Digital asset security verification and information monitoring method and system
Through a multi-dimensional classification model and adaptive verification algorithm, combined with digital asset association map and multi-algorithm adaptation layer, in-depth verification analysis and multi-factor risk assessment are carried out, and a blockchain verification proof mechanism is built, solving the difficulty of unified verification in existing technologies. The problems of different types of digital assets have achieved efficient and secure verification and monitoring of digital assets.
Patent Information
- Application Number
- CN202411559126.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-04
- Publication Date
- 2025-02-11
- Estimated Expiration
- 2044-11-04
AI Technical Summary
The existing technology is difficult to design a comprehensive digital asset verification system, which can uniformly verify the legitimacy and ownership of different types of digital assets, and handle the correlation and dynamics between digital assets, while being highly secure and tamper-proof. ability.
Classify digital assets through multi-dimensional classification models, design adaptive verification algorithms, build digital asset association maps, use graph neural network algorithm to calculate verification weights, discover transaction association rules and abnormal patterns, design multi-algorithm adaptation layers, and use intelligent data Analytical and conversion technology, conduct in-depth verification analysis and multi-factor risk assessment, and finally build a blockchain verification proof mechanism.
It realizes comprehensive, intelligent and scalable security verification and information monitoring of different types of digital assets, improves the security and credibility of digital assets, and ensures the tamper-proof and traceability of verification results.
Smart Images

Figure CN119416179A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of information technology, and in particular to a digital asset security verification and information monitoring method and system. Background Art
[0002] In the process of security verification of digital asset portfolios, since the portfolio contains multiple types of digital assets, each digital asset has its own unique attributes and verification methods, it is necessary to design a comprehensive verification system that can uniformly verify the legitimacy and ownership of different digital assets. At the same time, since there may be correlations between digital assets, for example, some digital assets may be derived from other digital assets, or there are transaction relationships between digital assets, this requires the verification system to be able to identify and process these correlations and perform weighted verification on the associated digital assets to ensure the integrity and credibility of the entire portfolio. In addition, different digital assets may use different encryption algorithms, consensus mechanisms and data structures. The verification system needs to be compatible with these differences and be able to track changes and updates of digital assets in real time. At the same time, since the value of digital assets is often high, the verification system also needs to have a high degree of security and anti-tampering capabilities, and can effectively resist various counterfeiting, tampering and attack behaviors to ensure the authenticity and reliability of the portfolio. Therefore, designing a comprehensive, reliable and secure digital asset portfolio verification system requires comprehensive consideration of multiple dimensions such as digital asset attributes, correlations, and dynamics, and the use of advanced cryptography, consensus algorithms and data analysis technologies to build a highly intelligent and adaptive verification framework to provide solid technical support for the management and trading of digital asset portfolios. Summary of the invention
[0003] The present invention provides a digital asset security verification and information monitoring method, which mainly includes: According to the type attributes of digital assets, a multi-dimensional classification model is used to classify digital assets, obtain the classification labels and feature vector representations of digital assets, and establish a digital asset feature database; Design an adaptive verification algorithm for the classified digital asset feature database, dynamically select matching verification rules and threshold parameters to achieve comprehensive verification of the legitimacy and ownership of digital assets, and generate digital asset verification results; Based on the digital asset verification results and the correlation between digital assets, a digital asset correlation graph is constructed, and the graph neural network algorithm is used to jointly model the related digital assets. Through message transmission and feature aggregation, the verification weight of the digital asset node is calculated; Obtain the transaction history data of digital assets from the digital asset association graph, use time series pattern mining technology to discover the association rules and abnormal patterns of digital asset transactions, and form a transaction verification rule library; Apply the transaction verification rule base to the digital asset verification process, combine the encryption algorithm and consensus mechanism of digital assets, design a multi-algorithm adaptation layer, and achieve compatible processing of different types of digital assets through algorithm registration and dynamic scheduling; According to the data structure of the algorithm adaptation layer, intelligent data analysis and conversion technology is used to map heterogeneous digital asset data into a unified feature space, generate standardized digital asset data, and form a digital asset feature database; Utilize the digital asset feature database to conduct in-depth verification and analysis, obtain the security attribute information of digital assets, adopt a multi-factor risk assessment model to dynamically assess the security level of digital assets, and generate digital asset security rating results; Based on the digital asset security rating and verification results, a blockchain verification and proof mechanism is constructed. By generating a hash value containing the verification process, results and timestamp, a hash proof chain for digital asset verification is formed to ensure the tamper-proof and traceability of the verification results.
[0004] The present invention provides a digital asset verification and information monitoring system, which mainly includes: Digital asset classification module, used to classify digital assets in multiple dimensions; Digital asset verification module, used to verify the legitimacy and ownership of digital assets; Digital asset association modeling module, used to build digital asset association graphs and weight calculations; Transaction verification rule mining module, used to discover association rules and anomalies of digital asset transactions; Multi-algorithm adaptation processing module, used to process different types of digital assets compatibly; The digital asset security assessment module is used to assess the security level of digital assets.
[0005] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses a digital asset security verification and information monitoring method. The present invention classifies digital assets in multiple dimensions, establishes a digital asset feature database, and designs an adaptive verification algorithm to comprehensively verify the legitimacy and ownership of digital assets. On this basis, the present invention constructs a digital asset association graph, calculates the verification weight of digital asset nodes using a graph neural network algorithm, and discovers the association rules and abnormal patterns of digital asset transactions through time series pattern mining technology to form a transaction verification rule library. The present invention also designs a multi-algorithm adaptation layer to achieve compatible processing of different types of digital assets, and uses intelligent data parsing and conversion technology to generate standardized digital asset data. Finally, the present invention uses deep verification analysis and a multi-factor risk assessment model to dynamically evaluate the security level of digital assets, and constructs a blockchain verification and proof mechanism to ensure the tamper-proof and traceability of the verification results. The present invention provides a comprehensive, intelligent, and scalable digital asset security assessment solution, which effectively improves the security and credibility of digital assets. BRIEF DESCRIPTION OF THE DRAWINGS
[0006] Figure 1 The present invention is a flowchart of a digital asset security verification and information monitoring method.
[0007] Figure 2 A schematic diagram of a digital asset security verification and information monitoring method and system of the present invention.
[0008] Figure 3 This is another schematic diagram of a digital asset security verification and information monitoring method and system of the present invention.
[0009] Figure 4 It is a structural schematic diagram of a digital asset security verification and information monitoring system of the present invention. DETAILED DESCRIPTION
[0010] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.
[0011] like Figure 1-4 In this embodiment, a digital asset security verification and information monitoring method and system may specifically include: Step S101, classify the digital assets according to the type attributes of the digital assets using a multi-dimensional classification model, obtain the classification labels and feature vector representations of the digital assets, and establish a digital asset feature database.
[0012] Acquire multi-dimensional attribute information of digital assets, cluster digital assets using K-means clustering algorithm, and obtain coarse-grained classification labels of digital assets; for each coarse-grained classification, extract the common features of digital assets of this type, and build a TF-IDF feature vector representation model for this type of digital assets; classify the feature vectors of digital assets using support vector machine (SVM) classification algorithm, and obtain fine-grained classification labels of digital assets; store the classification labels of digital assets and corresponding feature vectors in MongoDB database, and establish a mapping relationship between digital asset labels and features; when receiving a digital asset retrieval request from a user, obtain the digital asset classification label entered by the user, and obtain the digital asset corresponding to the digital asset classification label from the MongoDB database feature vector; calculate the cosine similarity between the user's portrait feature vector and the digital asset feature vector, and recommend the top-N digital assets with the highest similarity to the user; when there are new digital assets, extract the multi-dimensional attribute information of the new digital assets, obtain their feature vectors through the TF-IDF feature vector representation model, use the SVM classification model algorithm to determine their category, and insert the new digital asset classification label and feature vector information into the MongoDB database; regularly maintain and update the digital asset feature data in the MongoDB database, update its classification label and feature vector representation according to the dynamic changes of digital assets, reduce the update frequency of digital assets that have not changed for a long time, and increase the update frequency of digital assets that change frequently.
[0013] Specifically, digital assets can contain many dimensions, such as issuer, issuance time, total amount, circulation, consensus mechanism, application scenarios, etc. These dimensions can be quantified, for example, the issuance time can be represented by a timestamp, and the consensus mechanism can be represented by a unique hot encoding. Suppose that multi-dimensional attribute information of 1,000 digital assets is collected. Use the K-means algorithm to cluster these digital assets. Set the number of clusters K to 5, and the algorithm will divide the 1,000 digital assets into 5 clusters. Suppose that one of the clusters contains mainstream cryptocurrencies such as A digital currency, another cluster contains various decentralized tokens, and another cluster contains NFT-related digital assets. In this way, coarse-grained classification labels of digital assets are obtained, such as "mainstream cryptocurrencies", "decentralized tokens", "NFT assets", etc. The reason for coarse-grained classification is that direct fine-grained classification will increase the computational complexity, while coarse-grained classification can help narrow the scope and improve efficiency. For the coarse-grained classification of "mainstream cryptocurrencies", the common characteristics of this type of digital assets can be extracted. For example, they usually have a large market value, high trading volume, long history, and broad community consensus. The TF-IDF algorithm can be used to construct a feature vector representation model. The TF-IDF algorithm can evaluate the importance of a word to a document set or one of the documents in a corpus. Here, each digital asset can be regarded as a "document" and their attributes as "words". For example, the TF-IDF feature vector of "A digital currency" may contain features such as "high market value", "large trading volume", and "long history", and their weights are calculated by the TF-IDF algorithm. The support vector machine SVM algorithm is used to classify the feature vectors of digital assets to obtain fine-grained classification labels of digital assets. For example, under the coarse-grained classification of "mainstream cryptocurrency", it can be further subdivided into "POW consensus mechanism", "POS consensus mechanism", etc. Through the SVM algorithm, "A digital currency" can be divided into the fine-grained classification of "POW consensus mechanism". The classification labels (coarse-grained and fine-grained) and corresponding feature vectors of digital assets are stored in the MongoDB database. MongoDB is a non-relational database that can flexibly store various data types. A collection can be established to store the information of digital assets, including names, classification labels, feature vectors, etc. For example, the name of "A digital currency", the two classification labels of "mainstream cryptocurrency" and "POW consensus mechanism", and its TF-IDF feature vector can be stored in the MongoDB database. The advantage of doing this is that it is convenient for subsequent retrieval and update. When the user enters the classification label "POW consensus mechanism" for retrieval, the system will obtain the feature vectors of all digital assets belonging to the "POW consensus mechanism" from the MongoDB database. Assume that the user's portrait feature vector contains features such as "high security" and "high degree of decentralization".The system calculates the cosine similarity between the user portrait feature vector and each digital asset feature vector. The higher the cosine similarity, the higher the match between the user and the digital asset. The system recommends the top-N digital assets with the highest similarity to the user. When a new digital asset called "ABC Token" is added, its multi-dimensional attribute information is extracted. The feature vector of "ABC Token" can be obtained through the previously constructed TF-IDF feature vector representation model. Then, the trained SVM classification model is used to determine the category to which "ABC Token" belongs, such as "mainstream cryptocurrency" and "POS consensus mechanism". Finally, the classification label and feature vector information of "ABC Token" are inserted into the MongoDB database. The digital asset feature data in the MongoDB database is regularly maintained and updated. For digital assets with long-term stable properties such as A digital currency, the update frequency can be reduced. For some decentralized tokens that change frequently, the update frequency needs to be increased. This can ensure the accuracy and timeliness of the data in the database, thereby improving the effectiveness of the recommendation system.
[0014] Step S102, designing an adaptive verification algorithm for the classified digital asset feature database, and realizing comprehensive verification of the legitimacy and ownership of digital assets by dynamically selecting matching verification rules and threshold parameters, and generating digital asset verification results.
[0015] A digital asset feature database is obtained, which contains feature information of various digital assets; a preset candidate verification rule set and a threshold parameter range are obtained according to the coarse-grained classification labels of the classification information of the digital assets; a decision tree algorithm is used to select the verification rule subset that is most suitable for the digital asset to be verified from the candidate verification rule set. The feature selection of the decision tree is based on the feature information of the digital asset. By calculating the information gain ratio of each feature, the feature with the largest information gain ratio is selected as the split node, and a decision tree is recursively constructed until the preset stop condition is met; a random forest algorithm is used to dynamically adjust the threshold parameters of the verification rule subset. The random forest algorithm selects sample subsets and feature subsets at random. A set of decision trees is constructed, and the prediction results of the decision trees are integrated to obtain the adjustment value of the threshold parameter; the adjusted threshold parameter is used to determine whether the digital asset to be verified meets the preset ownership verification rule, and if so, the legitimacy of the digital asset to be verified is judged and verified; the weighted average method is used to calculate the comprehensive verification score of the digital asset to be verified, and the weight coefficient of the weighted average is preset according to the importance of ownership verification and legitimacy verification; the comprehensive verification score is compared with the preset threshold to obtain the final verification result of the digital asset to be verified; the final verification result, the characteristics of the digital asset to be verified, the ownership verification result and the legitimacy verification result information are written into the digital asset verification result database.
[0016] Specifically, a digital asset feature database is obtained, which contains feature information of various digital assets. For example, the database stores information of digital assets such as digital currency A, including issuance time, total amount, circulation, consensus mechanism, application scenarios, historical price trends, etc. This information constitutes the characteristics of digital assets and can be used in subsequent verification processes. According to the coarse-grained classification labels of digital assets (such as "mainstream cryptocurrency", "decentralized token", "NFT asset", etc.), a preset candidate verification rule set and threshold parameter range are obtained. For example, for "mainstream cryptocurrency", candidate verification rules may include "market value greater than 10 billion US dollars", "number of trading platforms greater than 100", "code open source and security audited", etc. The threshold parameter range can be the lower limit of market value, the lower limit of the number of trading platforms, etc. A decision tree algorithm is used to select the verification rule subset that is most suitable for the digital asset to be verified from the candidate verification rule set. Assume that a new digital asset named "ABC token" is to be verified, and its coarse-grained classification label is "mainstream cryptocurrency". The decision tree algorithm will select the most relevant verification rule based on the feature information of "ABC token", such as market value, number of trading platforms, whether the code is open source, etc. For example, if the market value of "ABC Token" is only 1 billion US dollars, then the rule of "market value greater than 10 billion US dollars" is not applicable and will be excluded by the decision tree algorithm. The decision tree algorithm is used because it can select the most appropriate verification rules according to the different characteristics of digital assets to improve the efficiency and accuracy of verification. The random forest algorithm is used to dynamically adjust the threshold parameters of the verification rule subset. Assume that the decision tree algorithm selects the rule of "the number of trading platforms is greater than 50" and the initial threshold is 50. The random forest algorithm will dynamically adjust this threshold by analyzing historical data. For example, if historical data shows that "mainstream cryptocurrencies" with more than 80 trading platforms are more reliable, the random forest algorithm will adjust the threshold to 80. The random forest algorithm is used because it can integrate the prediction results of multiple decision trees to improve the accuracy and stability of the threshold parameters. The adjusted threshold parameters are used to determine whether the digital asset to be verified meets the preset ownership verification rules. Assuming that "ABC Token" is listed on 85 trading platforms and meets the adjusted threshold parameters, it is considered to have passed the ownership verification. The purpose of ownership verification is to confirm the authenticity and credibility of digital assets and prevent forgery and fraud. If the ownership verification rules are met, the legality of the digital asset to be verified is judged and verified. Legality verification mainly examines whether the digital asset complies with relevant laws, regulations and regulatory policies. For example, verify whether the issuance of "ABC token" is compliant and whether it involves illegal activities. The weighted average method is used to calculate the comprehensive verification score of the digital asset to be verified. Assume that the weight of ownership verification is 7 and the weight of legality verification is 3.The ownership verification score of "ABC Token" is 100 points, and the legality verification score is 90 points, so its comprehensive verification score is 7*100+3*90=97 points. The weighted average method can comprehensively consider the importance of different verification dimensions and obtain a more comprehensive verification result. Compare the comprehensive verification score with the preset threshold to obtain the final verification result of the digital asset to be verified. Assuming the preset threshold is 90 points, the comprehensive verification score of "ABC Token" is 97 points, and the final verification result is "passed". Write the final verification result, the characteristics of "ABC Token", the ownership verification result and the legality verification result information into the digital asset verification result database. The advantage of doing this is that it is convenient for subsequent queries and analysis, and it is also convenient for regulatory agencies to supervise and manage. The information stored in the database can be used to track the history of digital assets and identify potential risks and problems.
[0017] Step S103, based on the digital asset verification results and the correlation between digital assets, a digital asset association graph is constructed, and a graph neural network algorithm is used to jointly model the associated digital assets. The verification weight of the digital asset node is calculated through message passing and feature aggregation.
[0018] The digital asset verification results and the correlation between digital assets are obtained, and a digital asset correlation graph is constructed, in which nodes represent digital assets and edges represent the correlation between digital assets; the digital asset correlation graph is preprocessed, missing values are filled, outliers are filtered, and attribute characteristics and structural characteristics of nodes are extracted; the preprocessed digital asset correlation graph is modeled using a graph convolutional neural network algorithm to obtain a graph convolutional neural network model; at each layer of the graph convolutional neural network model, the feature representation of the node is updated through a message passing mechanism according to the adjacency relationship between nodes; at the last layer of the graph convolutional neural network model, the feature representation of the digital asset node is obtained; the correlation weights between nodes are calculated through an attention mechanism, and the attention mechanism The system includes converting the feature representation of the node into attention weight, and aggregating the features of adjacent nodes by weighted summation; concatenating the feature representation and attention weight of the digital asset node as the input of the fully connected layer, and obtaining the verification weight of each digital asset node through the fully connected layer; judging the importance of the digital asset nodes according to the verification weight, marking the digital asset nodes above the threshold as important nodes, and marking the digital asset nodes below the threshold as suspicious nodes; focusing on verifying the important nodes and suspicious nodes, and adding the verification results as new node attributes to the digital asset association map; dynamically updating the feature representation and verification weight of the digital asset node according to the verification results, and retraining the graph convolutional neural network model to obtain continuously optimized digital asset verification results.
[0019] Specifically, the digital asset association graph aims to depict the relationship between digital assets and provide a more comprehensive perspective for risk assessment and verification. The nodes in the graph represent individual digital assets, such as digital currency A, a specific NFT collection, etc. The edges represent the associations between them, such as being based on the same underlying technology, a common developer team, being listed on the same trading platform, etc. After building the graph, preprocessing is required to improve the accuracy of the model. For example, if the market value data of a digital asset is missing, it can be filled with the historical market value data of the asset, or estimated with the average market value of assets of the same category. For outliers, such as a sudden surge in the trading volume of a newly issued token, it is necessary to analyze the reasons behind it. If it is caused by market manipulation factors, it should be filtered out. In addition, it is crucial to extract the attribute features and structural features of the nodes. Attribute features can be market value, trading volume, issuance time, etc., while structural features can be the degree of the node (the number of edges connected to the node), centrality, etc. These features will be used for subsequent graph convolutional neural network modeling. Graph convolutional neural network is a deep learning model specifically designed to process graph data. In each layer, the node exchanges information with its neighbor nodes and updates its own feature representation. For example, if an NFT project is associated with a reputable development team, the credibility of the NFT project may increase, which is reflected in the change of its feature representation. The attention mechanism can capture important relationships between nodes. For example, the relationship between a stablecoin and the fiat currency it is anchored to is very important, while the relationship with a small exchange is relatively unimportant. The attention mechanism assigns different weights to different relationships to highlight key information. The fully connected layer combines the feature representation of the node with the attention weight to generate a verification weight for each node. This weight reflects the importance of the node and its potential risk. Nodes above the threshold are marked as important nodes, and nodes below the threshold are marked as suspicious nodes. Important nodes and suspicious nodes are verified with emphasis. For example, for a token marked as suspicious, its transaction records, smart contract code, etc. can be investigated in depth to determine whether it has fraud risks. The verification results will be added to the graph as new node attributes. The model will be dynamically updated according to the new verification results. For example, if a token that was previously considered important is found to have a security vulnerability, its verification weight will be reduced, and the model will adjust its parameters accordingly to improve the accuracy of subsequent verification. This continuous optimization process ensures that the model can adapt to the ever-changing digital asset environment. Through the digital asset association graph and graph convolutional neural network, high-risk and suspicious digital assets can be identified more effectively. For example, a newly launched decentralized trading platform is associated with several suspicious project parties, and these project parties have previously issued tokens with asset security risks. Through the association graph and model analysis, the potential risks of this trading platform can be discovered in time and early warnings can be issued.As another example, an NFT project claims to have a strong technical background and community support, but graph analysis shows that it has a low correlation with other well-known projects and communities, which may suggest that its publicity is exaggerated and investors need to be cautious.
[0020] Step S104, obtaining the transaction history data of the digital asset from the digital asset association graph, using the time series pattern mining technology to discover the association rules and abnormal patterns of digital asset transactions, and forming a transaction verification rule library.
[0021] Obtain the digital asset association graph, obtain the target digital asset node and its attribute information from the digital asset association graph; obtain the associated digital assets associated with the target digital asset according to the preset association relationship; obtain the transaction history data of the target digital asset and the associated digital asset, and the transaction history data includes the transaction time, transaction amount and transaction counterparty; pre-process the transaction history data, and fill the missing values using the preset data filling rules; align the transaction time of the target digital asset and the associated digital asset to generate a normalized time series transaction data set; use the Apriori algorithm to mine frequent transaction patterns from the time series transaction data set; compare the frequent transaction patterns with the preset normal transaction patterns; The transaction pattern rules are compared to determine whether the frequent transaction pattern is an abnormal transaction pattern; if the frequent transaction pattern does not conform to the normal transaction pattern rules, the frequent transaction pattern is marked as an abnormal transaction pattern; the abnormal transaction pattern is manually reviewed, and the manually confirmed abnormal transaction pattern is added to the pre-built abnormal transaction pattern library; new digital asset transaction data is obtained, and the digital asset information associated with the new digital asset transaction data is obtained based on the digital asset association map; the new digital asset transaction data is matched with the abnormal transaction pattern in the abnormal transaction pattern library; if the new digital asset transaction data matches the abnormal transaction pattern in the abnormal transaction pattern library, an abnormal transaction alarm message is generated.
[0022] Specifically, the digital asset association map is like a huge network of relationships, with each node representing a digital asset, such as digital currency A or a certain NFT collection. The lines between the nodes represent the relationship between them, such as being based on the same underlying technology, a common development team, being listed on the same trading platform, and so on. Obtaining a digital asset association map is to draw such a network in order to better understand the relationship between digital assets, so as to conduct risk assessment and verification. For example, if you want to investigate a new NFT project called "XYZ", you can obtain the project node and its attribute information from the map, such as the issuance time, total amount, issuance price, etc. In order to more comprehensively assess the risk of the "XYZ" project, you need to obtain the digital assets associated with it. Suppose the map shows that the "XYZ" project is associated with a development team called "ABC", and the "ABC" team has previously developed another NFT project "DEF". Then, the "DEF" project is associated with the "XYZ" project, and the "DEF" project can also be included in the scope of investigation. Obtain historical transaction data of "XYZ" and "DEF", including transaction time, transaction amount, and counterparty. For example, "XYZ" sold 1 NFT for 1ETH on January 1, 2024, and the buyer's address is 0x123. "DEF" sold 1 NFT for 5ETH on December 1, 2023, and the buyer's address is 0x456. These transaction data will be used for subsequent analysis. There may be missing values in the transaction data, such as the amount of some transactions is not recorded. In this case, it is necessary to use the preset data filling rules to fill them. For example, the missing transaction amount can be filled with the historical average transaction price of the asset. Since the transaction time of different digital assets may be different, the transaction time needs to be aligned to generate a normalized time series transaction data set. For example, the transaction data of "XYZ" and "DEF" are aggregated by hour for subsequent analysis. The Apriori algorithm is used to mine frequent transaction patterns from the time series transaction data set. For example, it may be found that the transaction volume of "XYZ" and "DEF" often rises sharply in the same time period, which may suggest that there is some kind of related transaction between the two. Compare the mined frequent transaction patterns with the preset normal transaction pattern rules. Normal trading pattern rules can be formulated based on historical data and market experience. For example, the trading volume of an NFT project is usually related to factors such as its market popularity and community activity. If the trading volume of a project suddenly soars, but its market popularity and community activity have not increased significantly, there may be an anomaly. If the frequent trading pattern does not conform to the normal trading pattern rules, it will be marked as an abnormal trading pattern. For example, if the trading volumes of "XYZ" and "DEF" soared simultaneously, but neither of them released any good news and there were no obvious market drivers, it could be marked as an abnormal trading pattern. Manually review the abnormal trading pattern to confirm whether it is indeed abnormal.For example, analysts can conduct in-depth investigations into the transaction records, capital flows, etc. of "XYZ" and "DEF" to determine whether there is market manipulation or other violations. Add manually confirmed abnormal transaction patterns to the pre-built abnormal transaction pattern library. For example, add the synchronous transaction volume surge pattern of "XYZ" and "DEF" to the abnormal transaction pattern library for subsequent monitoring. Assume that new digital asset transaction data is generated and needs to be matched with the pattern in the abnormal transaction pattern library. For example, a new NFT project called "GHI" also has a synchronous surge in transaction volume with "XYZ" and "DEF". If the new digital asset transaction data matches the pattern in the abnormal transaction pattern library, an abnormal transaction alarm message is generated. For example, the system will issue an alarm: "GHI"'s transaction pattern matches the synchronous transaction pattern of "XYZ" and "DEF" in the known abnormal transaction pattern library, and there may be risks. In this way, potential abnormal trading behaviors can be discovered in a timely manner, the interests of investors can be protected, and the healthy development of the digital asset market can be maintained.
[0023] Step S105, apply the transaction verification rule base to the digital asset verification process, combine the encryption algorithm and consensus mechanism of the digital assets, design a multi-algorithm adaptation layer, and achieve compatible processing of different types of digital assets through algorithm registration and dynamic scheduling.
[0024] According to the type of digital asset, the corresponding transaction verification rules are obtained from the digital asset registration center; the transaction verification rules are converted into executable code logic using smart contract technology and deployed to the blockchain network; according to the blockchain network to which the digital asset belongs, the consensus mechanism adopted by the blockchain network is determined; when a digital asset transaction request is received, the smart contract is called to verify the digital asset transaction request, and the verification includes checking the legitimacy of the format, signature and attributes of the digital asset transaction request; if the digital asset transaction request is verified, the digital asset transaction request is broadcast to the consensus node in the blockchain network; the consensus node verifies and confirms the digital asset transaction request according to the consensus mechanism; cryptography technology is used The technology protects the transaction verification process and verifies the legitimacy of digital asset transaction requests without disclosing the identity information of both parties to the transaction. For cross-chain digital asset transactions, a cross-chain protocol is introduced to achieve atomic exchange of digital assets in different blockchain networks through transaction proof mechanism and digital asset mapping mechanism. Blockchain nodes propagate, synchronize and verify blocks and transactions through P2P networks to ensure the decentralization and fault tolerance of the network. According to the business rules and processes of digital assets, smart contract technology is used to solidify the business rules and processes into the code to achieve automated management of the entire life cycle of digital assets. After security audits and tests, smart contracts are deployed to the blockchain network and jointly executed and verified by multiple nodes.
[0025] Specifically, the digital asset registration center is like a household registration office for digital assets, recording information about various digital assets, including asset type, issuer, total amount, and so on. For example, the registration information of an NFT collection called "Art Coin" will include its name, issuer (such as an art studio), total circulation (such as 10,000), and the information of artworks corresponding to each NFT. Obtaining the corresponding transaction verification rules from the digital asset registration center is to find the corresponding transaction rules according to different types of digital assets. For example, for NFT collections such as "Art Coin", its transaction rules may be: only an integer number of NFTs can be transferred in each transaction, and the transaction cannot be divided. Smart contracts are like self-executing contracts that convert transaction verification rules into code logic and deploy them on the blockchain. For example, for the transaction rules of "Art Coin", the smart contract will contain a code to check whether the number of NFTs in each transaction is an integer. If someone tries to trade 0.5 "Art Coins", the smart contract will automatically block the transaction. The blockchain network is like a distributed ledger, and all transaction records will be recorded on the blockchain. Different blockchain networks use different consensus mechanisms, such as proof of work (PoW) or proof of stake (PoS). PoS determines who has the right to verify transactions based on the number of digital assets held by the node. Assuming that "art coins" are issued on a blockchain that uses the PoS consensus mechanism, nodes that hold more "art coins" will be more likely to obtain the right to verify transactions. When someone wants to trade "art coins", he will send a transaction request. This transaction request will be sent to the smart contract and verified by the smart contract. The smart contract will check whether the format and signature of the transaction request and the properties of the transaction are legal. For example, the smart contract will check whether the signature of the transaction request matches the sender's private key and whether the number of NFTs traded is an integer. If the transaction request passes the verification of the smart contract, it will be broadcast to the consensus nodes in the blockchain network. The consensus nodes will verify and confirm the transaction according to the consensus mechanism of the blockchain. For example, under the PoS mechanism, nodes with more coins will vote on the transaction, and if the majority of nodes agree, the transaction will be confirmed and added to the blockchain. Cryptography technology is like a lock that protects the security of transactions. It can verify the legitimacy of a transaction without disclosing the identity information of both parties. For example, zero-knowledge proof technology can prove the legitimacy of a transaction without revealing the specific content of the transaction. For cross-chain transactions, for example, if someone wants to exchange digital currency A for "art coins", a cross-chain protocol needs to be introduced. The cross-chain protocol is like a bridge that connects different blockchain networks. Through the transaction proof mechanism and digital asset mapping mechanism, the atomic exchange of digital assets in different blockchain networks can be achieved. For example, a user can exchange 1 digital currency A for 10 "art coins", and the cross-chain protocol will ensure that the transaction is either completed at the same time or fails at the same time.Blockchain nodes communicate through a P2P network, just like a decentralized network where each node can be connected to each other. This network structure can ensure the decentralization and fault tolerance of the network. Even if some nodes fail, the network can still operate normally. Smart contracts can also be used to manage the entire life cycle of digital assets, such as the issuance, trading, and destruction of "art coins". These business rules and processes can be solidified into the code of smart contracts to achieve automated management. For example, smart contracts can set the total issuance of "art coins" and automatically control the issuance process of NFTs. Before being deployed to the blockchain network, smart contracts need to undergo security audits and tests to ensure their security. Multiple nodes will jointly execute and verify smart contracts to further improve security. For example, in the smart contract of "art coins", a security audit process may be set, and only after passing the audit can it be officially deployed on the blockchain.
[0026] Step S106, based on the data structure of the algorithm adaptation layer, intelligent data analysis and conversion technology is used to map the heterogeneous digital asset data into a unified feature space, generate standardized digital asset data, and form a digital asset feature database.
[0027] The original data of heterogeneous digital assets are obtained, and the original data includes the attributes, parameters and configuration information of the digital assets; the original data is parsed by using an XML parser to extract the key attribute fields of the digital assets and construct a digital asset attribute feature vector; the digital asset attribute feature vector is normalized according to a pre-defined unified feature space, and attribute values of different scales and types are mapped to the interval [0, 1] to obtain a standardized digital asset attribute feature vector; the digital asset attributes that cannot be directly mapped to the unified feature space are converted into a feature representation form of the unified feature space by using a One-Hot encoding data conversion method; the standardized digital asset attribute feature vector is filled into the corresponding dimension of the unified feature space to obtain a feature vector representation of the digital asset in the unified feature space; according to the feature vector of the digital asset in the unified feature space, a standardized digital asset data record is generated according to the predefined data format and field name; the standardized digital asset data record is imported into a database to establish a digital asset feature data table.
[0028] Specifically, the data formats and fields of heterogeneous digital assets are different. In order to uniformly process and analyze these data, they need to be converted into standardized forms. First, it is necessary to obtain the raw data of heterogeneous digital assets, which are usually stored in different formats, such as database records, XML files, JSON files, etc. Taking artwork NFT and game props as examples, the data of artwork NFT may contain fields such as name, author, creation time, work description, image hash value, etc., while the data of game props may contain fields such as name, type, level, attribute, durability, etc. After obtaining the raw data, the data needs to be parsed and extracted. For data in XML format, an XML parser, such as a DOM parser or a SAX parser, can be used. For example, the data of artwork NFT may be stored in an XML file, in which the information of each NFT is represented by an XML element. The XML parser can traverse the XML file and extract the key attribute fields of each NFT, such as name, author, image hash value, etc. After extracting the key attribute fields, it is necessary to construct a digital asset attribute feature vector. For example, the feature vector of an artwork NFT can be expressed as: [name, author, creation time, work description, image hash value]. Since the attribute values of different digital assets may have different scales and types, for example, the creation time is a numeric type, while the name is a string type, the feature vector needs to be normalized. Normalization can map attribute values of different scales and types to a unified interval, such as [0, 1]. For example, the creation time can be converted to a proportional value relative to the earliest creation time, and the name can be converted to a normalized value of a string hash value. In this way, a standardized digital asset attribute feature vector can be obtained. For some digital asset attributes that cannot be directly mapped to a unified feature space, such as the type of game props, One-Hot encoding can be used for conversion. Assume that there are three types of game props: weapons, armor, and consumables. Then a three-dimensional vector can be used to represent the type of props, for example, weapons are represented as [1, 0, 0], armor is represented as [0, 1, 0], and consumables are represented as [0, 0, 1]. Fill the standardized digital asset attribute feature vector and the feature vector after One-Hot encoding into the corresponding dimensions of the unified feature space, and you can get the feature vector representation of the digital asset in the unified feature space. For example, the feature vector of an artwork NFT can be expressed as: [normalized value of name, normalized value of author, normalized value of creation time, normalized value of work description, normalized value of image hash value]. The feature vector of a game prop can be expressed as: [normalized value of name, normalized value of level, normalized value of attribute, normalized value of durability, 1, 0, 0] (assuming that the prop is a weapon). Based on the feature vector of digital assets in a unified feature space, standardized digital asset data records can be generated according to predefined data formats and field names.For example, the feature vectors of both artwork NFTs and game props can be converted into data records containing the following fields: asset ID, asset type, feature vector. Finally, the standardized digital asset data records are imported into the database to establish a digital asset feature data table. This allows for unified storage, management, and analysis of different types of digital assets. The benefit of this is that operations such as data retrieval, statistical analysis, and machine learning can be easily performed.
[0029] Step S107, using the digital asset feature database, conduct in-depth verification and analysis to obtain security attribute information of digital assets, adopt a multi-factor risk assessment model to dynamically assess the security level of digital assets, and generate a digital asset security rating.
[0030] The digital asset feature information in the digital asset feature database is obtained to obtain a multi-dimensional feature vector of the digital asset to be evaluated; a training data set is constructed using the digital asset feature vector, and the training data set is trained through a deep neural network model to obtain a deep representation of the digital asset features; the deep representation of the digital asset features is input into a pre-constructed security attribute discrimination model, and the security attribute information of the digital asset is obtained through reasoning with the security attribute discrimination model, wherein the security attribute information includes confidentiality, integrity and availability levels; a pairwise comparison matrix is constructed using a hierarchical analysis method based on the obtained digital asset security attribute information, and the maximum eigenvalue pair of the pairwise comparison matrix is calculated. The corresponding eigenvector is obtained and normalized to obtain the weight of each assessment factor; the security attribute information of the digital asset and the weight of the assessment factor are weighted and summed to obtain the risk score of the digital asset under each assessment dimension; the risk score of the digital asset is tracked regularly and dynamically through the scheduled task module. When the score exceeds the preset threshold, the digital asset security level adjustment process is triggered and the digital asset security rating database is updated; the real-time security rating information of the digital asset is read from the digital asset security rating database, and a digital asset security assessment report is automatically generated. The report includes the basic information of the digital asset, the security attribute level, the risk score and the security rating change record.
[0031] Specifically, the digital asset feature database stores multi-dimensional feature vectors of each digital asset. For example, for an artwork NFT, its feature vector may include image hash value, author, creation time, artistic style, etc. For a game prop, its feature vector may include name, type, attribute, rarity, etc. Obtaining these multi-dimensional features of the digital asset to be evaluated forms the feature vector of the asset. For example, the feature vector of a game prop "Flaming Sword" can be ["Flaming Sword", "Weapon", "Attack +10", "Rare"]. Using the existing digital asset feature vectors, a training data set can be constructed. For example, a large amount of artwork NFT data, including their feature vectors and corresponding market prices, can form a training data set. Then, a deep neural network model is built through the TensorFlow framework to train the training data set, and a deep representation of digital asset features is obtained to train this data set. The deep neural network can obtain a deep representation of digital asset features by learning the complex relationship between feature vectors and prices. This is equivalent to mapping the original feature vector to a new space, in which similar digital assets will be closer. For example, after the original feature vector of "Flame Sword" is processed by a deep neural network, it may become a high-dimensional vector, such as [2,8,1,9,…], which can more effectively express the characteristics of "Flame Sword". Using historical data, a security attribute discrimination model for digital assets is constructed through a logistic regression model. By inputting the deep representation of digital asset characteristics into the security attribute discrimination model, the confidentiality, integrity and availability levels of the digital asset can be inferred. The security attribute discrimination model can determine the risk level of theft, tampering or inaccessibility of digital assets based on the deep characteristics of digital assets. For example, if the image hash value of an artwork NFT is easily tampered with, its integrity level will be low. If a game prop is easily copied, its confidentiality level will also be low. Assuming that the deep features of "Flame Sword" are input into the model, the model infers that its confidentiality level is "high", the integrity level is "medium", and the availability level is "high". Next, the pairwise comparison matrix is constructed using the hierarchical analysis method, and the eigenvector corresponding to the maximum eigenvalue of the pairwise comparison matrix is calculated and normalized to obtain the weight of each evaluation factor. For example, for game props, you may pay more attention to their availability and rarity, so the weights of availability and rarity will be higher. Assume that after the hierarchical analysis method calculation, the weights of confidentiality, integrity and availability are 2, 3, and 5 respectively. The security attribute information of the digital asset is weighted and summed with the evaluation factor weights to obtain the risk score of the digital asset under each evaluation dimension. The confidentiality, integrity and availability levels of "Flame Sword" are "high", "medium" and "high" respectively, and the corresponding scores can be set to 1, 5, and 1. Then the risk score of "Flame Sword" = 2*1+3*5+5*1=85.Digital assets are dynamically tracked regularly (e.g., every day) through the system's scheduled task module. If the risk score of "Flame Sword" exceeds the preset threshold (e.g., 9), the digital asset security level adjustment process is triggered and the digital asset security rating database is updated. For example, the system may lower the integrity level of "Flame Sword" to "low". The digital asset security rating database records the real-time security rating information of digital assets. This information can be read from the database to automatically generate a digital asset security assessment report. For example, the assessment report of "Flame Sword" will include its basic information (name, type, etc.), security attribute level (high, medium, low), risk score (85), and security rating change record (e.g., the integrity level is lowered from "medium" to "low"). Such a report can help users understand the security status of digital assets and make corresponding decisions.
[0032] Step S108, based on the digital asset security rating and verification results, a blockchain verification and certification mechanism is constructed. By generating a hash value including the verification process, results and timestamp, a hash certification chain for digital asset verification is formed to ensure the tamper-proof and traceability of the verification results.
[0033] Obtain the security rating and verification results of digital assets and use them as the original data of the verification proof; add the verification process description and current timestamp to the original verification data to generate complete verification proof data; perform SHA256 hash operation on the verification proof data to obtain a unique hash value as the fingerprint of this verification; concatenate the current verification hash value with the previous verification hash value to build a hash proof chain to form a continuous verification sequence; write the verification hash value and hash proof chain into the blockchain through the Ethereum smart contract; define verification rules and reward and punishment mechanisms through smart contracts to incentivize multiple nodes to participate in verification and reach a consensus on the verification results; use the blockchain browser to query and trace verification data to obtain historical verification information of digital assets; verify the authenticity of the current verification data by comparing historical verification hash values; synchronize the verification data in the blockchain to the local database, use SQL to perform statistical analysis on the verification data, calculate the security indicators of digital assets, evaluate the historical security status of digital assets, and generate a security analysis report.
[0034] Specifically, the digital asset security rating results, such as the confidentiality of "Flame Sword" is "high", the integrity is "medium", the availability is "high", the risk score is 85, and the verification results, such as "Verification passed", will be used as the original data of the verification proof. Add a description of the verification process, such as "A multi-factor risk assessment was performed on the feature vector of 'Flame Sword'", and the current timestamp, such as "October 17, 2024 10:00:00", to generate complete verification proof data. Perform a SHA256 hash operation on the complete verification proof data to obtain a unique hash value, such as "e4b58eb4642304a2270307021b3f42f3", as the fingerprint of this verification. This hash value can be used to uniquely identify this verification. The current verification hash value "e4b58eb4642304a2270307021b3f42f3" is concatenated with the previous verification hash value, such as "a1b2c3d4e5f678901234567890abcdef", for example "e4b58eb4642304a2270307021b3f42f3|a1b2c3d4e5f678901234567890abcdef", to construct a hash proof chain to form a continuous verification sequence. The hash proof chain can trace the historical record of verification. The verification hash value "e4b58eb4642304a2270307021b3f42f3" and the hash proof chain "e4b58eb4642304a2270307021b3f42f3|a1b2c3d4e5f678901234567890abcdef" are written into the blockchain through the Ethereum smart contract to ensure the immutability and traceability of the data. The smart contract defines the verification rules, such as the format of the verification data, the verification frequency, etc., as well as the reward and punishment mechanism, such as rewarding nodes that verify correctly and punishing nodes that verify incorrectly. These mechanisms can motivate multiple nodes to participate in the verification and reach a consensus on the verification results. For example, it is stipulated that each node participating in the verification needs to pledge a certain amount of ether. If the verification result is correct, it can get a reward; if the verification result is wrong, the pledged ether will be deducted. Use the blockchain browser to query and trace the verification data, such as querying the verification record of "Flame Sword" in the past month to obtain the historical verification information of digital assets. This information includes the timestamp of each verification, the verification hash value, the security rating results, etc. By comparing the historical verification hash values, for example, comparing the current hash value "e4b58eb4642304a2270307021b3f42f3" with its previous hash value stored in the blockchain, the authenticity of the current verification data can be verified. If the hash values do not match, it means that the data may have been tampered with. Synchronize the verification data in the blockchain, such as the historical verification records of "Flame Sword", to the local database.Use SQL to perform statistical analysis on verification data, such as calculating the average risk score and security rating change trend of "Flame Sword" in the past month, and evaluate the historical security status of digital assets. Generate a security analysis report based on the statistical analysis results. For example, the security analysis report of "Flame Sword" can include its security rating change trend chart, average risk score, and possible security risks in the past month. This information can help users understand the historical security status of digital assets and make corresponding decisions. For example, if the risk score of "Flame Sword" continues to rise, users may need to take measures to improve its security.
[0035] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A digital asset security verification and information monitoring method, characterized in that: The method comprises: According to the type attributes of digital assets, a multi-dimensional classification model is used to classify digital assets, obtain the classification labels and feature vector representations of digital assets, and establish a digital asset feature database; for the classified digital asset feature database, an adaptive verification algorithm is designed to achieve comprehensive verification of the legitimacy and ownership of digital assets by dynamically selecting matching verification rules and threshold parameters, and generate digital asset verification results; based on the digital asset verification results and the correlation between digital assets, a digital asset association graph is constructed, and a graph neural network algorithm is used to jointly model the associated digital assets, and the verification weights of digital asset nodes are calculated through message passing and feature aggregation; the transaction history data of digital assets is obtained from the digital asset association graph, and the association rules and abnormal patterns of digital asset transactions are discovered using time series pattern mining technology to form a transaction verification rule library; the transaction verification rule library is applied to digital assets. In the digital asset verification process, a multi-algorithm adaptation layer is designed in combination with the encryption algorithm and consensus mechanism of digital assets. Through algorithm registration and dynamic scheduling, compatible processing of different types of digital assets is achieved. According to the data structure of the algorithm adaptation layer, intelligent data parsing and conversion technology is used to map heterogeneous digital asset data to a unified feature space, generate standardized digital asset data, and form a digital asset feature database. The digital asset feature database is used to conduct in-depth verification and analysis to obtain the security attribute information of digital assets. A multi-factor risk assessment model is used to dynamically evaluate the security level of digital assets and generate digital asset security rating results. Based on the digital asset security rating results and verification results, a blockchain verification and proof mechanism is constructed. By generating a hash value containing the verification process, results and timestamp, a hash proof chain for digital asset verification is formed to ensure the tamper-proof and traceability of the verification results.
2. The method according to claim 1, characterized in that The digital assets are classified using a multi-dimensional classification model according to the type attributes of the digital assets, the classification labels and feature vector representations of the digital assets are obtained, and a digital asset feature database is established, including: Obtain multi-dimensional attribute information of digital assets, cluster digital assets using K-means clustering algorithm, and obtain coarse-grained classification labels of digital assets; For each coarse-grained classification, extract the common features of this type of digital assets and build a TF-IDF feature vector representation model for this type of digital assets; The support vector machine (SVM) classification algorithm is used to classify the feature vectors of digital assets to obtain fine-grained classification labels of digital assets; The classification labels and corresponding feature vectors of digital assets are stored in the MongoDB database, and a mapping relationship between digital asset labels and features is established; When receiving a digital asset search request from a user, obtaining the digital asset classification label input by the user, and obtaining the digital asset feature vector corresponding to the digital asset classification label from the MongoDB database; Calculate the cosine similarity between the user's portrait feature vector and the digital asset feature vector, and recommend the top-N digital assets with the highest similarity to the user; When there are new digital assets, extract the multi-dimensional attribute information of the new digital assets, obtain their feature vectors through the TF-IDF feature vector representation model, use the SVM classification model algorithm to determine their categories, and insert the new digital asset classification labels and feature vector information into the MongoDB database; Regularly maintain and update the digital asset feature data in the MongoDB database, and update its classification labels and feature vector representations based on the dynamic changes of digital assets. For digital assets that have not changed for a long time, reduce their update frequency, and for digital assets that change frequently, increase their update frequency.
3. The method according to claim 2, characterized in that The adaptive verification algorithm is designed for the classified digital asset feature database, and the digital asset legitimacy and ownership are comprehensively verified by dynamically selecting matching verification rules and threshold parameters, and the digital asset verification results are generated, including: Obtaining a digital asset feature database, which contains feature information of various digital assets; According to the coarse-grained classification labels of the classification information of the digital assets, a preset candidate verification rule set and a threshold parameter range are obtained; A decision tree algorithm is used to select the most suitable verification rule subset for the digital asset to be verified from the candidate verification rule set. The feature selection of the decision tree is based on the feature information of the digital asset. By calculating the information gain ratio of each feature, the feature with the largest information gain ratio is selected as the split node, and the decision tree is recursively constructed until the preset stop condition is met; The random forest algorithm is used to dynamically adjust the threshold parameters of the validation rule subset. The random forest constructs multiple decision trees by randomly selecting sample subsets and feature subsets, and then integrates the prediction results of multiple decision trees to obtain the adjustment value of the threshold parameter. Adopt the adjusted threshold parameters to determine whether the digital asset to be verified meets the preset ownership verification rules. If so, the legitimacy of the digital asset to be verified is determined and verified; The comprehensive verification score of the digital asset to be verified is calculated using a weighted average method, and the weight coefficient of the weighted average is preset according to the importance of ownership verification and legality verification; Compare the comprehensive verification score with the preset threshold to obtain the final verification result of the digital asset to be verified; The final verification results, the characteristics of the digital assets to be verified, the ownership verification results and the legality verification results are written into the digital asset verification result database.
4. The method according to claim 1, characterized in that: Based on the digital asset verification results and the correlation between digital assets, a digital asset correlation graph is constructed, and a graph neural network algorithm is used to jointly model the associated digital assets. Through message transmission and feature aggregation, the verification weight of the digital asset node is calculated, including: Obtaining the digital asset verification results and the correlation between digital assets, and constructing a digital asset correlation graph, wherein nodes in the digital asset correlation graph represent digital assets, and edges represent correlation relationships between digital assets; Preprocess the digital asset association graph, fill in missing values, filter outliers, and extract the attribute features and structural features of nodes; The graph convolutional neural network algorithm is used to model the preprocessed digital asset association graph to obtain a graph convolutional neural network model; In each layer of the graph convolutional neural network model, the feature representation of the nodes is updated through the message passing mechanism according to the adjacency relationship between the nodes; In the last layer of the graph convolutional neural network model, the feature representation of the digital asset node is obtained; The relevance weights between nodes are calculated through the attention mechanism, which includes converting the feature representation of the node into attention weights and aggregating the features of adjacent nodes by weighted summation; The feature representation and attention weight of the digital asset node are concatenated as the input of the fully connected layer, and the verification weight of each digital asset node is obtained through the fully connected layer; The importance of digital asset nodes is judged according to the verification weight, and digital asset nodes above the threshold are marked as important nodes, and digital asset nodes below the threshold are marked as suspicious nodes; Focus on verifying important nodes and suspicious nodes, and add the verification results as new node attributes to the digital asset association map; The feature representation and verification weight of the digital asset node are dynamically updated according to the verification results, and the graph convolutional neural network model is retrained to obtain continuously optimized digital asset verification results.
5. The method according to claim 1, characterized in that: The digital asset transaction history data is obtained from the digital asset association graph, and the association rules and abnormal patterns of digital asset transactions are discovered using time series pattern mining technology to form a transaction verification rule base, including: Obtain the digital asset association map, and obtain the target digital asset node and its attribute information from the digital asset association map; According to the preset association relationship, obtain the associated digital assets associated with the target digital assets; Obtain the transaction history data of the target digital asset and the associated digital assets, including transaction time, transaction amount and transaction counterparty; Preprocess the transaction history data and fill in the missing values using the preset data filling rules; Align the transaction time of the target digital asset and the associated digital asset to generate a standardized time-series transaction data set; Apriori algorithm is used to mine frequent trading patterns from time series trading data sets; Compare the frequent transaction pattern with the preset normal transaction pattern rules to determine whether the frequent transaction pattern is an abnormal transaction pattern; If the frequent trading pattern does not conform to the normal trading pattern rules, the frequent trading pattern is marked as an abnormal trading pattern; Manually review abnormal transaction patterns and add manually confirmed abnormal transaction patterns to the pre-built abnormal transaction pattern library; Acquire new digital asset transaction data, and acquire digital asset information associated with the new digital asset transaction data based on the digital asset association graph; Match new digital asset transaction data with abnormal transaction patterns in the abnormal transaction pattern library; If the new digital asset transaction data matches the abnormal transaction pattern in the abnormal transaction pattern library, an abnormal transaction alarm message is generated.
6. The method according to claim 1, characterized in that The transaction verification rule base is applied to the digital asset verification process, combined with the encryption algorithm and consensus mechanism of digital assets, and a multi-algorithm adaptation layer is designed to achieve compatible processing of different types of digital assets through algorithm registration and dynamic scheduling, including: According to the digital asset type, obtain the corresponding transaction verification rules from the digital asset registration center; Use smart contract technology to convert transaction verification rules into executable code logic and deploy it to the blockchain network; Determine the consensus mechanism adopted by the blockchain network to which the digital asset belongs; When receiving a digital asset transaction request, the smart contract is called to verify the digital asset transaction request, including checking the legitimacy of the format, signature, and attributes of the digital asset transaction request; If the digital asset transaction request is verified, the digital asset transaction request will be broadcast to the consensus node in the blockchain network; Consensus nodes verify and confirm digital asset transaction requests based on the consensus mechanism; Cryptography technology is used to protect the transaction verification process, verifying the legitimacy of digital asset transaction requests without disclosing the identity information of both parties to the transaction; For cross-chain digital asset transactions, a cross-chain protocol is introduced to achieve atomic exchange of digital assets in different blockchain networks through transaction proof mechanism and digital asset mapping mechanism; Blockchain nodes use P2P networks to propagate, synchronize, and verify blocks and transactions, ensuring the decentralization and fault tolerance of the network. According to the business rules and processes of digital assets, smart contract technology is used to solidify the business rules and processes into the code to achieve automated management of the entire life cycle of digital assets; After security audit and testing, smart contracts are deployed to the blockchain network and jointly executed and verified by multiple nodes.
7. The method according to claim 1, characterized in that According to the data structure of the algorithm adaptation layer, intelligent data parsing and conversion technology is used to map heterogeneous digital asset data to a unified feature space, generate standardized digital asset data, and form a digital asset feature database, including: Obtaining the original data of heterogeneous digital assets, including the attributes, parameters and configuration information of digital assets; Use XML parser to parse the original data, extract the key attribute fields of digital assets, and construct the digital asset attribute feature vector; According to the pre-defined unified feature space, the digital asset attribute feature vector is normalized, and the attribute values of different scales and types are mapped to the interval [0, 1] to obtain the standardized digital asset attribute feature vector; One-Hot encoding data conversion method is used to convert digital asset attributes that cannot be directly mapped to a unified feature space into a feature representation of the unified feature space; Fill the standardized digital asset attribute feature vector into the corresponding dimension of the unified feature space to obtain the feature vector representation of the digital asset in the unified feature space; Generate standardized digital asset data records based on the feature vectors of digital assets in a unified feature space and in accordance with predefined data formats and field names; Import standardized digital asset data records into the database and create a digital asset feature data table.
8. The method according to claim 1, characterized in that: The digital asset feature database is used to conduct in-depth verification and analysis, obtain security attribute information of digital assets, adopt a multi-factor risk assessment model, dynamically assess the security level of digital assets, and generate digital asset security rating results, including: Acquire digital asset feature information in a digital asset feature database to obtain a multi-dimensional feature vector of the digital asset to be evaluated; The digital asset feature vector is used to construct a training data set, and the training data set is trained through a deep neural network model to obtain a deep representation of the digital asset features; Inputting the deep representation of digital asset features into a pre-built discriminant model, and obtaining security attribute information of the digital asset through reasoning of the security attribute discriminant model, wherein the security attribute information includes confidentiality, integrity, and availability levels; Based on the acquired digital asset security attribute information, a pairwise comparison matrix is constructed using the hierarchical analysis method, the eigenvector corresponding to the maximum eigenvalue of the pairwise comparison matrix is calculated, and normalized to obtain the weight of each evaluation factor; The security attribute information of the digital asset is weighted and summed with the evaluation factor weights to obtain the risk score of the digital asset in each evaluation dimension; The system's scheduled task module is used to regularly and dynamically track the risk scores of digital assets. When the score exceeds the preset threshold, the digital asset security level adjustment process is triggered and the digital asset security rating database is updated; The real-time security rating results of digital assets are read from the digital asset security rating database, and a digital asset security assessment report is automatically generated. The report includes basic information of digital assets, security attribute level, risk score and security rating result change record.
9. The method according to claim 1, characterized in that: Based on the digital asset security rating results and verification results, a blockchain verification and certification mechanism is constructed to form a hash certification chain for digital asset verification by generating a hash value containing the verification process, results and timestamp, thereby ensuring the tamper-proof and traceability of the verification results, including: Obtain the security rating and verification results of digital assets and use them as the original data for verification proof; For the original verification data, add the verification process description and current timestamp to generate complete verification proof data; Perform SHA256 hash operation on the verification proof data to obtain a unique hash value as the fingerprint of this verification; Concatenate the current verification hash value with the previous verification hash value to build a hash proof chain to form a continuous verification sequence; The verification hash value and hash proof chain are written to the blockchain through the Ethereum smart contract; Defining verification rules and reward and punishment mechanisms through smart contracts encourages multiple nodes to participate in verification and reach consensus on the verification results; Use blockchain browsers to query and trace verification data and obtain historical verification information of digital assets; Verify the authenticity of the current verification data by comparing the historical verification hash values; Synchronize the verification data in the blockchain to the local database, use SQL to perform statistical analysis on the verification data, calculate the security indicators of digital assets, evaluate the historical security status of digital assets, and generate a security analysis report.
10. A digital asset security verification and information monitoring system, characterized in that: The system includes: a digital asset classification module for multi-dimensional classification of digital assets; a digital asset verification module for verifying the legitimacy and ownership of digital assets; a digital asset association modeling module for building a digital asset association map and weight calculation; a transaction verification rule mining module for discovering association rules and anomalies of digital asset transactions; a multi-algorithm adaptation processing module for compatible processing of different types of digital assets; and a digital asset security assessment module for conducting security level assessment of digital assets.
Citation Information
Patent Citations
Industrial chain risk assessment model and method based on graph neural network, and medium
CN117236698A
Risk identification method based on knowledge graph
CN117520565A
Power grid health assessment and analysis method based on multiple modes
CN118657404A
Systems and methods for building a unified asset graph
US20230281468A1
System and method for identity-based fraud detection
US7458508B1
Cited By
Bank abnormal transaction detection method based on multi-modal data fusion
CN119850218A
A Bank Abnormal Transaction Detection Method Based on Multimodal Data Fusion
CN119850218B
Data interaction sharing method and system suitable for digital asset management
CN120596282A
Consensus method and system based on data assets
CN120631983A
A consensus method and system based on data assets
CN120631983B
