Intelligent data asset evaluation system and method based on multi-dimensional data analysis

Abstract

The invention relates to the technical field of data asset assessment, and particularly discloses a data asset intelligent assessment system and method based on multi-dimensional data analysis, and the system comprises an association network construction module which constructs a data asset multi-dimensional dynamic association network based on the multi-dimensional features of all data assets in a plurality of time steps; an evaluation context capturing module captures a multi-dimensional evaluation context under a current evaluation scene label in the data asset multi-dimensional dynamic association network based on an evaluation strategy; an asset static evaluation module determines static value evaluation results of all data assets under the current evaluation scene label based on the multi-dimensional evaluation context under the current evaluation scene label; a variation factor determination module determines a spatio-temporal evolution variation factor of each data asset based on the multi-context time sequence multi-dimensional dissimilatory feature set; the data asset comprehensive evaluation module generates a data asset evaluation result based on the spatio-temporal evolution variation factors and the static value evaluation results of all the data assets; and a reliable basis is provided for management decision of data assets.

Description

Technical Field

[0001] This invention relates to the field of asset valuation technology, and in particular to a data asset intelligent valuation system and method based on multi-dimensional data analysis. Background Technology

[0002] In today's digital age, data has become one of the most critical assets for enterprises and organizations. With the rapid development of information technology, the amount of data accumulated across industries has exploded, and the types of data have become increasingly diverse, encompassing structured, semi-structured, and unstructured data. These data assets contain enormous commercial value and strategic significance, playing a decisive role in enterprise decision-making, market competition, and business innovation. Accurately assessing the value of data assets has become crucial. Through scientific and reasonable data asset assessment, enterprises can clearly understand the status of their data assets, identify which data has high value, and which data needs further mining or cleaning, thereby achieving optimized allocation and effective utilization of data assets. This not only helps improve operational efficiency and reduce costs but also creates new business opportunities and enhances core competitiveness. With the continuous development of big data, artificial intelligence, and other technologies, intelligent data asset assessment systems and methods based on multi-dimensional data analysis have emerged, possessing broad application prospects. This system can fully utilize advanced data analysis technologies to conduct in-depth analysis and assessment of data assets from multiple dimensions, breaking the limitations of traditional single-dimensional assessments, providing enterprises with more comprehensive, accurate, and intelligent data asset assessment results, and promoting deeper development of enterprise digital transformation.

[0003] However, existing data asset assessment technologies cannot comprehensively and dynamically represent the complex relationships between data assets, nor can they accurately capture the multi-dimensional assessment context based on different assessment scenario labels and corresponding assessment strategies, resulting in a lack of targeted assessment. Regarding the changes in data assets over time and space, existing technologies struggle to reflect the dynamic changes in data assets, thus failing to generate comprehensive and accurate data asset assessment results and failing to meet enterprises' needs for the scientific management and effective utilization of data assets.

[0004] Therefore, this invention proposes a data asset intelligent evaluation system and method based on multi-dimensional data analysis. Summary of the Invention

[0005] This invention provides an intelligent data asset assessment system and method based on multi-dimensional data analysis. Leveraging the multi-dimensional features of all data assets across multiple time steps, it constructs a multi-dimensional dynamic correlation network of data assets, presenting a panoramic view of the dynamic relationships between them and laying a comprehensive information foundation for assessment. By identifying the current assessment scenario label, and based on the corresponding assessment strategy, it accurately captures the multi-dimensional assessment context within the aforementioned network, ensuring the assessment closely aligns with the specific scenario requirements. Using the captured context, it determines the static value assessment results of all data assets in the current scenario, providing a basic value measurement. It generates a multi-contextual, time-series, multi-dimensional variation feature set, thereby determining the spatiotemporal evolution variation factors of each data asset, fully considering the changes of assets over time and space. By integrating the spatiotemporal evolution variation factors and the static value assessment results, it generates comprehensive and accurate data asset assessment results, providing a solid basis for the scientific management and rational decision-making of data assets, and helping enterprises to deeply explore the potential value of data assets.

[0006] This invention provides a data asset intelligent assessment system based on multi-dimensional data analysis, comprising: The association network construction module is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps. The assessment context capture module is used to identify the current assessment scenario label and capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets based on the assessment strategy of the current assessment scenario label. The asset static valuation module is used to determine the static value valuation results of all data assets under the current valuation scenario label based on the multi-dimensional valuation context under the current valuation scenario label. The mutation factor determination module is used to generate a multi-contextual temporal multi-dimensional mutation feature set for all data assets under the current evaluation scenario label, and to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multi-dimensional mutation feature set. The data asset comprehensive evaluation module is used to generate data asset evaluation results for all data assets under the current evaluation scenario label, based on the spatiotemporal evolution and variation factors and static value evaluation results of all data assets.

[0007] Preferably, the associated network construction module includes: The multidimensional feature acquisition submodule is used to collect the data scale, update frequency, and access popularity of all data assets in real time at multiple time steps, and to measure the associated data volume and business impact of all data assets. The data scale, update frequency, access popularity, associated data volume, and business impact of all data assets at each time step are used as the multidimensional features of all data assets at each time step. The edge weight determination submodule is used to calculate the edge weight of the corresponding two data assets at each time step based on the information gain rate and co-occurrence frequency between each pair of data assets at each time step. The association network construction submodule is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps and the edge weights of every two data assets at multiple time steps.

[0008] Preferably, the evaluation context capture module includes: The scene tag recognition submodule is used to match the core elements in the user evaluation request with the feature word library of multiple preset scene tags and determine all matching scene tags and their corresponding matching degrees. Based on the matching degree, it is determined whether there are obviously matching scene tags. If so, the obviously matching scene tag is used as the pre-evaluation scene tag. Otherwise, the semantic similarity between the user evaluation request and the feature word library of each preset scene tag is calculated, and the preset scene tag with the maximum semantic similarity is used as the pre-evaluation scene tag. The module involves a node capture submodule, which is used to identify all core nodes in the multidimensional dynamic correlation network of data assets based on the evaluation strategy of the current evaluation scenario label, and to determine the starting point of the network among all core nodes based on the importance of all core nodes. The node multi-hop expansion submodule is used to perform multi-hop expansion along the association edges in the multi-dimensional dynamic association network of data assets based on the network starting point and expansion rules, so as to obtain all valid node paths. The effective path reorganization submodule is used to reorganize all effective node paths according to the evaluation dimensions of the evaluation strategy of the current evaluation scenario label, so as to obtain a weighted multidimensional evaluation context under the current evaluation scenario label.

[0009] Preferably, the static asset valuation module includes: The multidimensional context decomposition submodule is used to determine the basic attribute indicator evaluation context, business value indicator evaluation context, and strong correlation indicator evaluation context under the current evaluation scenario label based on the multidimensional evaluation context under the current evaluation scenario label, and to determine the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The first indicator quantification submodule is used to calculate the basic attribute evaluation indicators, business value indicators, and correlation strength indicators of all data assets under the current evaluation scenario label, based on the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The static value assessment submodule is used to determine the comprehensive weight of each indicator type based on the multi-dimensional assessment context under the current assessment scenario label. Based on the comprehensive weight of all indicator types, it assesses the basic attribute indicators, business value indicators, and correlation strength indicators of all data assets under the current assessment scenario label, and obtains the static value assessment results of all data assets under the current assessment scenario label.

[0010] Preferably, the variation factor determination module includes: The alienation feature generation submodule is used to perform alienation feature analysis based on the multidimensional assessment context under the current assessment scenario label and generate a multi-context temporal multidimensional alienation feature set for all data assets under the current assessment scenario label. The mutation factor determination submodule is used to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multidimensional mutation feature set.

[0011] Preferably, the alienation feature generation submodule includes: The second indicator quantification submodule is used to determine the basic attribute indicator set of the basic attribute indicator assessment network, the business value indicator set of the business value indicator assessment network, and the correlation strength indicator set of the strong correlation indicator assessment network based on the multi-dimensional assessment network under the current assessment scenario label. The Single Pathway Differentiation Analysis Submodule is used to perform single-pathway differentiation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation path, the business value indicator set of the business value indicator evaluation path, and the correlation strength indicator set of the strong correlation indicator evaluation path. It obtains the short-term fluctuation differentiation value, long-term trend differentiation value, and threshold breakthrough differentiation value of each data asset at each time step. The Interactive Alienation Analysis submodule is used to perform cross-network interactive alienation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation network, the business value indicator set of the business value indicator evaluation network, and the correlation strength indicator set of the strong correlation indicator evaluation network, to obtain the network collaborative alienation value and the dominant network influence alienation value between each pair of networks. The alienation feature determination submodule is used to generate a multi-dimensional alienation feature set of all data assets under the current evaluation scenario label based on the short-term fluctuation alienation value, long-term trend alienation value, threshold breakthrough alienation value, and alienation value of the synergistic alienation between pairs of threads and the alienation value of the dominant thread at each time step.

[0012] Preferably, the mutation factor determination submodule includes: The weight allocation unit is used to assign weights to each alienation feature in the multi-context temporal multidimensional alienation feature set based on the current evaluation scenario label. The alienation synthesis unit is used to calculate the synthesis alienation value of each data asset at each time step based on the weight of each alienation feature in the multi-context temporal multi-dimensional alienation feature set. The heterogeneity unit is used to standardize and trend-fit the comprehensive heterogeneity value of each data asset at each time step to obtain the spatiotemporal evolution heterogeneity factor of each data asset.

[0013] Preferably, the heterogeneous unit includes: The standardized sub-unit is used to standardize the comprehensive alienation value of each data asset at each time step, so as to obtain the standard comprehensive alienation value of each data asset at each time step. The linear fitting subunit is used to perform linear fitting on the standard comprehensive alienation value of each data asset at a preset number of time steps and determine the slope as the spatiotemporal evolution alienation factor of each data asset.

[0014] The preferred data asset comprehensive assessment module includes: The Fusion Weight Definition submodule is used to define the fusion weights of spatiotemporal evolution variation factors and static value assessment results; The asset fusion assessment submodule is used to generate data asset assessment results for all data assets under the current assessment scenario label based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, as well as the fusion weights of the spatiotemporal evolution variation factors and static value assessment results.

[0015] This invention provides a data asset intelligent assessment method based on multi-dimensional data analysis, comprising: Construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps; Identify the current assessment scenario label, and based on the assessment strategy of the current assessment scenario label, capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets; Based on the multi-dimensional assessment framework under the current assessment scenario label, determine the static value assessment results of all data assets under the current assessment scenario label; Generate a multi-contextual, temporal, and multi-dimensional alienation feature set for all data assets under the current evaluation scenario label, and determine the spatiotemporal evolution and alienation factor of each data asset based on the multi-contextual, temporal, and multi-dimensional alienation feature set. Based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, the data asset assessment results of all data assets under the current assessment scenario label are generated.

[0016] The beneficial effects of this invention compared to existing technologies are as follows: By leveraging the multidimensional features of all data assets across multiple time steps, a multidimensional dynamic correlation network of data assets is constructed, presenting a panoramic view of the dynamic relationships between data assets and laying a comprehensive information foundation for assessment. By identifying the current assessment scenario label, and based on the corresponding assessment strategy, the multidimensional assessment context is accurately captured within the aforementioned network, ensuring that the assessment closely aligns with the specific scenario requirements. Using the captured context, the static value assessment results of all data assets in the current scenario are determined, providing a basic value measurement. A multi-contextual temporal multidimensional variation feature set is generated, thereby determining the spatiotemporal evolution variation factors of each data asset, fully considering the changes of assets over time and space. By integrating the spatiotemporal evolution variation factors and the static value assessment results, a comprehensive and accurate data asset assessment result is generated, providing a solid basis for the scientific management and rational decision-making of data assets, and helping enterprises to deeply explore the potential value of data assets.

[0017] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a data asset intelligent evaluation system based on multi-dimensional data analysis in an embodiment of the present invention; Figure 2 This is a schematic diagram of the network construction module in an embodiment of the present invention; Figure 3 This is a schematic diagram of the evaluation context capture module in an embodiment of the present invention; Figure 4 This is a schematic diagram of the asset static valuation module in an embodiment of the present invention; Figure 5 This is a schematic diagram of the variation factor determination module in an embodiment of the present invention; Figure 6 This is a schematic diagram of the data asset comprehensive evaluation module in an embodiment of the present invention. Detailed Implementation

[0020] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0021] like Figure 1 As shown, this invention provides an implementation method for a data asset intelligent assessment system based on multi-dimensional data analysis, comprising: The association network construction module is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps. The assessment context capture module is used to identify the current assessment scenario label and capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets based on the assessment strategy of the current assessment scenario label. The asset static valuation module is used to determine the static value valuation results of all data assets under the current valuation scenario label based on the multi-dimensional valuation context under the current valuation scenario label. The mutation factor determination module is used to generate a multi-contextual temporal multi-dimensional mutation feature set for all data assets under the current evaluation scenario label, and to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multi-dimensional mutation feature set. The data asset comprehensive evaluation module is used to generate data asset evaluation results for all data assets under the current evaluation scenario label, based on the spatiotemporal evolution and variation factors and static value evaluation results of all data assets.

[0022] In this embodiment, data assets refer to various types of structured, semi-structured, and unstructured data accumulated by enterprises and organizations. For example, a company's customer information data, sales record data, and production process data all fall under the category of data assets.

[0023] In this embodiment, multidimensional features refer to the characteristics of data assets described from multiple aspects, including the data scale (data size), update frequency (how frequently data is updated), access popularity (how frequently data is accessed), as well as the measured amount of associated data (how much data is associated with the data asset) and business impact (the degree to which it affects the business of the enterprise).

[0024] In this embodiment, the multidimensional dynamic association network of data assets is constructed based on the multidimensional features of all data assets at multiple time steps and the edge weights between each pair of data assets at each time step. It presents a panoramic view of the dynamic relationships between data assets, where nodes represent data assets, edges represent the associations between data assets, and edge weights reflect the degree of association between two data assets at each time step.

[0025] In this embodiment, the current evaluation scenario label is used to identify a specific data asset evaluation scenario. Different evaluation scenario labels correspond to different evaluation needs, such as data asset transaction scenarios and data asset internal optimization scenarios. Each scenario label has its unique evaluation focus and strategy. For example, under the data asset transaction scenario label, the evaluation may focus more on the market value and potential benefits of the data asset; while under the internal optimization scenario label, the focus may be on analyzing the degree to which the data asset supports the enterprise's business processes.

[0026] In this embodiment, the evaluation strategy for the current evaluation scenario label consists of a series of evaluation methods and rules formulated for the current evaluation scenario label. Operations are performed within the multi-dimensional dynamic relational network of data assets based on this strategy, such as identifying core nodes, determining the starting point of the network, performing multi-hop node expansion, and reorganizing effective paths. For example, under the data asset investment value evaluation scenario label, the evaluation strategy might stipulate prioritizing the identification of data asset nodes related to investment decisions, determining the starting point of the network based on the importance of the nodes to the investment decision, and then expanding the node paths along the relational edges according to specific rules to obtain a multi-dimensional evaluation network reflecting the investment value of the data assets.

[0027] In this embodiment, the multi-dimensional evaluation network under the current evaluation scenario label is a weighted network structure retrieved from the multi-dimensional dynamic association network of data assets based on the evaluation strategy of the current evaluation scenario label. This structure is reorganized from the node paths in the association network according to the evaluation dimensions of the evaluation strategy, reflecting data asset information related to the current evaluation scenario from multiple dimensions. For example, in a specific evaluation scenario, the multi-dimensional evaluation network may include descriptions of the data assets from multiple dimensions such as the basic attributes, business value, and strong association dimensions. This information is presented through the relationships and weights of nodes and edges in the network.

[0028] In this embodiment, the static value assessment results of all data assets under the current assessment scenario label are used to measure the value of data assets in the current assessment scenario from a static perspective. For example, for a company's customer data assets, by analyzing their basic attribute indicators (such as data accuracy and completeness), business value indicators (such as contribution to customer conversion), and strong correlation indicators (such as the degree of correlation with other key business data) in the current assessment scenario, and combining the weights of each indicator, the static value assessment results are calculated to help the company understand the value of the customer data asset in the current scenario.

[0029] In this embodiment, the multi-context temporal multidimensional alienation feature set is based on a multidimensional evaluation context. It is formed by performing single-context and cross-context alienation feature analysis on different indicator sets, obtaining multiple alienation values ​​of each data asset at each time step, and reflecting the multidimensional changes of data assets over time.

[0030] In this embodiment, the spatiotemporal evolution variation factor of data assets is a numerical value that reflects the dynamic changes of data assets in the spatiotemporal dimension.

[0031] In this embodiment, the data asset assessment result is a result that integrates spatiotemporal evolution variation factors and static value assessment results, representing the value of the data asset.

[0032] like Figure 2 As shown, in order to construct a multidimensional dynamic association network of data assets that reflects the dynamic relationships of data assets by collecting specific multidimensional features of data assets at multiple time steps and calculating edge weights, a further association network construction module is proposed, including: The multidimensional feature acquisition submodule is used to collect the data scale, update frequency, and access popularity of all data assets in real time at multiple time steps, and to measure the associated data volume and business impact of all data assets. The data scale, update frequency, access popularity, associated data volume, and business impact of all data assets at each time step are used as the multidimensional features of all data assets at each time step. The edge weight determination submodule is used to calculate the edge weight of the corresponding two data assets at each time step based on the information gain rate and co-occurrence frequency between each pair of data assets at each time step. The association network construction submodule is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps and the edge weights of every two data assets at multiple time steps.

[0033] In this embodiment, real-time collection of data scale, update frequency, and access popularity of all data assets at multiple time steps refers to continuously acquiring the data volume, update frequency, and access frequency of each data asset at different points in time, thereby characterizing the state of the data assets at different times from multiple dimensions. For example, real-time monitoring of the daily (time step) growth scale, daily data update frequency, and daily query frequency of user purchase record data on an e-commerce platform.

[0034] In this embodiment, the amount of associated data and business impact of all data assets are measured. This involves assessing the amount of data associated with each data asset and its impact on the company's actual business operations. For example, analyzing customer information data assets involves assessing the amount of order data associated with them and the degree to which this customer information data influences sales decision-making.

[0035] In this embodiment, the information gain rate and co-occurrence frequency between each pair of data assets at each time step are considered. The information gain rate measures the degree to which the uncertainty about the other data asset is reduced when information about one data asset is known. For example, in customer information data and purchase behavior data, knowing the customer age distribution (one data asset) reduces the uncertainty of purchase behavior (the other data asset), which is the information gain. The information gain rate is normalized by dividing the information gain by the inherent information entropy of the affected data asset (which measures the degree of disorder in information). In this way, the obtained information gain rate can more reasonably reflect the relative strength of information transmission and dependence between two data assets, and is used to assess the tightness of their association. The co-occurrence frequency refers to how frequently two data assets appear together at the same time step. For example, in financial data, the information gain rate of customer credit rating data and loan record data at each month (time step), and the frequency with which they appear together each month.

[0036] In this embodiment, the edge weights of the two data assets at each time step are calculated based on the information gain rate and co-occurrence frequency between them. These two indicators are used to determine the degree of association between the two data assets at each time step through a specific calculation method. For example, the edge weights are obtained by weighting the information gain rate and co-occurrence frequency; a higher value indicates a stronger association between the two data assets at that time step.

[0037] In this embodiment, a multidimensional dynamic association network of data assets is constructed based on the multidimensional features of all data assets at multiple time steps and the edge weights of every two data assets at multiple time steps. This means that data assets are used as nodes, the characteristics of each node are described according to multidimensional features, and the edge weights represent the degree of association between nodes, thus building a network structure that can reflect the dynamic relationship between data assets.

[0038] like Figure 3 As shown, in order to determine the pre-evaluation scenario label through matching and judgment, and to capture core nodes in the association network, perform multi-hop expansion and path reorganization according to the evaluation strategy to obtain a weighted multi-dimensional evaluation context under the current evaluation scenario label, an evaluation context capture module is further proposed, including: The scene tag recognition submodule is used to match the core elements in the user evaluation request with the feature word library of multiple preset scene tags and determine all matching scene tags and their corresponding matching degrees. Based on the matching degree, it is determined whether there are obviously matching scene tags. If so, the obviously matching scene tag is used as the pre-evaluation scene tag. Otherwise, the semantic similarity between the user evaluation request and the feature word library of each preset scene tag is calculated, and the preset scene tag with the maximum semantic similarity is used as the pre-evaluation scene tag. The module involves a node capture submodule, which is used to identify all core nodes in the multidimensional dynamic correlation network of data assets based on the evaluation strategy of the current evaluation scenario label, and to determine the starting point of the network among all core nodes based on the importance of all core nodes. The node multi-hop expansion submodule is used to perform multi-hop expansion along the association edges in the multi-dimensional dynamic association network of data assets based on the network starting point and expansion rules, so as to obtain all valid node paths. The effective path reorganization submodule is used to reorganize all effective node paths according to the evaluation dimensions of the evaluation strategy of the current evaluation scenario label, so as to obtain a weighted multidimensional evaluation context under the current evaluation scenario label.

[0039] In this embodiment, a user assessment request refers to a user's request for the valuation of data assets, including requirements regarding the assessment objectives, scope, and focus. For example, a company may submit a request to the assessment system to assess the market value of its data assets for a transaction.

[0040] In this embodiment, the core elements of a user's evaluation request are the key information points within the request, such as the purpose of the evaluation and the type of data assets involved. For example, in the aforementioned transaction evaluation request, the core elements might be the specific category of data assets and the desired valuation direction.

[0041] In this embodiment, the feature word library for various preset scenario labels is pre-defined and contains a set of feature words related to different evaluation scenarios. For example, for the data asset investment scenario, the word library may contain words such as "expected returns" and "risk assessment".

[0042] In this embodiment, the core elements of the user's evaluation request are matched with feature word libraries of various preset scenario tags, and all matching scenario tags and their corresponding matching degrees are determined. That is, the key information of the request is compared with each scenario word library to see which scenario tags fit and the degree of fit. For example, in an investment evaluation request, "expected returns" matches with the investment scenario word library, and the matching degree is calculated.

[0043] In this embodiment, the existence of a clearly matching scene tag is determined based on the matching degree. If the matching degree reaches a certain standard, it is considered that a clearly matching scene exists. For example, an average matching degree of more than 80% between multiple core elements and a feature word library of a certain preset scene tag is considered a clear match; if it is higher than this, the scene tag is determined.

[0044] In this embodiment, calculating the semantic similarity between the user's evaluation request and the feature lexicon of each preset scenario tag is a semantic analysis of the similarity between the request and each scenario lexicon to more accurately determine the appropriate scenario. For example, natural language processing techniques can be used to calculate semantic similarity.

[0045] In this embodiment, the evaluation strategy based on the current evaluation scenario label identifies all core nodes in the multi-dimensional dynamic association network of data assets. That is, according to the specific evaluation scenario strategy, the data asset nodes that are important to the evaluation are found in the network. For example, in an investment scenario, data asset nodes related to returns and risks are identified.

[0046] In this embodiment, the starting point of the network is determined from all core nodes based on their importance. This involves selecting a starting node from among the identified important nodes according to their degree of importance. For example, in an investment scenario, core nodes are ranked according to their impact on investment decisions, and the most important one is selected as the starting point of the network.

[0047] In this embodiment, the expansion rules are based on criteria for extending paths from the starting point of the network, such as limiting the type of extended edges and the number of hops. For example, it is stipulated that expansion can only be carried out along strongly related edges, with a maximum of 3 hops.

[0048] In this embodiment, multi-hop expansion is performed along the associated edges in the multi-dimensional dynamic association network of data assets based on the network starting point and expansion rules to obtain all valid node paths. That is, extending along the network from the starting point according to the rules yields a series of valid paths. For example, expanding from a selected starting point according to the rules results in multiple paths containing different data asset nodes. Assume there is a multi-dimensional dynamic association network of e-commerce data assets, where nodes represent different data assets, such as user information, product information, order information, logistics information, etc., and the edges between nodes represent the association relationships between data assets. Now, a network starting point is determined to be the "user information" data asset. The expansion rules are set as follows: expansion can only be performed along the associated edges directly related to sales business, and the maximum expansion is 3 hops. Starting from the "user information" starting point, the first hop may extend along the associated edge to the "order information" node, because a user places an order, which is a direct sales business association. The second hop, starting from "order information," can extend to the "product information" node according to the rules, because the order contains information about the purchased products, which is also related to sales business. The third step extends from "product information" to the "logistics information" node, since goods require logistics and delivery after sale, which also falls under the scope of sales operations. This yields an effective node path: User Information - Order Information - Product Information - Logistics Information.

[0049] Following the same rules, other paths can be obtained, such as extending directly from "user information" to "product browsing history" (first hop), then from "product browsing history" to "product recommendation information" (second hop), and finally from "product recommendation information" to "purchase conversion rate information" (third hop), which is another effective node path: user information - product browsing history - product recommendation information - purchase conversion rate information. Through this multi-hop extension method, a series of effective paths containing different data asset nodes can be obtained.

[0050] In this embodiment, all valid node paths are reorganized according to the evaluation dimensions of the evaluation strategy for the current evaluation scenario label to obtain a weighted multidimensional evaluation network under the current evaluation scenario label. This involves reorganizing the paths based on the evaluation strategy dimensions and assigning weights to each path or node to form a multidimensional evaluation network. For example, a series of valid node paths have already been obtained through multi-hop expansion, such as: Path 1: User Information - Purchase Frequency Information - Purchase Amount Information; Path 2: Product Information - Inventory Turnover Information - Sales Channel Information; Path 3: User Information - Return Rate Information - Customer Satisfaction Information.

[0051] For the user value evaluation dimension, paths 1 and 3 are highly relevant. Purchase frequency and purchase amount in path 1 directly reflect a user's spending power and value, while return rate and customer satisfaction in path 3 indirectly reflect a user's recognition of the platform and its long-term value. Therefore, in this dimension, path 1 is assigned a weight of 0.6, and path 3 is assigned a weight of 0.4.

[0052] For the sales efficiency assessment dimension, path 2 is more relevant, as inventory turnover rate and sales channel information directly affect sales efficiency. Therefore, path 2 is assigned a weight of 1.

[0053] After restructuring, under the user value dimension, path 1 and path 3 constitute the evaluation framework for this dimension; under the sales efficiency dimension, path 2 constitutes the evaluation framework, each with its corresponding weight. This forms a weighted, multi-dimensional evaluation framework under the current e-commerce evaluation scenario tags, enabling a more comprehensive assessment of the correlation and importance between data assets and evaluation objectives from different dimensions.

[0054] like Figure 4 As shown, in order to deconstruct the multi-dimensional assessment framework, quantify different indicator sets to calculate each indicator, and then combine the comprehensive weights of each indicator type to obtain the static value assessment results of all data assets under the current assessment scenario label, a static asset assessment module is further proposed, including: The multidimensional context decomposition submodule is used to determine the basic attribute indicator evaluation context, business value indicator evaluation context, and strong correlation indicator evaluation context under the current evaluation scenario label based on the multidimensional evaluation context under the current evaluation scenario label, and to determine the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The first indicator quantification submodule is used to calculate the basic attribute evaluation indicators, business value indicators, and correlation strength indicators of all data assets under the current evaluation scenario label, based on the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The static value assessment submodule is used to determine the comprehensive weight of each indicator type based on the multi-dimensional assessment context under the current assessment scenario label. Based on the comprehensive weight of all indicator types, it assesses the basic attribute indicators, business value indicators, and correlation strength indicators of all data assets under the current assessment scenario label, and obtains the static value assessment results of all data assets under the current assessment scenario label.

[0055] In this embodiment, based on the multi-dimensional evaluation context under the current evaluation scenario label, the evaluation contexts for basic attribute indicators, business value indicators, and strong correlation indicators are determined. That is, from the multi-dimensional evaluation context, based on the requirements of the evaluation scenario, contexts reflecting the basic attributes of data assets, business value, and the degree of correlation with other assets are respectively divided. For example, in an e-commerce data evaluation scenario, the evaluation context for basic attribute indicators might revolve around data accuracy and completeness; the evaluation context for business value indicators might revolve around sales contribution; and the evaluation context for strong correlation indicators might revolve around correlation with logistics data. Assuming a data asset evaluation scenario in social media: Basic Attribute Indicator Assessment Framework: The basic attributes of social media data relate to the quality characteristics of the data itself. For example, the authenticity of the data; false user information severely impacts the value of the data. And the timeliness of the data, such as the time of a user's most recent post. These constitute the basic attribute indicator assessment framework, focusing on aspects like data authenticity and timeliness, used to evaluate the fundamental quality of data assets.

[0056] Business Value Metric Assessment Framework: Business value reflects the contribution of data to the core business of social media. For example, user activity; highly active users bring more traffic and interaction, significantly enhancing the platform's commercial value. Another example is advertising revenue contribution, measured by analyzing the contribution of different user groups to ad clicks and conversions. This forms the business value metric assessment framework, focusing on user activity, advertising revenue contribution, etc., to evaluate the value of data assets to the business.

[0057] The framework for evaluating strong correlation indicators: Social media data is closely linked to other relevant data. For example, the correlation with third-party partner platforms, such as e-commerce platforms, is crucial; if social media can accurately push product links to potential buyers, this correlation is an important indicator. There's also the correlation with content creation platforms, assessing the dissemination and interaction of high-quality content across different platforms. These correlations with third-party platforms and content creation platforms form the framework for evaluating strong correlation indicators, reflecting the degree of connection between data assets and other assets.

[0058] In this embodiment, the basic attribute indicator set for the basic attribute indicator evaluation framework, the business value indicator set for the business value indicator evaluation framework, and the correlation strength indicator set for the strong correlation indicator evaluation framework are determined. This clarifies the specific data indicator sets used for evaluation within each framework. For example, the basic attribute indicator set may include data error rate and missing rate; the business value indicator set may include indicators related to user conversion rate and repurchase rate improvement; and the correlation strength indicator set may include quantifiable indicators related to specific associated data.

[0059] In this embodiment, based on the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context under the current evaluation scenario label, the basic attribute evaluation indicators, business value indicators, and correlation strength indicators of all data assets under the current evaluation scenario label are calculated respectively. That is, based on the indicators within each indicator set, the corresponding indicators are calculated for each data asset. For example, basic attribute evaluation indicators are calculated based on data error rate and missing rate; business value indicators are calculated based on user conversion rate, etc.; and correlation strength indicators are calculated based on correlation quantification indicators.

[0060] In this embodiment, the comprehensive weight of each indicator type is determined based on the multi-dimensional evaluation context under the current evaluation scenario label. This is done by determining the proportion of importance of different indicator types, such as basic attributes, business value, and correlation strength, in the overall evaluation according to the evaluation scenario. For example, in some scenarios, the weight of the business value indicator may be set to 0.5, the weight of the basic attribute indicator to 0.3, and the weight of the correlation strength indicator to 0.2.

[0061] In this embodiment, based on the comprehensive weights of all indicator types, the basic attribute evaluation indicators, business value indicators, and association strength indicators of all data assets under the current evaluation scenario label are used to obtain the static value evaluation results of all data assets under the current evaluation scenario label. That is, using the determined weights of each indicator, the different indicators of each data asset are weighted and summed to obtain the static value evaluation results of the data assets under this scenario. For example, if the basic attribute evaluation indicator of a certain data asset is 80 (obtained by weighted summation of data error rate and missing rate), the business value indicator is 90 (obtained by weighted summation of all indicators it contains), and the association strength indicator is 70 (obtained by weighted summation of all indicators it contains), with corresponding weights of 0.3, 0.5, and 0.2 respectively, the static value evaluation result is 80×0.3+90×0.5+70×0.2=83.

[0062] like Figure 5 As shown, in order to realize alienation feature analysis based on multidimensional evaluation context, generate a multi-context temporal multidimensional alienation feature set, and determine the spatiotemporal evolution alienation factor of each data asset, a further alienation factor determination module is proposed, including: The alienation feature generation submodule is used to perform alienation feature analysis based on the multidimensional assessment context under the current assessment scenario label and generate a multi-context temporal multidimensional alienation feature set for all data assets under the current assessment scenario label. The mutation factor determination submodule is used to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multidimensional mutation feature set.

[0063] To determine different indicator sets and perform single-context and cross-context alienation feature analysis, a multi-context temporal multi-dimensional alienation feature set for all data assets under the current assessment scenario label is generated based on the analysis results. Furthermore, a alienation feature generation submodule is proposed, including: The second indicator quantification submodule is used to determine the basic attribute indicator set of the basic attribute indicator assessment network, the business value indicator set of the business value indicator assessment network, and the correlation strength indicator set of the strong correlation indicator assessment network based on the multi-dimensional assessment network under the current assessment scenario label. The Single Pathway Differentiation Analysis Submodule is used to perform single-pathway differentiation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation path, the business value indicator set of the business value indicator evaluation path, and the correlation strength indicator set of the strong correlation indicator evaluation path. It obtains the short-term fluctuation differentiation value, long-term trend differentiation value, and threshold breakthrough differentiation value of each data asset at each time step. The Interactive Alienation Analysis submodule is used to perform cross-network interactive alienation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation network, the business value indicator set of the business value indicator evaluation network, and the correlation strength indicator set of the strong correlation indicator evaluation network, to obtain the network collaborative alienation value and the dominant network influence alienation value between each pair of networks. The alienation feature determination submodule is used to generate a multi-dimensional alienation feature set of all data assets under the current evaluation scenario label based on the short-term fluctuation alienation value, long-term trend alienation value, threshold breakthrough alienation value, and alienation value of the synergistic alienation between pairs of threads and the alienation value of the dominant thread at each time step.

[0064] In this embodiment, based on the multi-dimensional evaluation context under the current evaluation scenario label, the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context are determined. This involves extracting corresponding sets of data indicators from the multi-dimensional evaluation context for different evaluation dimensions. For example, when evaluating internet user data assets, the basic attribute indicator set may include indicators such as data accuracy and completeness; the business value indicator set may include indicators such as user activity and paid conversion rate; and the correlation strength indicator set may involve indicators of the degree of correlation with user behavior data and consumption data.

[0065] In this embodiment, single-pathway alienation feature analysis is performed based on the basic attribute indicator set of the basic attribute indicator evaluation path, the business value indicator set of the business value indicator evaluation path, and the correlation strength indicator set of the strong correlation indicator evaluation path. This yields the short-term fluctuation alienation value, long-term trend alienation value, and threshold breakthrough alienation value for each data asset at each time step. Specifically, for each path's indicator set, the changing characteristics of the data asset at different time steps are analyzed. For example, when analyzing user activity (business value indicator set), the short-term fluctuation alienation value reflects the difference between user activity in recent days and the average activity over the past week; the long-term trend alienation value reflects the deviation of the user activity trend over the past few months from the expected trend; and the threshold breakthrough alienation value is used to determine whether user activity has exceeded a set key threshold, such as the daily active user count reaching a specific value. Specifically, this includes: The calculation process for short-term volatility distortion is as follows: first, calculate the difference between the characteristic value of the current time step and the average of the characteristic values ​​of the previous 3 consecutive time steps; Divide this difference by the larger of the "current feature value" and the "average of the previous 3 time steps" to obtain the relative deviation ratio; Finally, the sign of the result is determined based on the sign of the above difference (a positive value indicates that the current feature value is higher than the average of the last 3 steps, i.e., higher than the normal value; a negative value indicates that the current feature value is lower than the average of the last 3 steps, i.e., lower than the normal value).

[0066] The calculation process for the long-term trend deviation value is as follows: For each data asset under each context, determine whether the current characteristic value deviates from the evolutionary pattern based on the long-term trend. First, select the feature values ​​of the last 10 time steps, fit a trend line using a quadratic function, and then use the trend line to predict the feature value of the current time step. Calculate the difference between the current actual feature value and the above predicted value, and divide the absolute value of the difference by the predicted value to obtain the relative deviation. The sign of the result is determined by the sign of the difference (a positive value indicates that the current feature value is higher than the predicted trend, and a negative value indicates that it is lower than the predicted trend). If the goodness of fit (R²) of the trend line is less than 0.8, the time step needs to be reselected and the trend line refitted to ensure the reliability of the trend prediction.

[0067] The calculation process for the threshold exceeding the deviation value is as follows: Based on historical data, determine the normal fluctuation range, judge whether the current feature value exceeds the range, and quantify the degree of deviation. First, based on the historical characteristic values ​​of the data assets under this context, calculate the 95% confidence interval (including the lower and upper limits, reflecting the boundaries of normal fluctuations). If the current eigenvalue is higher than the upper limit of the confidence interval, divide "current value - upper limit" by "upper limit - lower limit" to obtain the relative magnitude of the upward breakthrough. If the current feature value is lower than the lower limit of the confidence interval, divide "lower limit - current value" by "upper limit - lower limit" to obtain the relative magnitude of the downward breakthrough. If the current feature value is within the confidence interval, the result is 0 (indicating that it has not exceeded the normal range).

[0068] In this embodiment, cross-context interaction and alienation feature analysis is performed based on the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. This yields the context synergy alienation value and the dominant context influence alienation value between each pair of contexts. In other words, it studies the mutual influence relationship between indicator sets from different contexts. For example, analyzing the relationship between basic attribute indicators (such as data accuracy) and business value indicators (such as user conversion rate), the context synergy alienation value can represent the degree to which improved data accuracy synergistically promotes user conversion rate; the dominant context influence alienation value can determine which context plays a dominant role and the degree of influence in the interaction between these two contexts, judging whether improved data accuracy dominates the change in user conversion rate, or vice versa. Specifically, this includes: The calculation process for the network synergistic anisotropy value is as follows: quantifying whether the synergistic relationship between two different networks deviates from historical patterns. For any two threads, calculate the Pearson correlation coefficient (reflecting the current coordination strength) of the rate of change of their characteristics at the current time step. Calculate the difference between this correlation coefficient and the historical average correlation coefficient over the past 30 days (which reflects historical synergistic patterns); After taking the absolute value of the difference, the sign of the result is determined based on the relationship between the current correlation coefficient and the historical mean (a positive correlation coefficient indicates enhanced synergy, while a negative correlation coefficient indicates weakened synergy), thereby reflecting whether the association between the two threads is abnormal.

[0069] The calculation process for the influence of the dominant network on the alienation value is as follows: Assess whether the influence of non-dominant networks on the dominant network meets expectations. First, determine the "dominant thread" based on the evaluation scenario labels (e.g., in a transaction scenario, the business value thread is the core dominant thread). Calculate the feature change of the dominant context at the current time step, and the sum of the product of the feature change of all non-dominant contexts and their respective preset weights (based on scene label settings, reflecting the expected influence of non-dominant contexts on the dominant context). Subtracting the sum of the above products from the characteristic changes of the dominant network yields the deviation between the actual and expected impacts. Divide this deviation by the larger of "change in dominant network characteristics" and "1" (to avoid distorting the result due to an excessively small denominator) to reflect whether the influence of non-dominant networks on dominant networks deviates from expectations.

[0070] In this embodiment, based on the short-term fluctuation alienation value, long-term trend alienation value, threshold breakthrough alienation value, and the alienation value of the synergistic alienation between pairs of threads and the alienation value of the dominant thread's influence for all data assets at each time step, a multi-threaded temporal multi-dimensional alienation feature set for all data assets under the current evaluation scenario label is generated. The alienation values ​​obtained from single-thread and cross-thread analyses of each data asset at different time steps are integrated to form a comprehensive set of alienation features that reflect the multi-dimensional, time-varying alienation of data assets.

[0071] To determine the spatiotemporal evolution variation factor of each data asset by assigning weights, calculating the comprehensive variation value, and performing standardization and trend fitting, a variation factor determination submodule is further proposed, including: The weight allocation unit is used to assign weights to each alienation feature in the multi-context temporal multidimensional alienation feature set based on the current evaluation scenario label. The alienation synthesis unit is used to calculate the synthesis alienation value of each data asset at each time step based on the weight of each alienation feature in the multi-context temporal multi-dimensional alienation feature set. The heterogeneity unit is used to standardize and trend-fit the comprehensive heterogeneity value of each data asset at each time step to obtain the spatiotemporal evolution heterogeneity factor of each data asset.

[0072] In this embodiment, assigning weights to each alienation feature in the multi-dimensional alienation feature set based on the current evaluation scenario label means setting an importance ratio for different types of alienation values ​​in the multi-dimensional alienation feature set according to the specific evaluation scenario. For example, based on the scenario of "strategy evaluation for optimizing delivery efficiency", the platform believes that long-term trends are the most critical to strategy evaluation, so it assigns a weight of 0.4 to the long-term trend alienation value; short-term fluctuations have the second greatest impact, with a weight of 0.3; threshold breakthroughs, although occasional, have a significant impact, with a weight of 0.2; and the impact of contextual coordination and dominant contexts is relatively weak, with weights of 0.05 and 0.05 respectively.

[0073] In this embodiment, the comprehensive alienation value of each data asset at each time step is calculated based on the weight of each alienation feature in the multi-dimensional alienation feature set of multi-context time series. That is, the various alienation values ​​of each data asset at each time step are weighted and summed using the assigned weights. Taking the order delivery time-related data asset mentioned above as an example, assuming the short-term fluctuation alienation value is 0.3, the long-term trend alienation value is 0.2, the threshold breakthrough alienation value is 0.1, the context coordination alienation value is 0.25, and the dominant context influence alienation value is 0.15; its comprehensive alienation value at this time step = short-term fluctuation alienation value × corresponding weight + long-term trend alienation value × corresponding weight + threshold breakthrough alienation value × corresponding weight + context coordination alienation value × corresponding weight + dominant context influence alienation value × corresponding weight, which equals 0.21. If a single data asset has two or more context-coordinated alienation values ​​or dominant context-influence alienation values ​​(e.g., when there are three or more contexts), the average of these multiple context-coordinated alienation values ​​or dominant context-influence alienation values ​​can be calculated before participating in the weighted summation operation. This yields the comprehensive alienation value of the data asset at this time step, which is used to measure the overall degree of alienation of the data asset under the current evaluation scenario.

[0074] To standardize the comprehensive heterogeneity value of each data asset at each time step, and to determine the slope as a spatiotemporal evolution heterogeneity factor through linear fitting, a heterogeneity unit is further proposed, including: The standardized sub-unit is used to standardize the comprehensive alienation value of each data asset at each time step, so as to obtain the standard comprehensive alienation value of each data asset at each time step. The linear fitting subunit is used to perform linear fitting on the standard comprehensive alienation value of each data asset at a preset number of time steps and determine the slope as the spatiotemporal evolution alienation factor of each data asset.

[0075] In this embodiment, the comprehensive alienation value of each data asset at each time step is standardized to obtain a standard comprehensive alienation value for each data asset at each time step. This step aims to eliminate the impact of different units or value ranges in the comprehensive alienation values ​​of different data assets, making them comparable. For example, the comprehensive alienation values ​​of different data assets may be between 0 and 10, or between 0 and 100. Through standardization methods (such as normalization), they are all mapped to the same value range (such as 0-1) to obtain a standard comprehensive alienation value. Assume there are two data assets, A and B, in an e-commerce data evaluation scenario.

[0076] Data asset A has a calculated comprehensive alienation value of 8 at a certain time step, while data asset B has a comprehensive alienation value of 60 at the same time step. Because the various indicators used to calculate the comprehensive alienation value differ, their value ranges vary significantly: A's comprehensive alienation value is between 0 and 10, while B's is between 0 and 100, making a direct comparison meaningless.

[0077] Using the normalization method, the formula is: Standardized Comprehensive Differentiation Value = (Original Value - Minimum Value) / (Maximum Value - Minimum Value).

[0078] For data asset A, assuming its value range is 0-10, substituting into the formula, its standard comprehensive alienation value = (8-0) / (10-0) = 0.8.

[0079] For data asset B, the value range is 0-100, and its standard comprehensive alienation value = (60-0) / (100-0) = 0.6.

[0080] In this way, after standardization, the standard comprehensive alienation values ​​of A and B at this time step are both within the range of 0-1, eliminating the impact caused by the different value ranges. The degree of alienation of the two at this time step can be directly compared. 0.8 is greater than 0.6, indicating that the degree of alienation of data asset A at this time step is relatively higher than that of B.

[0081] In this embodiment, a linear fit is performed on the standard comprehensive alienation value of each data asset over a preset number of time steps, and the slope is determined as the spatiotemporal evolutionary variation factor for each data asset. By performing a linear fit on the standard comprehensive alienation value over a period of time (preset number of time steps), a straight line that best represents its changing trend is obtained, and the slope of the straight line is the spatiotemporal evolutionary variation factor. For example, if the change in the standard comprehensive alienation value of a data asset over the past 10 time steps (preset number of time steps) is observed, and a straight line is fitted with a slope of 0.05, it indicates that the data asset exhibits a certain changing trend in the spatiotemporal dimension, and this slope serves as the spatiotemporal evolutionary variation factor to measure its dynamic changes.

[0082] like Figure 6As shown, in order to achieve the fusion weighting of the defined spatiotemporal evolution variation factor and the static value assessment result, and based on this fusion weighting and the data asset assessment results generated by both for all data assets under the current assessment scenario label, a data asset comprehensive assessment module is further proposed, including: The Fusion Weight Definition submodule is used to define the fusion weights of spatiotemporal evolution variation factors and static value assessment results; The asset fusion assessment submodule is used to generate data asset assessment results for all data assets under the current assessment scenario label based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, as well as the fusion weights of the spatiotemporal evolution variation factors and static value assessment results.

[0083] In this embodiment, defining the fusion weight of spatiotemporal evolution variation factors and static value assessment results means determining the relative importance of spatiotemporal evolution variation factors and static value assessment results in the final data asset assessment based on the assessment purpose and scenario requirements. For example, when assessing data assets with high growth potential, spatiotemporal evolution variation factors may be given a higher weight; while for relatively stable data assets with clear current value, the static value assessment results will have a higher weight.

[0084] In this embodiment, the data asset evaluation result for all data assets under the current evaluation scenario label is generated based on the spatiotemporal evolution variation factor and static value assessment result of all data assets, as well as the fusion weight of the spatiotemporal evolution variation factor and static value assessment result. That is, the spatiotemporal evolution variation factor and static value assessment result are comprehensively calculated using the determined fusion weight. Assuming that the static value assessment result of a certain data asset is 80, the spatiotemporal evolution variation factor is 0.6, and the fusion weights are 0.4 and 0.6 respectively, then the evaluation result of this data asset = 80 × 0.4 + 80 × 0.6 × 0.6 = 32 + 28.8 = 60.8, thus comprehensively reflecting the value of the data asset in the current scenario.

[0085] This invention provides an implementation method for an intelligent data asset assessment method based on multi-dimensional data analysis, comprising: Construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps; Identify the current assessment scenario label, and based on the assessment strategy of the current assessment scenario label, capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets; Based on the multi-dimensional assessment framework under the current assessment scenario label, determine the static value assessment results of all data assets under the current assessment scenario label; Generate a multi-contextual, temporal, and multi-dimensional alienation feature set for all data assets under the current evaluation scenario label, and determine the spatiotemporal evolution and alienation factor of each data asset based on the multi-contextual, temporal, and multi-dimensional alienation feature set. Based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, the data asset assessment results of all data assets under the current assessment scenario label are generated.

[0086] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A data asset intelligent assessment system based on multi-dimensional data analysis, characterized in that: include: The association network construction module is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps. The assessment context capture module is used to identify the current assessment scenario label and capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets based on the assessment strategy of the current assessment scenario label. The asset static valuation module is used to determine the static value valuation results of all data assets under the current valuation scenario label based on the multi-dimensional valuation context under the current valuation scenario label. The mutation factor determination module is used to generate a multi-contextual temporal multi-dimensional mutation feature set for all data assets under the current evaluation scenario label, and to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multi-dimensional mutation feature set. The data asset comprehensive evaluation module is used to generate data asset evaluation results for all data assets under the current evaluation scenario label, based on the spatiotemporal evolution and variation factors and static value evaluation results of all data assets.

2. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 1, characterized in that, The network construction module includes: The multidimensional feature acquisition submodule is used to collect the data scale, update frequency, and access popularity of all data assets in real time at multiple time steps, and to measure the associated data volume and business impact of all data assets. The data scale, update frequency, access popularity, associated data volume, and business impact of all data assets at each time step are used as the multidimensional features of all data assets at each time step. The edge weight determination submodule is used to calculate the edge weight of the corresponding two data assets at each time step based on the information gain rate and co-occurrence frequency between each pair of data assets at each time step. The association network construction submodule is used to construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps and the edge weights of every two data assets at multiple time steps.

3. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 1, characterized in that, The evaluation context capture module includes: The scene tag recognition submodule is used to match the core elements in the user evaluation request with the feature word library of multiple preset scene tags and determine all matching scene tags and their corresponding matching degrees. Based on the matching degree, it is determined whether there are obviously matching scene tags. If so, the obviously matching scene tag is used as the pre-evaluation scene tag. Otherwise, the semantic similarity between the user evaluation request and the feature word library of each preset scene tag is calculated, and the preset scene tag with the maximum semantic similarity is used as the pre-evaluation scene tag. The module involves a node capture submodule, which is used to identify all core nodes in the multidimensional dynamic correlation network of data assets based on the evaluation strategy of the current evaluation scenario label, and to determine the starting point of the network among all core nodes based on the importance of all core nodes. The node multi-hop expansion submodule is used to perform multi-hop expansion along the association edges in the multi-dimensional dynamic association network of data assets based on the network starting point and expansion rules, so as to obtain all valid node paths. The effective path reorganization submodule is used to reorganize all effective node paths according to the evaluation dimensions of the evaluation strategy of the current evaluation scenario label, so as to obtain a weighted multidimensional evaluation context under the current evaluation scenario label.

4. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 1, characterized in that, The asset static valuation module includes: The multidimensional context decomposition submodule is used to determine the basic attribute indicator evaluation context, business value indicator evaluation context, and strong correlation indicator evaluation context under the current evaluation scenario label based on the multidimensional evaluation context under the current evaluation scenario label, and to determine the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The first indicator quantification submodule is used to calculate the basic attribute evaluation indicators, business value indicators, and correlation strength indicators of all data assets under the current evaluation scenario label, based on the basic attribute indicator set of the basic attribute indicator evaluation context, the business value indicator set of the business value indicator evaluation context, and the correlation strength indicator set of the strong correlation indicator evaluation context. The static value assessment submodule is used to determine the comprehensive weight of each indicator type based on the multi-dimensional assessment context under the current assessment scenario label. Based on the comprehensive weight of all indicator types, it assesses the basic attribute indicators, business value indicators, and correlation strength indicators of all data assets under the current assessment scenario label, and obtains the static value assessment results of all data assets under the current assessment scenario label.

5. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 1, characterized in that, The variable factor determination module includes: The alienation feature generation submodule is used to perform alienation feature analysis based on the multidimensional assessment context under the current assessment scenario label and generate a multi-context temporal multidimensional alienation feature set for all data assets under the current assessment scenario label. The mutation factor determination submodule is used to determine the spatiotemporal evolution mutation factor of each data asset based on the multi-contextual temporal multidimensional mutation feature set.

6. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 5, characterized in that, The alienation feature generation submodule includes: The second indicator quantification submodule is used to determine the basic attribute indicator set of the basic attribute indicator assessment network, the business value indicator set of the business value indicator assessment network, and the correlation strength indicator set of the strong correlation indicator assessment network based on the multi-dimensional assessment network under the current assessment scenario label. The Single Pathway Differentiation Analysis Submodule is used to perform single-pathway differentiation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation path, the business value indicator set of the business value indicator evaluation path, and the correlation strength indicator set of the strong correlation indicator evaluation path. It obtains the short-term fluctuation differentiation value, long-term trend differentiation value, and threshold breakthrough differentiation value of each data asset at each time step. The Interactive Alienation Analysis submodule is used to perform cross-network interactive alienation feature analysis based on the basic attribute indicator set of the basic attribute indicator evaluation network, the business value indicator set of the business value indicator evaluation network, and the correlation strength indicator set of the strong correlation indicator evaluation network, to obtain the network collaborative alienation value and the dominant network influence alienation value between each pair of networks. The alienation feature determination submodule is used to generate a multi-dimensional alienation feature set of all data assets under the current evaluation scenario label based on the short-term fluctuation alienation value, long-term trend alienation value, threshold breakthrough alienation value, and alienation value of the synergistic alienation between pairs of threads and the alienation value of the dominant thread at each time step.

7. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 5, characterized in that, The variant factor determination submodule includes: The weight allocation unit is used to assign weights to each alienation feature in the multi-context temporal multidimensional alienation feature set based on the current evaluation scenario label. The alienation synthesis unit is used to calculate the synthesis alienation value of each data asset at each time step based on the weight of each alienation feature in the multi-context temporal multi-dimensional alienation feature set. The heterogeneity unit is used to standardize and trend-fit the comprehensive heterogeneity value of each data asset at each time step to obtain the spatiotemporal evolution heterogeneity factor of each data asset.

8. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 7, characterized in that, Heterogeneous units, including: The standardized sub-unit is used to standardize the comprehensive alienation value of each data asset at each time step, so as to obtain the standard comprehensive alienation value of each data asset at each time step. The linear fitting subunit is used to perform linear fitting on the standard comprehensive alienation value of each data asset at a preset number of time steps and determine the slope as the spatiotemporal evolution alienation factor of each data asset.

9. The intelligent data asset evaluation system based on multi-dimensional data analysis according to claim 1, characterized in that, The data asset comprehensive assessment module includes: The Fusion Weight Definition submodule is used to define the fusion weights of spatiotemporal evolution variation factors and static value assessment results; The asset fusion assessment submodule is used to generate data asset assessment results for all data assets under the current assessment scenario label based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, as well as the fusion weights of the spatiotemporal evolution variation factors and static value assessment results.

10. A data asset intelligent evaluation method based on multi-dimensional data analysis, characterized in that: include: Construct a multidimensional dynamic association network of data assets based on the multidimensional features of all data assets at multiple time steps; Identify the current assessment scenario label, and based on the assessment strategy of the current assessment scenario label, capture the multidimensional assessment context under the current assessment scenario label in the multidimensional dynamic association network of data assets; Based on the multi-dimensional assessment framework under the current assessment scenario label, determine the static value assessment results of all data assets under the current assessment scenario label; Generate a multi-contextual, temporal, and multi-dimensional alienation feature set for all data assets under the current evaluation scenario label, and determine the spatiotemporal evolution and alienation factor of each data asset based on the multi-contextual, temporal, and multi-dimensional alienation feature set. Based on the spatiotemporal evolution variation factors and static value assessment results of all data assets, the data asset assessment results of all data assets under the current assessment scenario label are generated.

Images