A method and system for identifying cross-chain transaction associations in blockchain.

Abstract

This invention discloses a method and system for identifying cross-chain transaction associations in blockchain technology. To address the technical problems of low coverage of private cross-chain bridges and inaccurate cross-chain transaction matching in existing cross-chain association methods, this invention collects raw data from multiple blockchain sources and extracts transaction and price information; identifies cross-chain bridge entities based on object-time-amount consistency matching of price information; performs collision matching on the transaction fields involved in the entity to extract the cross-chain association data structure; and determines the identification result by associating and pairing token transaction information based on this structure. This invention can automatically mine public and private cross-chain bridge entities and perform high-precision association analysis on a wide range of cross-chain behaviors, including non-financial data and inconsistent sender-receiver relationships.

Description

Technical Field

[0001] This invention belongs to the field of blockchain technology, specifically relating to a method and system for identifying cross-chain transaction associations in blockchain. Background Technology

[0002] With the rapid development and widespread application of blockchain technology, diverse needs have led to the emergence of multiple types of blockchains, forming a multi-chain coexistence ecosystem. Centralized Finance (DeFi), as one of the mainstream applications of blockchain, suffers from liquidity fragmentation and interaction difficulties due to the existence of multiple chains. To break the "island effect" between different blockchains, cross-chain bridge technology has emerged. It allows users to transfer assets or information from one source chain to another target chain, becoming a crucial facility for transferring assets between different chains. Cross-chain bridges are typically implemented using a combination of on-chain contracts and off-chain verification mechanisms, allowing users to submit assets on the source chain and receive the corresponding assets or tokens on the target chain. When a user initiates a cross-chain transaction on an on-chain contract, the on-chain contract generates a series of runtime logs to record the user's cross-chain behavior. After the off-chain relayer detects the contract log of the user's request on the source chain, it calls the target chain contract. After receiving the relayer's message, the target chain contract executes the relevant cross-chain operations and records the logs after identity verification or cryptographic verification (see: Ou W, Huang S, Zheng J, et al. An overview on cross-chain: Mechanism, platforms, challenges and advances[J]. Computer Networks, 2022, 218: 109378.). Since cross-chain operations do not require real-name verification and often involve inter-chain delays and multiple corresponding transfers, cross-chain bridges are frequently used in illegal activities such as money laundering and hiding fund paths. Therefore, the automatic identification and association of cross-chain transactions and related entities has important judicial and security value.

[0003] Existing fund flow analysis systems can handle some cross-chain transaction analysis, but they still have two shortcomings when facing complex DeFi operations and multi-chain environments, making it difficult to accurately track cross-chain events: First, relying on manual labeling or external libraries is difficult to cover small-scale or private cross-chain bridges; second, they usually only target the amount / time characteristics of tokens or use only fixed retrieval methods to match cross-chain transactions, which is not robust enough when dealing with cross-chain behavior of small-scale cross-chain bridges or non-financial message payloads (see: Lin D, Wu J, Su Y, et al. Connector: Enhancing thetraceability of decentralized bridge applications via automatic cross-chain transaction association[J]. IEEE Transactions on Information Forensics and Security, 2025.). Therefore, there is an urgent need for a method based on on-chain "object-time-amount" consistency and combined with transaction fields (such as log topic, call data, event data) for heuristic matching to improve the recognition rate of cross-chain bridges and cross-chain token pairs, and to detect and distinguish non-financial cross-chain message transmissions. This invention combines cross-chain bridge and token pair detection based on object-time-amount consistency with cross-chain transaction association method based on graph maximum weight matching to further efficiently and accurately detect and analyze cross-chain entities and transactions. Summary of the Invention

[0004] The purpose of this invention is to solve the technical problems of low coverage of private cross-chain bridges and inaccurate cross-chain transaction matching in existing cross-chain associations. It proposes a blockchain cross-chain transaction association identification method and system, which extracts data structure through ternary consistency matching and field collision matching, and uses weighted graphs to achieve automatic detection and high-precision association of various cross-chain transactions and entities.

[0005] To achieve the above objectives, the present invention adopts the following technical solution.

[0006] A method for identifying cross-chain transaction associations in blockchain includes the following steps: Collect raw data from multiple blockchain sources and extract token transaction information and token price information; Based on the token price information, the token transaction information is matched for consistency to identify the cross-chain bridge entity; Collision matching is performed on the transaction fields involved in the cross-chain bridge entity to extract the cross-chain related data structure; Based on the cross-chain association data structure, the token transaction information is associated and matched to obtain the cross-chain transaction association identification result.

[0007] Furthermore, collect raw blockchain data from multiple sources, including: Collect block data, transaction data, event logs, and status change data through the blockchain interface; The block data includes the block number, timestamp, block hash, and block transaction list; The transaction data includes the transaction hash, sending address, receiving address, transfer amount, and input data; The event log includes the contract address, event signature hash, topic list, and event data; The status change data includes changes in address balance and changes in contract storage slots.

[0008] Furthermore, extract token transaction information, including: For the native tokens in the raw data, token transfer records are extracted by analyzing the transaction execution flow and removing invalid transactions; For non-native tokens in the raw data, extract transfer-related events from the event log to obtain token transfer information; Generate token transaction information in a unified format based on token transfer records and information.

[0009] Furthermore, extract token price information, including: Identify a decentralized exchange entity with liquidity on the corresponding blockchain; Send a price inquiry request at a specified block height to the decentralized exchange entity to obtain token price information.

[0010] Furthermore, based on the token price information, consistency matching of the token transaction information is performed to identify the cross-chain bridge entity, including: Scan target transactions on the target chain in chronological order, and simultaneously scan source chain transactions based on a time sliding window; Based on the token price information, calculate the amount error between the source chain transaction and the target transaction within the time sliding window; When the amount error is within the preset tolerance range and the transaction object address field is consistent, a matching pair between the source transaction and the target transaction is established; The frequency of occurrence of address pairs and token pairs involved in the matching pairs is statistically analyzed, and addresses with a frequency higher than a preset threshold are identified as cross-chain bridge entities.

[0011] Furthermore, the cross-chain related data structure is extracted, including: Obtain all event logs of source chain transactions and target chain transactions associated with the cross-chain bridge entity; Excluding preset event types in the event log that are unrelated to cross-chain messages, a set of events to be matched is obtained; Extract the event fields from the set of events to be matched, as well as the block data fields, transaction data fields, and contract call fields involved in the cross-chain bridge entity; Perform string matching on the extracted dimensional fields and record the field key names and their frequency of occurrence for each successful match; Select field key names that appear more frequently than a preset threshold to construct a cross-chain related data structure.

[0012] Furthermore, the token transaction information is associated and matched, including: Based on the cross-chain association data structure, extract the transactions to be associated with the cross-chain bridge entity; A weighted graph is constructed based on the degree of association of the transactions to be associated, where the transactions are graph nodes and the field matching scores are edge weights. The weighted graph is subjected to maximum matching to determine the cross-chain transaction association identification result.

[0013] Furthermore, a weighted graph is constructed based on the degree of association of the transactions to be related, including: Scan the transactions to be associated on the target chain in chronological order and maintain a time window queue of transactions from the source chain; The fields of the source transaction and the target transaction in the time window queue are matched, and the proportion of the matched fields in the cross-chain related data structure is calculated to obtain the correlation score. The correlation scores are used as weights to connect the corresponding transaction nodes, thus constructing a weighted graph.

[0014] Further, the weighted graph is subjected to maximum matching processing, including: Based on a greedy strategy, the edges connecting unmatched nodes are selected sequentially in the weighted graph in descending order of weight. The selected transaction pairs pointed to by the connecting edges are used as the association identification relationships to generate cross-chain transaction association identification results.

[0015] A blockchain cross-chain transaction association identification system includes: The data acquisition module is used to collect raw data from multiple blockchain sources and extract token transaction information and token price information. The cross-chain entity identification module is used to perform consistency matching of the token transaction information based on the token price information to identify the cross-chain bridge entity. The association structure extraction module performs collision matching on the transaction fields involved in the cross-chain bridge entity to extract the cross-chain association data structure; The transaction association identification module performs association pairing on the token transaction information based on the cross-chain association data structure to obtain the cross-chain transaction association identification result.

[0016] The present invention has achieved the following beneficial effects.

[0017] 1. This invention adopts a sliding window and dual-pointer statistical method based on the "object-time-amount" ternary consistency, which can automatically discover public and small-scale private cross-chain bridge entities, effectively solving the problem that existing technologies rely on manual labeling and are difficult to cover non-mainstream cross-chain facilities.

[0018] 2. This invention can automatically identify and construct cross-chain related data structures by performing heuristic full-word or containment matching on multi-dimensional fields, thereby accurately handling complex cross-chain architectures where the identities of fund custodians and message relayers are inconsistent.

[0019] 3. This invention constructs a weighted graph based on the extracted associated data structure and performs maximum matching processing, transforming the complex cross-chain association problem into a graph matching problem, and achieving high-precision detection of non-financial cross-chain messages and scenarios where the sender and receiver are inconsistent.

[0020] 4. This invention improves the robustness of association in complex DeFi environments by preprocessing multi-source blockchain data and combining it with token price information for tolerance matching, and can accurately analyze the verification relationship between token cross-chain and non-financial cross-chain data. Attached Figure Description

[0021] Figure 1 This is a flowchart of a blockchain cross-chain transaction association identification method in one embodiment; Figure 2 This is a block diagram of a blockchain cross-chain transaction association identification system in an embodiment. Detailed Implementation

[0022] To make the various technical features, advantages, or effects of the present invention more apparent and understandable, detailed descriptions are provided below through embodiments.

[0023] This invention provides a method for identifying cross-chain transaction associations in blockchain, the process of which is as follows: Figure 1 As shown, the specific steps include: Step S1: Collect raw data from multiple blockchain sources and extract token transaction information and token price information.

[0024] Specifically, this step involves collecting and preprocessing multi-chain data, pre-extracting all token transaction information within the blocks to be collected. The multi-chains that can be used in this invention include EVM-compatible blockchains such as the Ethereum mainnet, Binance Smart Chain, Polygon, Arbitrum, and Optimism.

[0025] In an optional embodiment of the present invention, step S1 may include: Step S11: Collect block data, transaction data, event logs, and state change data through the blockchain interface.

[0026] Specifically, the aforementioned multi-chain data can be collected through the JSONRPC API of EVM-compatible blockchains. Block data includes block number, timestamp, block hash, and block transaction list; transaction data includes transaction hash, sending address, receiving address, transfer amount, and input data; event logs include contract address, event signature hash, topic list, and event data; and state change data includes changes in address balance and contract storage slots.

[0027] In an optional embodiment of the present invention, for detailed transaction logs that need to be queried on demand during subsequent correlation analysis, the corresponding blockchain API can be used directly.

[0028] Step S12: For the native tokens in the original data, extract the token transfer records by analyzing the transaction execution flow and removing invalid transactions.

[0029] Specifically, for native tokens, the Trace API is used to analyze the transaction execution flow, remove invalid transactions such as those that are withdrawn or have zero value, and calculate the record of valid token transfers in the transaction. This record includes the sender, the receiver, and the amount.

[0030] Step S13: For non-native tokens in the original data, extract transfer-type events from the event log to obtain token transfer information.

[0031] Specifically, for non-native tokens such as ERC20 tokens, Transfer events are extracted directly from the block to obtain transfer information, including the token address, sender, receiver, and amount. Furthermore, all token type transfers require recording corresponding transaction data, including the transaction hash and block height.

[0032] Step S14: Generate token transaction information in a unified format based on token transfer records and token transfer information.

[0033] Step S15: Identify a decentralized exchange entity with liquidity on the corresponding blockchain.

[0034] Step S16: Send a price inquiry request at a specified block height to the decentralized exchange entity to obtain token price information.

[0035] Specifically, a decentralized exchange with good liquidity on the blockchain is selected, and a price inquiry request is sent to the exchange at the specified block version to obtain the price of the token under that block.

[0036] Step S2: Based on the token price information, perform consistency matching on the token transaction information to identify the cross-chain bridge entity.

[0037] Specifically, in order to extract possible cross-chain bridge entities from the original transaction data, this invention takes advantage of the fact that cross-chain transactions are mostly issued and received by the same user, and adopts a sliding window and two-pointer statistical method based on the consistency matching of object, time and amount to extract the intermediary entity in the cross-chain behavior, namely the cross-chain bridge.

[0038] In an optional embodiment of the present invention, step S2 may include: Step S21: Scan the target transactions on the target chain in chronological order, and simultaneously scan the source chain transactions based on a time sliding window.

[0039] Specifically, while scanning the target chain, the source chain's transactions are scanned simultaneously. A sliding window mechanism is used, maintaining a dual-pointer representation of the window's left and right endpoints, pointing to a list of source chain transaction information. The left pointer indicates the earliest possible transaction for the associated transaction, and the right pointer indicates the latest possible transaction. This pointer is dynamically updated as the target chain scanning progresses. During the scan, the transaction amounts for both the source and target chains need to be calculated based on the token price and stored.

[0040] In an optional embodiment of the present invention, the right pointer moves with the current target chain transaction time and is updated to no more than the latest time of the target chain transaction, while the left pointer ensures that all transactions within the sliding window appear within the last 5 hours; for each target chain transaction, possible matching source chain transactions are searched within the window range.

[0041] Step S22: Based on the token price information, calculate the amount error between the source chain transaction and the target transaction within the time sliding window.

[0042] Specifically, the queue searches for source chain transfers whose amounts are within a given tolerance range and whose transaction counterparties have identical fields. Here, the transaction counterparty refers to the address sending or receiving the tokens.

[0043] In an optional embodiment of the invention, the monetary error includes a relative error or an absolute error; for example, the tolerance range is a relative error of no more than 1% or an absolute error of no more than the equivalent of US$1 to US$10.

[0044] Step S23: When the amount error is within the preset tolerance range and the transaction object address field is consistent, establish a matching pair between the source transaction and the target transaction.

[0045] In an optional embodiment of the present invention, the sending and receiving address pairs and token pairs of the matched source and target transfer pairs are counted. When multiple candidates are encountered, the count of each candidate is normalized by 1 / n to avoid noise amplification. Here, n represents the number of source chain candidate transactions matched by the target chain transaction.

[0046] Step S24: Calculate the frequency of occurrence of address pairs and token pairs involved in the matching pair, and identify addresses with a frequency higher than a preset threshold as cross-chain bridge entities.

[0047] Specifically, the most frequently occurring transfer paths are statistically analyzed, i.e., the tuples consisting of possible cross-chain bridge entity addresses and transfer tokens, and then filtered by a quantity threshold to identify cross-chain bridge addresses and supported cross-chain token pairs.

[0048] In an optional embodiment of the invention, the quantity threshold is set to a count greater than or equal to 10. This method can detect public and private cross-chain bridges.

[0049] Step S3: Perform collision matching on the transaction fields involved in the cross-chain bridge entity and extract the cross-chain related data structure.

[0050] Specifically, this step can associate common user token cross-chain transactions and heuristically extract cross-chain association data structures from them. For example, it can discover the event types and corresponding event parameters of the cross-chain behavior in the transaction log, which can be used to match non-transfer cross-chain transactions in the future.

[0051] In an optional embodiment of the present invention, step S3 may include: Step S31: Obtain all event logs of source chain transactions and target chain transactions associated with the cross-chain bridge entity.

[0052] Specifically, for all associated transactions of a certain cross-chain bridge entity, obtain all event logs of all associated cross-chain transactions on the source chain and the target chain.

[0053] Step S32: Exclude preset event types in the event log that are unrelated to cross-chain messages to obtain a set of events to be matched.

[0054] Specifically, by manually searching the existing event database, common event types unrelated to cross-chain messages, such as Transfer, Approval, and Swap events, are excluded. While these events are fewer in number, they are numerous, and this method effectively reduces a large number of interfering events.

[0055] Step S33: Extract the event fields from the set of events to be matched, as well as the block data fields, transaction data fields, and contract call fields involved in the cross-chain bridge entity.

[0056] Specifically, all event logs of transactions on the source and target chains are collected, including the event initiator, topic, and data fields; simultaneously, information fields from block data, transaction data, and contract calls are collected, such as transaction hash, blockchain number, contract caller, contract calldata, and data hash value. These data entries typically do not exceed 100.

[0057] Step S34: Perform string matching on the extracted fields of each dimension, and record the field key names and their frequency of occurrence for the successfully matched fields.

[0058] Specifically, the collected source chain and target chain data entries are enumerated, and the information fields are matched with full words, i.e., the field values ​​are exactly the same; or the string inclusion is matched, i.e., a substring of one field value contains another field value.

[0059] In an optional embodiment of the present invention, for a successfully matched field, the key names of the corresponding fields of the source chain and the target chain are recorded, such as information type, event initiator, topic position, and included in the statistical data corresponding to the cross-chain entity.

[0060] Step S35: Select field key names that appear more frequently than a preset threshold and construct a cross-chain related data structure.

[0061] Specifically, after completing all matches for the cross-chain entity, statistical data is analyzed. Fields appearing more frequently than a certain threshold in the statistical results are selected, representing data structures that are likely to represent key information about the cross-chain association. The address field is the address of the cross-chain bridge entity.

[0062] In an optional embodiment of the present invention, a preset threshold is set at 70% or more of the number of occurrences of the cross-chain bridge transaction, to avoid the impact of cross-chain bridge management transactions. In this method, the event initiator is also used as a cross-chain data matching field to address the inconsistency between the identities of the fund custodian and the message relayer in complex cross-chain bridge architectures, thereby detecting non-financial cross-chain requests.

[0063] Step S4: Based on the cross-chain associated data structure, the token transaction information is associated and matched to obtain the cross-chain transaction association identification result.

[0064] Specifically, this step can combine cross-chain associated data structures to check and match a wider range of cross-chain transactions, such as non-financial data and transactions where the sender and receiver are inconsistent. This invention is primarily based on Ethereum and EVM-compatible blockchains, but can also be extended to other blockchain architectures.

[0065] In an optional embodiment of the present invention, step S4 may include: Step S41: Based on the cross-chain associated data structure, extract the transactions to be associated with the cross-chain bridge entity.

[0066] Specifically, based on the cross-chain associated data structure obtained in the above steps, all transactions related to the cross-chain bridge address are re-extracted.

[0067] Step S42: Construct a weighted graph based on the degree of association of the transactions to be associated, where the transactions are graph nodes and the field matching scores are edge weights.

[0068] In an optional embodiment of the present invention, step S42 may include: Step S421: Scan the transactions to be associated on the target chain in chronological order and maintain a time window queue of transactions from the source chain.

[0069] Specifically, all transactions on the target chain are scanned in chronological order; a queue of source chain transactions is maintained for each target chain transfer, representing a time window of transactions, such as the most recent 5 hours. All source chain and target chain transactions are treated as graph nodes.

[0070] Step S422: Match the fields of the source transaction and the target transaction in the time window queue, calculate the proportion of the matched fields in the cross-chain related data structure, and obtain the correlation score.

[0071] Specifically, all transactions in the queue are scanned, and the relevant fields of the target transaction and the source transaction are matched based on the association information stored in the association data structure library.

[0072] Step S423: Connect the corresponding transaction nodes with the correlation score as the weight to construct a weighted graph.

[0073] Specifically, a subset of transaction pairs with higher scores are retained, for example, only those with scores above a certain threshold and slightly lower than the highest score. These scores are used as edge weights, and the two transactions are added to the weighted graph with a connected edge relationship.

[0074] Step S43: Perform maximum matching on the weighted graph to determine the cross-chain transaction association identification result.

[0075] Specifically, for multi-chain scenarios, the graph is a general graph and can be matched using methods such as the Blossom Algorithm.

[0076] In an optional embodiment of the present invention, step S43 may include: Step S431: Based on a greedy strategy, select the edges connecting unmatched nodes in the weighted graph in descending order of weight.

[0077] Specifically, since the cross-chain entity discovery module has extracted highly accurate association fields, and most points in the graph have a unique maximum matching scheme, a greedy strategy is adopted to continuously select the highest weight edge of the remaining unmatched points in the graph as the association relationship.

[0078] Step S432: The selected transaction pair pointed to by the edge is used as the association identification relationship to generate the cross-chain transaction association identification result.

[0079] Specifically, the above method can accurately analyze matching relationships, including non-financial data and inconsistencies between sender and receiver, and can also be used for the verification of cross-chain data of general tokens.

[0080] This invention also provides a blockchain cross-chain transaction association identification system, such as... Figure 2 As shown, it includes: The data acquisition module is used to collect raw data from multiple blockchain sources and extract token transaction information and token price information. The cross-chain entity identification module is used to perform consistency matching of token transaction information based on token price information and identify cross-chain bridge entities. The association structure extraction module performs collision matching on the transaction fields involved in the cross-chain bridge entity to extract the cross-chain association data structure; The transaction association identification module associates and matches token transaction information based on the cross-chain association data structure to obtain the cross-chain transaction association identification result.

[0081] Method performance testing: This invention was applied to the transaction history of the Ethereum mainnet and Polygon network in November 2023. Due to limitations of the blockchain API network, step S1 takes approximately 8 hours to complete. In a serial execution, steps S2 to S4 take approximately 1 hour to complete. The system identified a total of 95 cross-chain bridge related addresses. Using a public address tag library for comparison, 77% of the addresses were identified as cross-chain bridges with existing public tags, while the remaining 23% were untagged addresses. There were no known false positives indicating that the addresses were not cross-chain bridges. The system successfully identified the top ten cross-chain bridges operating on these two networks. Through manual verification, the identified cross-chain transaction structures were found to be largely consistent with the actual situation, and two cross-chain bridge entities supporting non-financial message cross-chain transactions were successfully detected.

[0082] Although the present invention has been disclosed above with reference to embodiments, it is not intended to limit the present invention. Appropriate modifications or equivalent substitutions made by those skilled in the art to the technical solutions of the present invention should be covered within the protection scope of the present invention, which is defined by the claims.

Claims

1. A method for identifying cross-chain transaction association of a blockchain, characterized in that, Includes the following steps: Collect raw data from multiple blockchain sources and extract token transaction information and token price information; Based on the token price information, the token transaction information is matched for consistency to identify the cross-chain bridge entity; Collision matching is performed on the transaction fields involved in the cross-chain bridge entity to extract the cross-chain related data structure; Based on the cross-chain association data structure, the token transaction information is associated and matched to obtain the cross-chain transaction association identification result.

2. The method of claim 1, wherein, Collect raw blockchain data from multiple sources, including: Collect block data, transaction data, event logs, and status change data through the blockchain interface; The block data includes the block number, timestamp, block hash, and block transaction list; The transaction data includes the transaction hash, sending address, receiving address, transfer amount, and input data; The event log includes the contract address, event signature hash, topic list, and event data; The status change data includes changes in address balance and changes in contract storage slots.

3. The method of claim 1, wherein, Extract token transaction information, including: For the native tokens in the raw data, token transfer records are extracted by analyzing the transaction execution flow and removing invalid transactions; For non-native tokens in the raw data, extract transfer-related events from the event log to obtain token transfer information; Generate token transaction information in a unified format based on token transfer records and information.

4. The method as described in claim 1, characterized in that, Extract token price information, including: Identify a decentralized exchange entity with liquidity on the corresponding blockchain; Send a price inquiry request at a specified block height to the decentralized exchange entity to obtain token price information.

5. The method as described in claim 1, characterized in that, Based on the token price information, consistency matching of the token transaction information is performed to identify cross-chain bridge entities, including: Scan target transactions on the target chain in chronological order, and simultaneously scan source chain transactions based on a time sliding window; Based on the token price information, calculate the amount error between the source chain transaction and the target transaction within the time sliding window; When the amount error is within the preset tolerance range and the transaction object address field is consistent, a matching pair between the source transaction and the target transaction is established; The frequency of occurrence of address pairs and token pairs involved in the matching pairs is statistically analyzed, and addresses with a frequency higher than a preset threshold are identified as cross-chain bridge entities.

6. The method as described in claim 1, characterized in that, Extract the cross-chain related data structure, including: Obtain all event logs of source chain transactions and target chain transactions associated with the cross-chain bridge entity; Excluding preset event types in the event log that are unrelated to cross-chain messages, a set of events to be matched is obtained; Extract the event fields from the set of events to be matched, as well as the block data fields, transaction data fields, and contract call fields involved in the cross-chain bridge entity; Perform string matching on the extracted dimensional fields and record the field key names and their frequency of occurrence for each successful match; Select field key names that appear more frequently than a preset threshold to construct a cross-chain related data structure.

7. The method as described in claim 1, characterized in that, Associating and matching the token transaction information includes: Based on the cross-chain association data structure, extract the transactions to be associated with the cross-chain bridge entity; A weighted graph is constructed based on the degree of association of the transactions to be associated, where the transactions are graph nodes and the field matching scores are edge weights. The weighted graph is subjected to maximum matching to determine the cross-chain transaction association identification result.

8. The method as described in claim 7, characterized in that, Construct a weighted graph based on the degree of association of the transactions to be related, including: Scan the transactions to be associated on the target chain in chronological order and maintain a time window queue of transactions from the source chain; The fields of the source transaction and the target transaction in the time window queue are matched, and the proportion of the matched fields in the cross-chain related data structure is calculated to obtain the correlation score. The correlation scores are used as weights to connect the corresponding transaction nodes, thus constructing a weighted graph.

9. The method as described in claim 7, characterized in that, The maximum matching process for the weighted graph includes: Based on a greedy strategy, the edges connecting unmatched nodes are selected sequentially in the weighted graph in descending order of weight. The selected transaction pairs pointed to by the connecting edges are used as the association identification relationships to generate cross-chain transaction association identification results.

10. A blockchain cross-chain transaction association identification system, characterized in that, include: The data acquisition module is used to collect raw data from multiple blockchain sources and extract token transaction information and token price information. The cross-chain entity identification module is used to perform consistency matching of the token transaction information based on the token price information to identify the cross-chain bridge entity. The association structure extraction module performs collision matching on the transaction fields involved in the cross-chain bridge entity to extract the cross-chain association data structure; The transaction association identification module performs association pairing on the token transaction information based on the cross-chain association data structure to obtain the cross-chain transaction association identification result.

Images