (t) ≤ maintenance cost threshold is met.

[0151] 2. Data preparation (core feature engineering as shown in Table 2 below)

[0152]

[0153] Table 2

[0154] 3. Model Architecture Design

[0155] from tensorflow.keras.models import Sequential

[0156] from tensorflow.keras.layers import LSTM, Dense, Dropout

[0157] model = Sequential([

[0158] LSTM(units=64, input_s , return_sequences=True),

[0159] Dropout(0.3),

[0160] LSTM(units=32),

[0161] Dense(16, activation='relu'),

[0162] Dense(1) outputs the annual income value ])

[0164] Input dimension: (sliding window length, number of features)

[0165] Output: Forecast value of earnings for the next year (rolling forecast to the end year).

[0166] 4. Economic Life End Determination Algorithm

[0167] def predict_lifecycle(model, init_data, cost_threshold):

[0168] lifecycle = []

[0169] current_input = init_data initial input sequence

[0170] year = 0

[0171] while True:

[0172] pred = model.pre [0][0] Forecast of next year's earnings

[0173] lifecycle.append(pred)

[0174] Determine termination condition (benefit < maintenance cost threshold)

[0175] if pred <= cost_threshold:

[0176] break

[0177] Rolling update input: remove the earliest year and add forecast year data

[0178] current_input = np. ent_input[-1,-1, 0] = pred Update the revenue feature

[0179] year += 1

[0180] return lifecycle, year Returns the lifecycle and end year.

[0181] 5. Key optimization strategies

[0182] Multi-scale feature fusion:

[0183] Short-term features: quarterly volatility (using 1D-CNN to extract local patterns);

[0184] Long-term characteristics: 3-year moving average return trend;

[0185] Dynamic threshold mechanism:

[0186] Dynamic thresholds based on business growth

[0187] cost_threshold = base_cost .

[0188] Non-stationarity handling:

[0189] First-order difference stationary: return_diff(t) = return(t) return(t-1);

[0190] Logarithmic transformation reduces heteroskedasticity.

[0191] 6. Verification and evaluation indicators (as shown in Table 3 below)

[0192]

[0193] Table 3

[0194] 7. Business Implementation Scenarios

[0195] Data asset decision dashboard: (Visualization module must include: benefit decay curve, lifespan warning, cost-benefit ratio heat map)

[0196] Dynamic depreciation model:

[0197] Annual depreciation rate = 1 (forecast earnings for the year / peak earnings) 0.5 .

[0198] Through this framework, enterprises can achieve: early warning of data asset failure risks 1-3 years in advance; adjust data maintenance resource investment based on predictions; and guide data asset restructuring or elimination decisions.

[0199] Specifically, such as Figure 2 As shown, the present invention is based on the LLM large language model training to learn the evaluation mechanism of the asset organic fusion evaluation system to generate an LLM-asset organic evaluation model; through the model, the enterprise data asset data generated by the real-time collection enterprise is organically evaluated to generate the data evaluation value of the enterprise data asset data, thereby quickly managing the industrial ecological chain of the enterprise data asset data.

[0200] First, based on the different core indicators in the system (cost model (TC), maintenance cost model (MC), and capital cost rate model (ROIC), data on five adjustment parameters such as technology maturity, scenario adaptability, value density, technology correction coefficient, and timeliness index, and asset data related to revenue forecasts), we collected enterprise data asset datasets corresponding to the indicators and performed feature processing:

[0201] (1) Feature encoding:

[0202] Structured data: TabTransformer architecture (6-head attention);

[0203] Unstructured data: RoBERTa-base text encoder;

[0204] Numerical features: quantile transformation + standardization;

[0205] (2) Multi-expert system:

[0206] Cost Expert: 3-layer MLP (256 / 128 / 64 hidden units);

[0207] Market Expert: GraphSAGE+GAT hybrid network;

[0208] Revenue Expert: TCN Temporal Convolutional Network (dilation factor = 2) [0:5] ).

[0209] (3) Meta-learning fusion layer:

[0210] Dynamic Weight Generator:

[0211] w i = softmax(ϕ i / ∑ϕ j ),

[0212] ϕ i : Expert contribution based on SHAP value (sliding window mean);

[0213] Divergence detection mechanism: When the expert output differs by >15%, manual review is triggered.

[0214] Next, call an LLM large language model (you can call the framework of an existing open source model) as the base model and input the feature data into the base model for training:

[0215] Training data configuration:

[0216] Positive sample: 5,000+ evaluated data assets (covering 8 major industries including finance and manufacturing)

[0217] Negative samples: Synthesizing abnormal cases through generative adversarial networks (GANs)

[0218] Data enhancement: MixUp algorithm (α=0.4) is used to improve the performance of small sample scenarios

[0219] The hyperparameter settings are shown in Table 4 below:

[0220]

[0221] Table 4

[0222] Training Acceleration Technology:

[0223] Adopt ZeRO-3 stage optimizer state segmentation;

[0224] Activate checkpoint technology (set checkpoints every 2 floors);

[0225] Mixed precision training (FP16 + dynamic loss scaling).

[0226] By training and generating the LLM-asset organic evaluation model, it is possible to conduct an organic asset evaluation of the enterprise data assets at each level of the mechanism based on the learned evaluation mechanisms, and output the corresponding data evaluation value.

[0227] When deploying application models, the three assessment dimensions can be organically integrated with LLM's complex pattern recognition capabilities to achieve a leap forward from single-point assessment to ecosystem value management. Based on the assessed value of enterprise data assets, each enterprise's data asset can be stored, managed, and operated in a hierarchical manner, enabling the management of the enterprise data asset's industrial ecosystem and improving the effectiveness of digital services.

[0228] Preferably, the IVI-BVI-PVI three-dimensional indicator evaluation matrix includes:

[0229] The IVI matrix unit includes the following internal industry characteristic indicators:

[0230] The field missing rate used to characterize data completeness,

[0231] The error rate used to characterize the accuracy of the data,

[0232] Cross-system conflict rate used to characterize data consistency;

[0233] The BVI Matrix unit includes the following business industry characteristic indicators:

[0234] The Spearman correlation coefficient with the core KPI used to characterize the strategic support,

[0235] The proportion of covered business process nodes used to characterize the scenario penetration rate,

[0236] Data update delay time used to characterize timeliness;

[0237] The PVI matrix unit includes the following performance industry characteristic indicators:

[0238] The annual revenue per TB of data used to represent the ROI conversion rate,

[0239] The proportion of new products driven by innovation is used to characterize the contribution of innovation.

[0240] The size of the dominant data alliance (number of nodes ≥ 50) used to characterize ecological control power.

[0241] The steps for constructing the IVI-BVI-PVI three-dimensional indicator evaluation matrix include defining three core dimensions, collecting relevant data, calculating indicator values ​​and integrating them into an evaluation framework.

[0242] First, the three units of the matrix are clarified: IVI (internal industry characteristic indicators), BVI (business industry characteristic indicators) and PVI (performance industry characteristic indicators), each unit contains three specific indicators.

[0243] Secondly, identify and collect necessary data sources, such as internal databases, business system logs, financial records, etc., to ensure that the data covers the calculation requirements of all indicators.

[0244] Then, the value of each indicator is calculated item by item, using a standardized formula to ensure comparability.

[0245] Next, the indicator values ​​are integrated into a three-dimensional matrix, and the scores of each dimension can be displayed through visualization tools (such as radar charts or heat maps) to comprehensively evaluate the health, business value and performance contribution of data assets.

[0246] Finally, establish regular evaluation mechanisms, such as quarterly or annual updates, to monitor changes in metrics and optimize data management strategies.

[0247] In the IVI matrix, internal industry characteristic indicators focus on fundamental data attributes and are calculated as follows: The field missing rate is used to indicate data completeness and is calculated as the number of missing fields divided by the total number of fields, multiplied by 100%. The formula is (number of missing fields / total number of fields) × 100%. The error rate is used to indicate data accuracy and is calculated as the number of incorrect data entries (e.g., data points that failed verification) divided by the total number of data entries, multiplied by 100%. The formula is (number of incorrect data entries / total number of data entries) × 100%. The cross-system conflict rate is used to indicate data consistency and is calculated as the number of points with inconsistent data values ​​between different systems divided by the total number of data points, multiplied by 100%. The formula is (number of conflicting data points / total number of data points) × 100%. Conflicting data points must be identified through inter-system comparison.

[0248] In the BVI matrix unit, the business industry characteristic indicator emphasizes the support role of data for the business. Its calculation method is as follows: The Spearman correlation coefficient with the core KPI is used to characterize the strategic support degree. The calculation is based on the rank correlation between the data indicator value and the core KPI value of the business (such as sales or user growth rate). The formula is Spearman correlation coefficient ρ = 1 [6 × Σ(rank difference square) / (n × ], where n is the number of samples, and rank difference refers to the difference in ranking between the data value and the KPI value. The proportion of business process nodes covered is used to indicate scenario penetration. It is calculated as the number of business process nodes covered by data (such as order processing or inventory management nodes) divided by the total number of nodes, then multiplied by 100%. The formula is (number of covered nodes / total number of nodes) × 100%. Data update latency is used to indicate timeliness. It is calculated as the average difference between the data update timestamp and the event occurrence timestamp. The formula is Σ(update timestamp × event occurrence timestamp) / total number of events, and the unit is usually hours or days.

[0249] In the PVI matrix, the Performance Industry Characteristic Indicator measures the economic and innovative benefits of data. Its calculation method is as follows: Annual revenue per TB of data represents ROI conversion rate and is calculated as annual total revenue (e.g., data-driven revenue) divided by the total amount of data (in TB). The formula is: Annual total revenue / Total amount of data (TB). The proportion of new products driven represents innovation contribution and is calculated as the number of new data-driven products (e.g., products developed based on data analysis) divided by the total number of new products, multiplied by 100%. The formula is: (Number of new products driven / Total number of new products) × 100%. Ecosystem control is measured by the size of the dominant data alliance and is calculated as the number of nodes in the alliance (must be ≥ 50). The formula is: Number of nodes. This indicator directly uses the actual number of nodes in the dominant alliance, but screening is required to ensure sustained influence by selecting those that meet the scale requirements.

[0250] Preferably, S4, based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix, analyzing the industry characteristic indicators of the enterprise data asset data, including:

[0251] Parsing the enterprise data asset data and classifying it into internal asset data, business asset data, and performance asset data;

[0252] Invoking the LLM large language model to calculate and analyze the internal value, business value, and performance value of the internal asset data, business asset data, and performance asset data based on the matrix cells of the IVI-BVI-PVI three-dimensional indicator evaluation matrix;

[0253] The internal value, business value and performance value are weighted and summed to generate the industry characteristic indicators of the enterprise data asset data.

[0254] The steps for analyzing the industry characteristic indicators of enterprise data assets by combining the IVI-BVI-PVI three-dimensional indicator evaluation matrix are as follows:

[0255] 1. Data analysis and classification

[0256] Technical process:

[0257] First, collect the original enterprise data assets and perform data cleaning and preprocessing; then classify them through the classification engine:

[0258] Rule 1: Internal asset data;

[0259] Rule 2: Business asset data;

[0260] Rule 3. Performance asset data.

[0261] Instructions:

[0262] Data cleaning: Use Python (Pandas / Spark) to handle missing values, duplicate data, and format standardization;

[0263] Classification rule definition (example):

[0264] Internal asset data: organizational structure, IT system logs, and employee skills database;

[0265] Business asset data: order flow, customer profiles, and supply chain records;

[0266] Performance asset data: financial reports, KPI achievement rates, market share data;

[0267] Automated tool (pseudocode example): keyword-based classification

[0268] def classify_data(row):

[0269] if "cost" in row['tags'] or "efficiency" in row['tags']:

[0270] return "Internal"

[0271] elif "customer" in row['tags'] or "transaction" in row['tags']:

[0272] return "Business"

[0273] elif "revenue" in row['tags'] or "kpi" in row['tags']:

[0274] return "Performance".

[0275] 2. LLM-driven 3D value calculation

[0276] Technical architecture:

[0277] Internal asset data is sent to the IVI calculation module;

[0278] Business asset data is sent to the BVI calculation module;

[0279] Performance asset data is sent to the PVI calculation module.

[0280] The three calculation results are sent to the LLM analysis engine.

[0281] Key operations:

[0282] LLM prompt word design (taking IVI calculation as an example): "Evaluate the intrinsic value (IVI) of the data asset based on the following dimensions:

[0283] Dimension 1: Data integrity (weight 0.3): Check the field missing rate and timeliness;

[0284] Dimension 2: Governance maturity (weight 0.4) analyzes metadata completeness and access control policies;

[0285] Dimension 3: Operation and maintenance costs (weight 0.3) calculates storage / computing resource consumption;

[0286] Output: IVI score from 0-100 and summary of basis".

[0287] API call example:

[0288] import openai

[0289] response = openai.ChatCompletion.create(

[0290] model="gpt-4-turbo",

[0291] mess act_score(response['choices'][0]['message']['content']).

[0292] 3. Three-dimensional value integration and the generation of industrial characteristics

[0293] Computational model:

[0294] Industry characteristic index = α IVI + β BVI + γ PVI;

[0295] (α+β+γ=1, weights are dynamically configured according to corporate strategy);

[0296] Execution process:

[0297] Weight configuration (example):

[0298] Internal value (IVI): α = 0.2 (in the manufacturing scenario);

[0299] Business value (BVI): β = 0.5;

[0300] Performance value (PVI): γ = 0.3;

[0301] Aggregate calculation: final_score = + + .

[0302] Feature interpretation:

[0303] Generate radar chart visualizations (using Matplotlib / Tableau);

[0304] LLM automatically generates a diagnostic report: "Current Data Asset Industry Characteristics (Score 73):

[0305] √ Outstanding Business Value (BVI=85): Leading in customer data utilization;

[0306] × Performance value needs to be improved (PVI=60): Insufficient financial relevance".

[0307] 3D matrix configurability:

[0308] Create a JSON configuration file to dynamically modify the indicator weight:

[0309] "IVI": {"weights": {"Integrity":0.3, "Governance":0.4, "Cost":0.3}},

[0310] "BVI": {"weights": {"liquidity":0.5, "scenario coverage":0.5}}

[0311] }.

[0312] LLM precision optimization solution

[0313] Use Few-shot learning to inject industry examples;

[0314] Develop a verification module (triggering manual review when the score fluctuation is >15%);

[0315] Safety and compliance design:

[0316] Data desensitization: Apache Griffin + encryption sandbox;

[0317] Audit trail: records all LLM call logs.

[0318] The output deliverables are shown in Table 5 below:

[0319]

[0320] Table 5

[0321] Therefore, this paper uses a technical path of structured classification → LLM intelligent evaluation → dynamic weighted fusion to ensure that industry characteristic indicators are both objectively quantifiable and business-interpretable, making them adaptable to typical scenarios such as manufacturing and finance. For implementation, it is recommended to prioritize enterprise-level LLM frameworks such as NVIDIA NeMo to ensure service stability.

[0322] Preferably, S5, pushing the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data, includes:

[0323] The front-end accesses the enterprise target users' demand information on the required enterprise data assets;

[0324] Calling the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including: cost-oriented keywords, market reference keywords, revenue forecast keywords and / or indicator keywords;

[0325] Constructing data retrieval logic, and generating corresponding asset demand prompt words based on the demand keywords, and inputting the asset demand prompt words into the LLM large language model;

[0326] The LLM large language model retrieves the enterprise data asset data whose data evaluation value and industry characteristic indicators match the demand information from the database according to the prompt word;

[0327] Output the enterprise data asset data and feed it back to the front end.

[0328] Combined with the previous LLM application principle, the specific implementation steps are as follows:

[0329] 1. Front-end access to enterprise target user demand information

[0330] Collect target enterprise users' demand information for the required enterprise data assets through front-end interfaces (such as forms, API interfaces, etc.). This demand information may include descriptions of the data types, application scenarios, and value orientations that users want to obtain.

[0331] 2. Call the LLM large language model to analyze demand keywords

[0332] Use the LLM large language model to analyze the received demand information and extract key demand words. These keywords are divided into the following categories:

[0333] Cost-oriented keywords: such as "collection cost", "maintenance cost", etc.

[0334] Market reference keywords: such as "transaction price", "market valuation", etc.

[0335] Earnings forecast keywords: such as "expected earnings", "economic life", etc.

[0336] Indicator keywords: such as "data integrity", "ROI conversion rate", etc.

[0337] The LLM model uses natural language understanding (NLU) technology to perform semantic analysis on demand information. The extracted keywords will serve as the basis for subsequent retrieval logic.

[0338] 3. Build data retrieval logic and generate asset demand prompts

[0339] Based on the extracted demand keywords, data retrieval logic is constructed. For example, for the "cost-oriented keyword," the system can prioritize data assets related to acquisition costs or maintenance costs; for the "ROI conversion rate" keyword, it can retrieve performance data related to annual revenue per TB.

[0340] Prompt word generation: The system generates corresponding asset requirement prompt words based on the required keywords. For example, "Please search for enterprise data assets whose data evaluation value meets the cost-oriented dimension of 'low maintenance cost' and whose data integrity is higher than 90%."

[0341] The design of prompt words combines the language understanding and reasoning capabilities of the LLM model. The generation of prompt words can be completed through template or dynamic generation.

[0342] 4. Retrieve matching enterprise data asset data from the LLM model

[0343] The generated asset requirement prompt words are input into the LLM large language model, and the model retrieves data assets that meet the conditions from the database based on the prompt words. The search conditions include:

[0344] Data evaluation value (such as cost orientation, market reference, revenue forecast, etc.);

[0345] Industry characteristic indicators (such as the indicators in the IVI-BVI-PVI three-dimensional indicator matrix).

[0346] The search algorithm supports multi-dimensional filtering and weighted matching. The database query results are further screened by the LLM model to ensure matching accuracy.

[0347] 5. Output data assets and feed them back to the front end

[0348] Retrieved enterprise data assets are output and provided to target enterprise users via a front-end interface. Output can be structured data (e.g., JSON / Excel) or visual reports (e.g., radar charts, heat maps, etc.). Output data is anonymized to ensure security. Visualization tools (e.g., Matplotlib, Tableau, etc.) are used to generate intuitive diagnostic reports.

[0349] This invention uses the LLM large language model to deeply analyze demand information, accurately extracting key words from users' core needs. It then combines this with a multi-dimensional evaluation system (such as cost orientation, market reference, and revenue forecast) to match data assets, significantly improving the accuracy of demand matching. Users can quickly access data assets that are highly aligned with their needs, reducing the time and cost of manual screening.

[0350] This technical solution achieves intelligent push and circulation of data assets through dynamic data evaluation value and industry characteristic indicators. It no longer relies on traditional static storage and call models, but instead implements dynamic management based on user needs.

[0351] The circulation efficiency and operational value of data assets have been significantly improved, which can better meet the business needs of enterprises.

[0352] The introduction of the IVI-BVI-PVI three-dimensional indicator evaluation matrix makes the industry-specific indicators of data assets both objectively quantifiable and business-interpretable. For example, internal value (IVI) emphasizes the fundamental attributes of data, business value (BVI) focuses on the data's support for the business, and performance value (PVI) measures the economic benefits of data. This allows enterprises to more clearly understand the actual value and potential uses of data assets, enabling more informed decision-making.

[0353] The present invention can also enhance the flexibility and scalability of the system:

[0354] The technical solution supports dynamic adjustment of weight configurations (such as α, β, and γ values) to adapt to the needs of different industries and scenarios. Furthermore, indicator weights and evaluation logic can be flexibly modified through a JSON configuration file. The system can quickly adapt to the specific needs of various industries, such as manufacturing and finance, and has strong versatility and scalability.

[0355] Therefore, the technical solution of the present invention realizes the efficient management and dynamic circulation of enterprise data assets through a closed-loop process of front-end access, LLM model analysis, data retrieval and intelligent push. Its core advantages are:

[0356] 1. Accurate demand matching: Extract demand keywords through the LLM model to ensure that data assets are highly consistent with user needs.

[0357] 2. Intelligent circulation: Based on a two-dimensional value assessment system and a three-dimensional indicator matrix, it promotes the dynamic management and operation of data assets.

[0358] 3. Strong business adaptability: Dynamic weight configuration and visualization tools enhance the system's flexibility and interpretability. This not only improves the processing and transfer efficiency of data assets, but also builds a more comprehensive industrial ecosystem service system for enterprises.

[0359] like Figure 3 As shown, on the other hand, an enterprise data asset digitization service device for industrial ecological chain management based on AI is provided, and the enterprise data asset digitization service device for industrial ecological chain management based on AI is used to implement the above-mentioned enterprise data asset digitization service method for industrial ecological chain management based on AI, and the device includes:

[0360] The collection end is used to collect the demand information of enterprise target users for the required enterprise data assets through the front end;

[0361] The backend server is configured to call the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including cost-oriented keywords, market reference keywords, revenue forecast keywords, and / or indicator keywords; construct data retrieval logic, and generate corresponding asset demand prompt words based on the demand keywords, and input the asset demand prompt words into the LLM large language model; and the LLM large language model retrieves enterprise data assets from a database based on the prompt words, whose data evaluation values ​​and industry characteristic indicators match the demand information;

[0362] Message middleware is used to push the enterprise data asset data to the enterprise target users.

[0363] Please understand the main interactions and principles of the device in conjunction with the previous method steps.

[0364] Figure 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, such as Figure 4 As shown, the electronic device 410 may include a first processor 2001 .

[0365] Optionally, the electronic device 410 may further include a memory 2002 and a transceiver 2003 .

[0366] The first processor 2001, the memory 2002 and the transceiver 2003 may be connected via a communication bus.

[0367] The following combination Figure 4 The components of the electronic device 410 are described in detail.

[0368] The first processor 2001 is the control center of the electronic device 410 and can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs) or one or more field programmable gate arrays (FPGAs).

[0369] Optionally, the first processor 2001 can execute various functions of the electronic device 410 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0370] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Figure 4 CPU0 and CPU1 are shown in FIG.

[0371] In a specific implementation, as an embodiment, the electronic device 410 may also include multiple processors, such as Figure 4 1 and 2. The first processor 2001 and the second processor 2004 are shown in FIG. Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). A processor herein can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0372] The memory 2002 is used to store the software program for executing the solution of the present invention, and is controlled by the first processor 2001 for execution. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0373] Alternatively, the memory 2002 may be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, an optical disc storage (including a compact disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 may be integrated with the first processor 2001 or exist independently and accessed through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0374] The transceiver 2003 is used to communicate with a network device or a terminal device.

[0375] Optionally, the transceiver 2003 may include a receiver and a transmitter ( Figure 4The receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0376] Optionally, the transceiver 2003 may be integrated with the first processor 2001, or may exist independently and communicate with the first processor 2001 through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0377] It should be noted that Figure 4 The structure of the electronic device 410 shown in the figure does not constitute a limitation on the router. The actual knowledge structure recognition device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0378] In addition, the technical effects of the electronic device 410 can refer to the technical effects of the enterprise data asset digitalization service method for industrial ecosystem management based on AI described in the above method embodiment, and will not be repeated here.

[0379] It should be understood that the first processor 2001 in the embodiment of the present invention may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor, or the processor may be any conventional processor, etc.

[0380] It should also be understood that the memory in the embodiments of the present invention may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0381] The above embodiments can be implemented in whole or in part via software, hardware (e.g., circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0382] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0383] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0384] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0385] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0386] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0387] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of the device or unit, which can be electrical, mechanical or other forms.

[0388] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0389] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0390] If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical disks.

[0391] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. An enterprise data asset digital service method for industrial ecosystem management based on AI, characterized by: The method comprises: S1. Build an asset evaluation system that integrates cost-oriented evaluation, market reference evaluation, and income forecast evaluation; S2. Based on the LLM large language model training, the evaluation mechanism of the asset organic integration evaluation system is learned to generate an LLM-asset organic evaluation model; S3. Collect enterprise data asset data generated by the enterprise in real time, and input the enterprise data asset data into the LLM-asset organic valuation model to perform asset organic valuation, thereby generating a data valuation value for the enterprise data asset data; S4. Analyze the industry characteristic indicators of the enterprise data asset data based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix; S5. Push the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data.

2. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 1 is characterized in that: The evaluation model of the cost-oriented evaluation is as follows: ， in: : The cost of collecting and maintaining enterprise data assets; TC‌: total acquisition cost; ‌MC‌: annual maintenance cost; ‌ROIC‌: rate of cost of capital; ‌U‌: Use premium factor.

3. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 1 is characterized in that: The evaluation model of the market reference evaluation is as follows: ， in: : The current market reference price of enterprise data assets; ‌: The transaction price of the reference asset; ‌: Adjustment coefficient of the i-th dimension (5 dimensions: technology maturity, scenario adaptability, value density, technology correction coefficient and timeliness index).

4. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 1 is characterized in that: The evaluation model for the revenue forecast evaluation is as follows: ， in: NPV is the expected return on an enterprise’s data assets; R t ‌: expected return in year t; ‌λ t ‌: risk attenuation factor; ‌r‌: industry discount rate; NPV uses LSTM to predict the economic life cycle and expected returns of enterprise data assets.

5. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 1 is characterized in that: The IVI-BVI-PVI three-dimensional indicator evaluation matrix includes: The IVI matrix unit includes the following internal industry characteristic indicators: The field missing rate used to characterize data completeness, The error rate used to characterize the accuracy of the data, Cross-system conflict rate used to characterize data consistency; The BVI Matrix unit includes the following business industry characteristic indicators: The Spearman correlation coefficient with the core KPI used to characterize the strategic support, The proportion of covered business process nodes used to characterize the scenario penetration rate, Data update delay time used to characterize timeliness; The PVI matrix unit includes the following performance industry characteristic indicators: The annual revenue per TB of data used to represent the ROI conversion rate, The proportion of new products driven by innovation is used to characterize the contribution of innovation. The size of the dominant data alliance (number of nodes ≥ 50) used to characterize ecological control power.

6. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 4 is characterized in that: S4. Based on the constructed IVI-BVI-PVI three-dimensional indicator evaluation matrix, analyze the industry characteristic indicators of the enterprise data asset data, including: Parsing the enterprise data asset data and classifying it into internal asset data, business asset data, and performance asset data; Invoking the LLM large language model to calculate and analyze the internal value, business value, and performance value of the internal asset data, business asset data, and performance asset data based on the matrix cells of the IVI-BVI-PVI three-dimensional indicator evaluation matrix; The internal value, business value and performance value are weighted and summed to generate the industry characteristic indicators of the enterprise data asset data.

7. The enterprise data asset digital service method for industrial ecosystem management based on AI according to claim 1 is characterized in that: S5. Pushing the enterprise data asset data to target enterprise users based on the data evaluation value and industry characteristic indicators of the enterprise data asset data, including: The front-end accesses the enterprise target users' demand information on the required enterprise data assets; Calling the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including: cost-oriented keywords, market reference keywords, revenue forecast keywords and / or indicator keywords; Constructing data retrieval logic, and generating corresponding asset demand prompt words based on the demand keywords, and inputting the asset demand prompt words into the LLM large language model; The LLM large language model retrieves the enterprise data asset data whose data evaluation value and industry characteristic indicators match the demand information from the database according to the prompt word; Output the enterprise data asset data and feed it back to the front end.

8. An enterprise data asset digitization service device for industrial ecosystem management based on AI, the enterprise data asset digitization service device for industrial ecosystem management based on AI being used to implement the enterprise data asset digitization service method for industrial ecosystem management based on AI as described in any one of claims 1 to 7, characterized in that: The device comprises: The collection end is used to collect the demand information of enterprise target users for the required enterprise data assets through the front end; The backend server is configured to call the LLM large language model to analyze the demand keywords for the required enterprise data assets in the demand information, including cost-oriented keywords, market reference keywords, revenue forecast keywords, and / or indicator keywords; construct data retrieval logic, and generate corresponding asset demand prompt words based on the demand keywords, and input the asset demand prompt words into the LLM large language model; and the LLM large language model retrieves enterprise data assets from a database based on the prompt words, whose data evaluation values ​​and industry characteristic indicators match the demand information; Message middleware is used to push the enterprise data asset data to the enterprise target users.

9. An electronic device, characterized in that: The electronic device comprises: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program code, which can be called by a processor to execute the method according to any one of claims 1 to 7.

Images