A digital asset management method based on blood relationship topology calculation

Abstract

This invention relates to a digital asset management method based on kinship topology computation, comprising: extracting semantic features of data Dx to generate a high-dimensional fingerprint; retrieving the most similar source data node in a graph database based on the high-dimensional fingerprint; determining the data Dx to obtain a determination result; issuing an authorization certificate on the blockchain based on the determination result; and creating the node and evolutionary edge corresponding to the data Dx in the graph database; recursively querying the kinship path in the graph database and extracting contribution weights based on the revenue P of the data Dx; uploading the contribution weights to the blockchain's ledger contract; calculating the profit share of each level of participants; transferring funds according to the profit share; and providing feedback on the settlement result. This invention solves the technical problems of unfair distribution, control failure, and difficulty in value estimation in the flow of data elements.

Description

Technical Field

[0001] This invention relates to the field of data asset technology, and specifically to a digital asset management method based on lineage topology calculation. Background Technology

[0002] With the booming development of the digital economy, data has become the fifth major factor of production. In December 2022, the "20 Measures for Data" clearly proposed the reform framework of "separation of three rights for data," namely, establishing the right to hold data resources, the right to process and use data, and the right to operate data products. However, in the actual implementation process, the data factor market not only faces the initial challenges of "difficulty in confirming rights and difficulty in tracing the source," but has also entered the deep waters of "difficulty in distributing benefits, difficulty in cross-domain management, and difficulty in value assessment." Currently, existing technologies have the following significant limitations in complex cross-entity, multi-level data asset transfer scenarios: 1. Blockchain-based static evidence storage technology: Current technologies commonly utilize blockchain for data hashing and notarization. While this solves the problem of "immutable ownership information," it has limitations when dealing with dynamic asset management: "Data plagiarism" and the lack of originality assessment: Simple hash comparison can only identify completely identical data. Without the linkage between similarity algorithms and lineage maps, infringers can disguise data as "new original data" by slightly adjusting the data format or adding or deleting bytes, resulting in the illegal dilution of the original ownership.

[0003] The "discretization" of access control: Existing authorization is mostly a static, point-to-point action. When the authorization of source data expires or is revoked for compliance reasons, the blockchain ledger struggles to automatically cascade and block hundreds or thousands of downstream derivative data products, resulting in extremely weak extended control over data assets and significant compliance risks.

[0004] 2. Traditional data lineage management and analysis systems: While traditional metadata systems can record the flow of data, their limitation lies in "only recording, not executing": Lack of physical enforcement mechanisms: Existing lineage analysis technologies are mostly used for offline "risk identification" or "impact analysis." In real-time asset management scenarios, these systems cannot automatically trigger fund return settlement based on lineage paths, nor can they perform millisecond-level physical shutdown of non-compliant assets in the transfer chain when anomalies are detected.

[0005] The disconnect between lineage and value calculation: Current lineage management treats data merely as a topological record, failing to extract structural features of the graph (such as node reference depth, frequency of dependency, and similarity decay coefficient) to transform them into pricing factors for production factors. This results in data asset transactions still relying on manual negotiation, lacking an objective, algorithm-driven dynamic pricing mechanism.

[0006] 3. Fragmentation of the data element revenue distribution mechanism: In scenarios involving the fusion and processing of multi-source data, a final data product often incorporates contributions from multiple original data sources. Existing profit distribution (revenue sharing) largely relies on pre-set offline contracts with fixed percentages. Due to the lack of quantifiable means of measuring the actual technical contributions during data flow (such as the proportion of features retained), it is difficult to achieve fair, dynamic, and automated profit sharing, severely hindering the enthusiasm of data holders to participate in the flow of resources.

[0007] 4. The complexity of ownership penetration and cascading control: As the data industry chain extends, the authorization chain is highly susceptible to breakage. Achieving robust control—allowing for "one-click authorization revocation at the source and immediate circuit breaking across the entire chain"—within complex, network-like derivative relationships is a major technical bottleneck in today's heterogeneous environments.

[0008] In summary, the current data asset management field urgently needs a new technological architecture that bridges the gap between "static recording" and "dynamic governance." This patent application proposes a "chain-graph collaborative" digital asset management system based on data lineage graphs and the separation of powers. This system not only ensures originality through "similarity monitoring," but also leverages graph topological features to drive "dynamic pricing," utilizes a lineage contribution algorithm to execute "intelligent revenue sharing," and achieves "cascading circuit breaking and extended control" of permissions through graph path traversal. This invention aims to address industry pain points such as "unfair distribution, control failure, and difficulty in value estimation" in the flow of data elements. Summary of the Invention

[0009] To address the aforementioned problems in the existing technology, this invention provides a digital asset management method based on lineage topology calculation. The technical problem to be solved by this invention is achieved through the following technical solution: This invention provides a digital asset management method based on lineage topology calculation, comprising: S1. In response to the registration request, extract the semantic features of the data Dx uploaded by the original holder A to generate a high-dimensional fingerprint. Based on the high-dimensional fingerprint, search for the most similar source data node in the graph database. Based on the similarity between the data Dx and the source data node, determine the data Dx to obtain a determination result. Based on the determination result, issue an authorization certificate on the blockchain and create the node and evolution edge corresponding to the data Dx in the graph database, where x < y. S2. Based on the revenue P of the data Dx, the rights confirmation and control center recursively queries the bloodline path to the graph database and extracts the contribution weight. The contribution weight is then uploaded to the blockchain's ledger contract. The ledger contract calculates the profit share of each level of participants, transfers funds according to the profit share, and reports the settlement result back to the rights confirmation and control center.

[0010] In one embodiment of the present invention, extracting semantic features of the data Dx uploaded by the original holder A to generate a high-dimensional fingerprint includes: The semantic features of the data Dx uploaded by the original holder A are extracted and mapped to the same vector space, and then mapped to the same hash code space to obtain a high-dimensional fingerprint with semantic awareness characteristics. In one embodiment of the present invention, the semantic features of the data Dx uploaded by the original holder A are extracted and mapped to the same vector space, and then mapped to the same hash code space to obtain a high-dimensional fingerprint with semantic awareness characteristics, including: The data Dx uploaded by the original holder A is input into the CLIP model. The CLIP model is used to extract semantic features from the data Dx. The extracted semantic features are mapped to the same vector space to obtain a semantic feature vector. Then, the semantic feature vector is mapped to the same hash code space through an MLP network to obtain a high-dimensional fingerprint of the data Dx. In one embodiment of the present invention, the most similar source data node is retrieved from the graph database based on the high-dimensional fingerprint, and a determination result is obtained by judging the data Dx based on the similarity between the data Dx and the source data node, including: Calculate the similarity between the high-dimensional fingerprint of the data Dx and the high-dimensional fingerprint of each source data node in the graph database. Select the source data node with the highest similarity as the most similar source data node. Determine whether the data Dx is original data or whether the data Dx has the L2 level permission processing authorization of the most similar source data node when the highest similarity is within the preset similarity interval [x%, y%]. If so, use the feature retention degree of the data Dx to the most similar source data node as the initial contribution weight of the data Dx. Here, L2 level permission means that the system allows the generation of tradable operating rights certificates for data assets marked with this level, and supports their subsequent evolution and automatic accounting in the graph database. In one embodiment of the present invention, the data Dx is determined based on the relationship between the maximum similarity and a preset similarity interval [x%, y%]. The determination includes whether the data Dx is original data or whether the data Dx has L2-level permission authorization from the most similar source data node when the maximum similarity is within the preset similarity interval [x%, y%]. If so, the feature retention degree of the data Dx on the most similar source data node is used as the initial contribution weight of the data Dx, including: The system determines the relationship between the maximum similarity and the preset similarity interval [x%, y%]. If the maximum similarity is less than x%, the data Dx is determined to be original data, and the data Dx is set to have either an L1-level permission or an L2-level permission. If the maximum similarity is greater than y%, the data Dx is not uploaded. If the maximum similarity is greater than or equal to x% and less than or equal to y%, the system continues to determine whether the data Dx has L2-level permission authorization from the most similar source data node. If not, the data Dx is not uploaded. If yes, the system calculates the feature retention rate of the data Dx to the most similar source data node and uses the feature retention rate as the initial contribution weight of the data Dx. Here, L1-level permission means that the system prohibits the generation of independent tradable certificates for data assets marked with this level and locks their evolution path to unrestricted nodes in the graph database. In one embodiment of the present invention, based on the determination result, an authorization certificate is issued on the blockchain, and nodes and evolutionary edges corresponding to the data Dx are created in the graph database, including: For data Dx that is original data or has L2-level permission permission status, a permission certificate is issued on the blockchain, and a node corresponding to the data Dx is created in the graph database. The initial contribution weight is written as an attribute into the evolution edge between the node of the data Dx and the most similar source data node, and the data Dx is marked as active in the graph database. In one embodiment of the present invention, based on the revenue P of the data Dx, the rights confirmation and control center recursively queries the graph database for kinship paths and extracts contribution weights, including: When product Dc, formed from the data Dx, completes a transaction and generates revenue P, operator C submits a payment settlement request to the rights confirmation and control center. The rights confirmation and control center then initiates a depth-first traversal request to the graph database to retrieve the complete lineage path from the data Dx to the product Dc, and to extract the contribution weights of records on all evolutionary edges in the lineage path. In one embodiment of the present invention, the contribution weight is uploaded to the blockchain's ledger contract, which calculates the profit share of each participant, transfers funds according to the profit share, and feeds back the settlement results to the rights confirmation and control center, including: The rights confirmation and control center will transfer the extracted contribution weight to the ledger contract deployed on the blockchain. The ledger contract will calculate the profit share of each level of participants in the bloodline path according to a preset algorithm. The revenue sharing contract executes transfer actions on the blockchain, transferring funds to the accounts of participants at each level according to the calculated profit share, and reporting the settlement results back to the rights confirmation and control center. In one embodiment of the present invention, after step S2, the method further includes: S3. In response to the change operation of revoking the authorization of data asset Da, locate all downstream data nodes derived from the data asset Da in the lineage graph, and issue control instructions to the blockchain to modify the contract status corresponding to the downstream data nodes derived from the data asset Da to the restricted state, thereby realizing physical-level circuit breaker control. In one embodiment of the present invention, in response to a change operation involving the revocation of authorization of a data asset Da, all downstream data nodes derived from the data asset Da are located in the lineage graph, including: The original holder A initiates a change operation on the blockchain to revoke the authorization of the data asset Da. The chain graph collaboration controller captures the change operation and calls the graph database to search in order to locate all downstream data nodes derived from the data asset Da in the lineage graph. Compared with the prior art, the beneficial effects of the present invention are as follows: This invention achieves fair revenue distribution based on lineage contribution, solving the pain point of "difficult revenue sharing." Instead of relying on subjective offline contract negotiations, this invention quantifies the substantive contributions of each level of original holder and processor to the final data product through the objective technical indicator of "feature retention" fixed in the graph database. The invention ensures transparent distribution: by using lineage graphs for reverse tracing, it automatically calculates contribution weights and drives smart contracts to execute millisecond-level automatic revenue sharing. This invention incentivizes originality: this mechanism ensures that even after multiple levels of processing, the holders of the original data still receive derivative revenue commensurate with their technological contributions, significantly enhancing the liquidity of the data element market.

[0011] This invention achieves end-to-end "cascading circuit breaker" extended control, solving the pain point of "difficult compliance management." By establishing a strong synchronization mechanism between "blockchain certificate status" and "graph topology path," this invention empowers the source rights holder with physical-level control over downstream derivative assets. This invention possesses immediacy and scalability: when the source data authorization expires or is revoked, the system uses a breadth-first search algorithm to locate all downstream dependent nodes within milliseconds and automatically locks the associated transaction smart contracts. This invention achieves physical-level interception: this "one-click revocation, network-wide circuit breaker" mechanism completely solves the industry problem of "source out of control and downstream proliferation" in data element flow, ensuring the end-to-end transparent transmission of data compliance requirements.

[0012] This invention establishes an algorithm-driven dynamic asset pricing model, addressing the pain point of "difficulty in value assessment." It innovatively utilizes graph databases to extract "topological characteristic factors" of nodes (such as out-degree / citation frequency, lineage depth, and scarcity coefficient) to construct an objective, automated valuation system. The pricing of this invention is objective: asset prices dynamically fluctuate with market reuse rate and lineage scarcity, eliminating subjective errors from manual assessment. This invention reduces transaction costs: the dynamic pricing engine enables real-time asset quotations, significantly shortening the matching cycle for data asset transactions and improving the resource allocation efficiency of the factor market.

[0013] This invention achieves a high-performance architecture of "computational collaboration" and "state synchronization," optimizing operating costs. It fully leverages the respective advantages of graph databases (emphasizing computation and path) and blockchain (emphasizing execution and trust), achieving a significant leap in system performance. This invention improves retrieval and computation efficiency: when processing tens of millions of complex lineage relationships, parameter extraction and path traversal based on graph features remain stable at the hundreds of milliseconds level, with retrieval efficiency exceeding that of traditional relational databases by more than 70%. This invention provides balanced storage and reliable execution: by storing complex evolutionary weights and topological parameters in a graph database, and recording only core clearing credentials and control instructions on the blockchain, it effectively avoids the "state explosion" of the blockchain, significantly reducing maintenance costs in heterogeneous environments while ensuring data immutability.

[0014] This invention accurately identifies hidden infringements, safeguarding the originality of data resource ownership. Utilizing cross-modal feature extraction and a dual-threshold dynamic judgment matrix, this invention automates the interception of minor infringing behaviors such as "data plagiarism" during the registration process. Compared to traditional hash comparison, it improves the identification rate by approximately 45% for similar data processed through format conversion, addition, deletion, and obfuscation, ensuring the "cleanliness" of the rights confirmation system from the source.

[0015] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a digital asset management method based on lineage topology calculation provided by the present invention; Figure 2 This invention provides a logical architecture diagram of an asset management system based on graph topology features. Figure 3 This is a schematic diagram of the process for determining the weight and initial contribution of data Dx provided by the present invention; Figure 4 This is a schematic diagram of an automatic revenue clearing and intelligent profit distribution process in a multi-level processing scenario provided by the present invention; Figure 5 This is a schematic diagram of a cascading circuit breaker and extended control process triggered by the revocation of source authorization, provided by the present invention; Figure 6 This is a flowchart of a dynamic pricing and recursive revenue sharing algorithm based on bloodline contribution provided by the present invention; Figure 7 This invention provides a schematic diagram of a cascaded fuse and extension control principle based on topology state synchronization. Figure 8 This invention provides a data asset ownership and management status evolution diagram throughout its entire lifecycle. Detailed Implementation

[0017] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0018] Example 1 Existing technological solutions for data ownership confirmation generally focus on the preservation and recording of static ownership, but they lack the technological capacity to address issues such as the quantification of benefits during data processing and evolution, end-to-end cascading circuit breaker control, and automated pricing based on topology structures. In large-scale, multi-entity data assetization scenarios, existing solutions struggle to achieve dynamic value assessment and automated security governance within the framework of "separation of three rights," severely hindering the release of data's value as a production factor.

[0019] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a digital asset management method based on lineage topology calculation provided by the present invention. The present invention provides a digital asset management method based on lineage topology calculation. This embodiment of the digital asset management method runs on a server equipped with an NVIDIA® Tesla V100 GPU, using a PyTorch 2.0 and Neo4j 5.x environment. The digital asset management method based on lineage topology calculation includes: S1. In response to the registration request, extract the semantic features of the data Dx uploaded by the original holder A to generate a high-dimensional fingerprint. Based on the high-dimensional fingerprint, search for the most similar source data node in the graph database. Based on the similarity between data Dx and the source data node, determine the data Dx to obtain the determination result. Based on the determination result, create the corresponding source data node and evolution edge in the graph database, and generate the permission certificate corresponding to the data asset node on the blockchain, x < y.

[0020] In this embodiment, Figure 2This demonstrates the deep coupling relationships and data flow among the various functional components within the system. The system architecture uses the rights confirmation and control center as its hub, transforming the underlying lineage graph features into executable contract logic. At the infrastructure layer, the graph database stores, manages, and computes the relationships between data. In the graph database, data is stored in a graph structure, including high-dimensional fingerprints of data nodes and edges. Each data node represents an independent data asset, and edges represent relationships such as derivation, dependency, and processing between nodes. The graph database not only stores static lineage nodes but also dynamically extracts topological feature factors (topological feature factors refer to structured feature parameters extracted from the data lineage topology graph (such as node centrality, path length, etc.) through its built-in graph computing engine to support the system in executing complex business logic decisions through preset algorithms without manual intervention. These include the node's out-degree in the lineage network (reflecting the frequency of resource dependence) and path depth (reflecting processing complexity). These technical parameters are fed back in real-time to the dynamic pricing engine and intelligent revenue sharing module at the control layer via standard interfaces. Meanwhile, the blockchain layer not only handles certificate storage but also serves as the hard execution environment for asset liquidation, working closely with cascading circuit breakers. This system architecture achieves deep synchronization between the physical layer's data flow evolution and the logical layer's asset allocation and access control, ensuring that every asset action is driven by objective topological data, rather than subjective static rules.

[0021] Optionally, the graph database is Neo4j.

[0022] In a specific embodiment, such as Figure 3 As shown, step S1 may include: S11. In response to the registration request, extract the semantic features of the data Dx uploaded by the original holder A, map them to the same vector space, and then map them to the same hash code space to obtain a high-dimensional fingerprint with semantic awareness characteristics.

[0023] Specifically, the data Dx uploaded by the original holder A is input into the CLIP model (Contrastive Language-Image Pre-training, a pre-trained model based on contrastive text-image pairs). The CLIP model is used to extract semantic features from the data Dx, and the extracted semantic features are mapped to the same vector space to obtain semantic feature vectors. Then, the semantic feature vectors are mapped to the same hash code space through an MLP network to obtain the high-dimensional fingerprint of the data Dx, which is a hash value.

[0024] Furthermore, the data Dx can be text or an image.

[0025] S12. Calculate the similarity between the high-dimensional fingerprint of data Dx and the high-dimensional fingerprint of each source data node stored in the graph database. Select the source data node with the maximum similarity as the most similar source data node. Determine whether data Dx is original data or whether data Dx has the L2 level permission authorization of the most similar source data node when the maximum similarity is in the preset similarity interval [x%, y%]. If so, use the feature retention of data Dx on the most similar source data node as the initial contribution weight of data Dx. The feature retention is the similarity between data Dx and the most similar source data node.

[0026] Specifically, firstly, the similarity between the high-dimensional fingerprint of data Dx and the high-dimensional fingerprint of each source data node stored in the graph database is calculated. The maximum similarity is selected from all the calculated similarities, and the source data node corresponding to the maximum similarity is taken as the most similar source data node. Then, the relationship between the maximum similarity and the preset similarity interval [x%, y%] is judged. If the maximum similarity is less than x%, data Dx is determined to be original data, and data Dx is set to have L1 or L2 permission status. If the maximum similarity is greater than y%, it is considered plagiarism and data Dx is not uploaded. If the maximum similarity is greater than or equal to x% and less than or equal to y%, it is further judged whether data Dx has L2 permission processing authorization of the most similar source data node. If not, data Dx is not uploaded. If so, the feature retention degree of data Dx to the most similar source data node is calculated, and the feature retention degree is used as the initial contribution weight of data Dx.

[0027] In this embodiment, L1 level permissions prohibit the system from generating independent tradable certificates for data assets marked at this level and lock their evolution path to unrestricted nodes in the graph database. L1 level refers to the permission control level set for original digital assets or highly sensitive assets. In this system, assets determined to be L1 level are only granted "internal closed-loop processing" permissions. The system forcibly prohibits the generation of independent circulation certificates for these assets through underlying logic and locks their evolution path to unrestricted nodes in the lineage graph, ensuring that the data is only used in the controlled environment of the authorizing party and strictly prohibiting its circulation in the external market. L2 level permissions allow the system to generate tradable operating rights certificates for data assets marked at this level and support their subsequent evolution and automatic accounting in the graph database. L2 level refers to the operating permission level set for derivative digital assets that have been de-identified, aggregated, or have significant original contributions. In this system, assets determined to be L2 level are granted "value-added processing" and "derivative rights confirmation" permissions. The system supports the generation of tradable operating rights certificates for assets of this level on the blockchain, and allows subsequent path evolution and automated revenue sharing in the lineage graph, supporting the element-based market circulation of assets.

[0028] Furthermore, the similarity calculation method is to calculate the similarity using cosine similarity.

[0029] Furthermore, x% is taken as 5%-10%, and y% is taken as 85%-95%.

[0030] S13. For data Dx that is original data or has L2 level permission permission status, issue permission certificates on the blockchain, create a node corresponding to data Dx in the graph database, write the initial contribution weight as an attribute into the evolution edge between the node of data Dx and the most similar source data node, and mark data Dx as active in the graph database.

[0031] Specifically, after the rights are confirmed, the blockchain issues a permission certificate, specifically a three-rights certificate. At this point, the node containing data Dx is marked as active in the graph database, and its initial contribution weight serves as the input parameter for the subsequent revenue sharing algorithm.

[0032] In this embodiment, the three rights certificates include the right to hold data resources, the right to process and use data, and the right to operate data products. The right to hold data resources refers to the data subject's actual control, management, and possession of the original data resources, and is a fundamental right in the rights confirmation system. The right to process and use data refers to the right, after obtaining authorization from the holder, to clean, de-identify, analyze, and further develop the original data to form intermediate data or data products. The right to operate data products refers to the right to trade, distribute, and profit from the processed data products that have commercial value.

[0033] This step demonstrates the processing logic when data Dx enters the system. Compared to traditional evidence storage, the method provided in this step focuses on generating initial contribution weights through similarity metric standardization, thus providing a technical basis for subsequent value-added revenue distribution.

[0034] S2. Based on the revenue P of data Dx, the rights confirmation and control center recursively queries the lineage path to the graph database and extracts the contribution weight. The contribution weight is then uploaded to the blockchain's ledger contract. The ledger contract calculates the profit share of each level of participants, transfers funds according to the profit share, and reports the settlement results back to the rights confirmation and control center.

[0035] In a specific embodiment, such as Figure 4 As shown, step S2 may include: S2.1 When product Dc, formed from data Dx, is traded and generates revenue P, operator C submits a payment settlement request to the rights confirmation and control center. The rights confirmation and control center initiates a depth-first traversal request to the graph database to retrieve the complete lineage path from data Dx to product Dc, and to extract the contribution weight of records on all evolutionary edges in the lineage path.

[0036] Specifically, this step is implemented through recursive querying of lineage paths. When operator C develops product Dc based on data Dx uploaded by the original holder A and completes a transaction to generate revenue P, the rights confirmation and control center initiates a depth-first traversal request to the graph database. The system automatically retrieves the attributes of all evolutionary edges from data Dx to the final product Dc to extract all recorded contribution weights.

[0037] S2.2 The Confirmation and Control Center will transfer the extracted contribution weights to the ledger contract deployed on the blockchain. The ledger contract will calculate the share of interests of each level of participants in the lineage path according to the preset algorithm.

[0038] Specifically, this step is executed through automatic liquidation logic. The rights confirmation and control center transmits the extracted contribution weight to the ledger contract on the blockchain. The ledger contract calculates the share of benefits due to each party according to a preset algorithm (e.g., ledger amount = total revenue * path retention weight). The path retention weight is the contribution weight.

[0039] S2.3 The revenue sharing contract executes the transfer action on the blockchain, transferring the calculated profit share to the accounts of the participants at each level, and reporting the settlement results to the rights confirmation and control center.

[0040] Specifically, this step enables real-time settlement of funds. The smart ledger contract automatically executes transfer actions on the blockchain, transferring funds to the accounts of participants at each level according to the calculated profit share, and reporting the settlement results back to the rights confirmation and control center, completing hash-based evidence storage.

[0041] This step enables automatic settlement and intelligent revenue sharing in multi-level processing scenarios, realizing "transaction as revenue sharing," and technically solving the pain points of difficult and opaque profit distribution when multiple entities participate in data flow.

[0042] S3. In response to the change operation of revoking the authorization of data asset Da, locate all downstream data nodes derived from data asset Da in the lineage graph, issue control instructions to the blockchain to modify the contract status of the downstream data nodes derived from data asset Da to the restricted state, and realize physical-level circuit breaker control.

[0043] In a specific embodiment, such as Figure 5 As shown, step S3 may include: S3.1, Original holder A initiates a change operation on the blockchain to revoke the authorization of data asset Da. The chain graph collaboration controller captures the change operation and calls the graph database to search in order to locate all downstream data nodes derived from data asset Da in the lineage graph.

[0044] Specifically, the original holder A initiates a change operation on the blockchain to revoke the authorization of the data asset Da. The chain graph collaboration controller captures this change operation and immediately invokes the breadth-first search algorithm of the graph database to locate all downstream data nodes derived from the data asset Da in the lineage graph.

[0045] S3.2. Issue control commands to the blockchain to modify the contract status of the downstream data nodes derived from the data asset Da to a restricted state, thereby achieving physical-level circuit breaker control.

[0046] Specifically, the system no longer just stays at the "early warning" level, but directly issues control instructions to the blockchain to set the smart contract status of all related downstream assets to a restricted state such as "invalid" or "suspended operation".

[0047] This embodiment provides a system that utilizes lineage graphs to achieve rapid control of downstream assets when the authorization of data Dx expires or a violation occurs. Even if operator C's product Dc is stored in a heterogeneous cloud, physical-level "extended control" over the flow of data elements is achieved because its transaction and verification entry points are locked by blockchain contracts.

[0048] like Figure 6 As shown, Figure 6 This document details how the system leverages the non-linear retrieval capabilities of graph databases to automate value allocation and pricing when processing asset transactions. When a consumer initiates a purchase request, the system first executes a reverse tracing algorithm to locate all upstream dependent nodes of the target asset in Neo4j. During this process, the system extracts two core technical variables: first, the "feature retention rate" of each processing stage, generated and solidified as edge attributes by the similarity detection engine during the registration phase, used to quantify the technical contribution weights of each level of entity; and second, the "topology out-degree" of the data source node, used to assess the scarcity of this original element across the entire network. The pricing engine automatically generates a dynamic price P by integrating these variables. After the transaction is completed, the intelligent settlement module executes recursive settlement logic based on the contribution weights of the entire path, automatically allocating funds proportionally to the corresponding multi-level wallets on the blockchain. This process is completely free from manual intervention, solving the technical challenge of revenue distribution in multi-entity, long-chain processing scenarios by transforming "lineage relationships" into "settlement parameters."

[0049] like Figure 7As shown in the diagram, this diagram details the technical path by which the system achieves strong cross-entity, end-to-end access control. Its core logic lies in establishing a "real-time mirror synchronization" between the graph node state and the blockchain contract interface. When the original data provider executes a revocation or modification instruction on the blockchain, the system not only modifies the certificate state of that specific node but also immediately invokes a cascading circuit breaker to initiate a graph scan. The circuit breaker uses a breadth-first search (BFS) algorithm to traverse all evolution paths below the node, identifying all affected second-level and higher-level derivative asset nodes. Based on the downstream asset list fed back by the graph database, the system automatically locks the transaction contract interfaces of these assets on the blockchain (i.e., a physical circuit breaker), immediately depriving them of trading and access capabilities. This extended control mechanism achieves a physical-level control effect of "source failure, full-path blocking," overcoming the technical shortcomings of traditional systems in tracing and controlling data after cross-domain transfers, and providing a closed-loop technical barrier for compliant data circulation.

[0050] like Figure 8 As shown, this embodiment, through a multi-stage evolution model, details the full lifecycle management logic of data assets from initial access to secure market exit, focusing on the real-time mapping relationship between asset status and the "lineage graph" and "blockchain ledger." Asset Initialization Stage (Registration Access State): When data Dx enters the system, the similarity detection engine first performs originality verification. After successful verification, the system initializes node Dx in the Neo4j graph database and assigns it initial technical attributes (such as L2 circulation level). At this time, the asset is in a "pending activation" state. Certificate Activation Stage (Rights Confirmation and Circulation State): The chain graph collaborative controller issues a three-rights separation certificate (holding, processing, and operation) for Dx on the blockchain. At this time, the "valid certificate" on the chain is logically anchored to the "legitimate node" in the graph, and the asset officially becomes eligible for market trading. Value-Added Evolution Stage (Lineage Update State): The asset is processed into derivative Dy during circulation. The system extracts the feature retention ratio of Dy to Dx in real time and writes this "contribution weight" into the edge attributes of the graph. At this point, the asset's state evolves from a "single node" to a "networked lineage structure," providing a foundation for subsequent calculations. Value Realization Phase (Dynamic Monetization State): When a transaction occurs, the system calls the graph computing engine to extract topological features (out-degree, depth), generates a quote through the dynamic pricing engine, and drives the smart contract to execute automatic revenue sharing based on lineage weights. At this time, the asset is in an active value exchange state. Security Termination Phase (Cascading Circuit Breaker State): When the source data experiences compliance risks or authorization expires, the cascading circuit breaker is activated. The system identifies all downstream dependent nodes using a breadth-first traversal algorithm and forcibly locks the corresponding on-chain contract interface to an "invalid" state. This phase marks the formal exit of the asset and its derivatives from market circulation, achieving a secure closed loop across the entire chain.

[0051] Combination Figure 2 , Figure 6 This embodiment utilizes a graph database to maintain a dynamically evolving data lineage graph. When an asset enters the trading process, the dynamic pricing engine extracts the structured parameters of the target node from the graph computing engine. Specifically, the system extracts the "out-degree" (reflecting data popularity and reuse rate) and "topology depth" (reflecting processing level) of the node and its upstream nodes in the lineage network. The pricing model uses these topology factors as independent variables, combined with a preset decay function, to automatically generate the real-time market price of the asset. This process achieves algorithm-driven objective valuation.

[0052] like Figure 6 As shown, during the transaction proceeds settlement phase, the smart revenue sharing module initiates a reverse tracing process. The system automatically reads all parent nodes and their processing edge attributes corresponding to the asset in the lineage graph. The edge attributes record the "feature retention rate" (i.e., the quantified technical contribution weight) generated by the similarity detection algorithm during the registration phase. Based on the full-path weight distribution returned by the algorithm, the smart contract automatically and recursively settles the transaction funds to the digital wallets of the original data holders and processors at each level. This ensures that, in multi-party collaborative scenarios, the distribution of benefits always aligns with the actual technical contributions of each party.

[0053] like Figure 7 As shown, this embodiment achieves extended control over downstream derivative assets through a "strong state binding" mechanism. When the original rights holder triggers a "permission revocation" or "certificate invalidation" command on the blockchain, the cascading circuit breaker simultaneously acquires this event. The system utilizes a breadth-first search algorithm in the graph database to quickly retrieve all downstream dependent nodes (including second- and multi-level derivative assets) along the lineage graph of the failed node. Subsequently, the circuit breaker issues a control command to the blockchain, forcibly placing the smart contract interfaces for all related derivative products into a "suspended" state. This mechanism technically ensures that changes in source rights can penetrate the entire chain in real time, achieving physical-level compliance interception.

[0054] like Figure 8As shown, during the initialization and activation phase, the system not only records ownership on the blockchain, but more importantly, it pre-defines the asset's "evolutionary boundary (L2 level)" in the graph database, laying the technical groundwork for all subsequent automated controls. Entering the value-added evolution and value realization phase, this embodiment demonstrates its core algorithmic capabilities: the system no longer relies on static paper contracts, but instead extracts "feature retention" and "topological out-degree" from the graph's edge attributes in real time as parameters, driving on-chain smart contracts to execute dynamic pricing and multi-level recursive revenue sharing, ensuring that every share of revenue is traceable and automatically executed. The most critical innovation lies in the secure termination phase. Unlike traditional systems that can only revoke a single certificate, this solution, when the source state deviates (such as authorization revocation), utilizes the path dependencies of the graph database to achieve "cascading transmission" of permission states through a breadth-first traversal algorithm. This design ensures that even if data has undergone multiple processing steps and exists in multiple heterogeneous products, the system can still accurately locate each affected end contract and execute physical-level circuit breaker interception. This closed-loop process fully embodies the technological evolution of this invention from "rights confirmation and evidence preservation" to "proactive security governance," significantly enhancing the technical strength and patentability of the solution.

[0055] To address the shortcomings of existing technologies that rely solely on hash comparison to fail to identify plagiarized similar data, this invention aims to establish a pre-detection mechanism based on feature vector extraction and dynamic similarity retrieval. This mechanism can identify infringing assets with excessively high similarity during the registration process, thereby protecting the originality of data resource ownership.

[0056] To address the problem that existing revenue-sharing mechanisms rely on static contracts and cannot quantify data processing contributions, this invention aims to extract data lineage evolution paths and feature retention coefficients from graph databases, and use smart contracts to automatically execute revenue-sharing algorithms based on "lineage contribution," thereby solving the problem of automated settlement of revenue attribution in multi-entity collaborative scenarios.

[0057] To address the "control circuit breaker" problem in existing technologies where the flow of downstream derivative assets cannot be immediately blocked after the source authorization is revoked, this invention aims to establish permission synchronization logic based on topology paths using a "chain-graph collaboration" architecture. When the source node fails, the system can automatically trigger a "cascading circuit breaker" across the entire operational interface within milliseconds, ensuring the scalability and real-time nature of data asset control.

[0058] To address the shortcomings of existing data transaction pricing, such as strong subjectivity and lack of real-time dimensions, this invention aims to extract "topological characteristic factors" (such as out-degree of reference and lineage depth) of data nodes from graph databases, construct a dynamic pricing function, and realize the automatic adjustment of the value of data assets according to market reuse rate and scarcity, thereby reducing the transaction costs of factor circulation.

[0059] To address the problem of existing technologies where ownership certificates and actual processing flows are not separate entities, this invention aims to automatically complete the ownership penetration verification of the entire path before executing any business or revenue-sharing behavior by anchoring the bloodline map and blockchain credentials in real time, thus providing a closed-loop technical governance architecture for the "separation of three rights".

[0060] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0061] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings, disclosure, and appended claims in carrying out the claimed invention. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0062] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, any modifications made without departing from the inventive concept should be considered within the scope of protection of the present invention.

Claims

1. A digital asset management method based on lineage topology computation, characterized in that, include: S1. In response to the registration request, extract the semantic features of the data Dx uploaded by the original holder A to generate a high-dimensional fingerprint. Based on the high-dimensional fingerprint, search for the most similar source data node in the graph database. Based on the similarity between the data Dx and the source data node, determine the data Dx to obtain a determination result. Based on the determination result, issue an authorization certificate on the blockchain and create the node and evolution edge corresponding to the data Dx in the graph database, where x < y. S2. Based on the revenue P of the data Dx, the rights confirmation and control center recursively queries the bloodline path to the graph database and extracts the contribution weight. The contribution weight is then uploaded to the blockchain's ledger contract. The ledger contract calculates the profit share of each level of participants, transfers funds according to the profit share, and reports the settlement result back to the rights confirmation and control center.

2. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, Extract semantic features from the data Dx uploaded by the original holder A to generate a high-dimensional fingerprint, including: The semantic features of the data Dx uploaded by the original holder A are extracted and mapped to the same vector space, and then mapped to the same hash code space to obtain a high-dimensional fingerprint with semantic awareness characteristics.

3. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, Extract the semantic features of the data Dx uploaded by the original holder A, map them to the same vector space, and then map them to the same hash code space to obtain a high-dimensional fingerprint with semantic awareness characteristics, including: The data Dx uploaded by the original holder A is input into the CLIP model. The CLIP model is used to extract semantic features from the data Dx. The extracted semantic features are mapped to the same vector space to obtain a semantic feature vector. Then, the semantic feature vector is mapped to the same hash code space through an MLP network to obtain the high-dimensional fingerprint of the data Dx.

4. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, Based on the high-dimensional fingerprint, the most similar source data node is retrieved in the graph database. The data Dx is then judged based on its similarity to the source data node to obtain a judgment result, including: Calculate the similarity between the high-dimensional fingerprint of the data Dx and the high-dimensional fingerprint of each source data node in the graph database. Select the source data node with the highest similarity as the most similar source data node. Determine whether the data Dx is original data or whether the data Dx has the L2 level permission processing authorization of the most similar source data node when the highest similarity is within the preset similarity interval [x%, y%]. If so, use the feature retention degree of the data Dx to the most similar source data node as the initial contribution weight of the data Dx. Here, L2 level permission means that the system allows the generation of tradable operating rights certificates for data assets marked with this level, and supports their subsequent evolution and automatic accounting in the graph database.

5. The digital asset management method based on lineage topology calculation according to claim 4, characterized in that, The data Dx is determined based on the relationship between the maximum similarity and the preset similarity interval [x%, y%]. This determines whether the data Dx is original data or whether, when the maximum similarity is within the preset similarity interval [x%, y%], the data Dx possesses L2-level permission authorization from the most similar source data node. If so, the feature retention degree of the data Dx towards the most similar source data node is used as the initial contribution weight of the data Dx, including: The system determines the relationship between the maximum similarity and the preset similarity interval [x%, y%]. If the maximum similarity is less than x%, the data Dx is determined to be original data, and the data Dx is set to have either an L1-level permission or an L2-level permission. If the maximum similarity is greater than y%, the data Dx is not uploaded. If the maximum similarity is greater than or equal to x% and less than or equal to y%, the system continues to determine whether the data Dx has L2-level permission authorization from the most similar source data node. If not, the data Dx is not uploaded. If yes, the system calculates the feature retention rate of the data Dx to the most similar source data node and uses the feature retention rate as the initial contribution weight of the data Dx. Here, L1-level permission means that the system prohibits the generation of independent tradable certificates for data assets marked with this level and locks their evolution path to unrestricted nodes in the graph database.

6. The digital asset management method based on lineage topology calculation according to claim 5, characterized in that, Based on the determination result, an authorization certificate is issued on the blockchain, and the node and evolutionary edge corresponding to the data Dx are created in the graph database, including: For data Dx that is original data or has L2-level permission permission status, a permission certificate is issued on the blockchain, and a node corresponding to the data Dx is created in the graph database. The initial contribution weight is written as an attribute into the evolution edge between the node of the data Dx and the most similar source data node, and the data Dx is marked as active in the graph database.

7. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, Based on the revenue P of the data Dx, the rights confirmation and control center recursively queries the graph database for kinship paths and extracts contribution weights, including: When product Dc, formed from the data Dx, completes a transaction and generates revenue P, operator C submits a payment settlement request to the rights confirmation and control center. The rights confirmation and control center then initiates a depth-first traversal request to the graph database to retrieve the complete lineage path from the data Dx to the product Dc, and to extract the contribution weights of records on all evolutionary edges in the lineage path.

8. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, The contribution weight is uploaded to the blockchain's ledger contract, which calculates the profit share of each participant, transfers funds according to the profit share, and reports the settlement results back to the rights confirmation and control center, including: The rights confirmation and control center will transfer the extracted contribution weight to the ledger contract deployed on the blockchain. The ledger contract will calculate the profit share of each level of participants in the bloodline path according to a preset algorithm. The revenue sharing contract executes transfer actions on the blockchain, transferring funds to the accounts of participants at each level according to the calculated profit share, and reporting the settlement results back to the rights confirmation and control center.

9. The digital asset management method based on lineage topology calculation according to claim 1, characterized in that, Following step S2, the following is also included: S3. In response to the change operation of revoking the authorization of data asset Da, locate all downstream data nodes derived from the data asset Da in the lineage graph, and issue control instructions to the blockchain to modify the contract status corresponding to the downstream data nodes derived from the data asset Da to the restricted state, thereby realizing physical-level circuit breaker control.

10. The digital asset management method based on lineage topology calculation according to claim 9, characterized in that, In response to a change operation that revoks the authorization of data asset Da, locate all downstream data nodes derived from said data asset Da in the lineage graph, including: The original holder A initiates a change operation on the blockchain to revoke the authorization of the data asset Da. The chain graph collaboration controller captures the change operation and calls the graph database to search in order to locate all downstream data nodes derived from the data asset Da in the lineage graph.

Images