Granted Patent B1

US 12,443,954 · App. 19/249,061 · Granted Oct 14, 2025

Inventors: Vijay Madisetti (Alpharetta, GA); Arshdeep Bahga (Chandigarh, IN)

Assignee: Vijay Madisetti

G06Q20/389G06Q20/381H04L9/3236H04L9/3268H04L9/50H04L2209/56

Abstract

Claims (44)

1. A method for generating risk assessment scores for blockchain addresses and smart contracts, comprising:

constructing a graph representation comprising a plurality of nodes, each node of the plurality of nodes representing one of a blockchain address or a smart contract address on a blockchain, and a plurality of edges, each edge of the plurality of edges representing a transaction between an externally-owned blockchain account on the blockchain network and a smart contract address on the blockchain network;

applying one or more pattern recognition algorithms to the graph representation to identify one or more identified sub-graphs within the graph representation comprising one or more groups of blockchain account addresses and smart contract addresses having relationship structures indicative of potential fraudulent activity;

executing one or more filtering contracts configured to generate analysis results by analyzing at least one of transaction flows, logic patterns, and behavioral characteristics of blockchain account addresses and smart contract addresses within each identified sub-graph of the one or more identified sub-graphs; and

determining a risk score for each blockchain account address and smart contract address comprised by an identified sub-graph of the one or more identified sub-graphs based on the analysis results from the filtering contracts, the risk score quantifying probabilistic fraud risk along a plurality of dimensions.

2. The method of claim 1 wherein the blockchain account addresses represented by the plurality of nodes each comprise externally owned accounts having account addresses, public-private key pairs, and balances; and

the smart contract addresses represented by the plurality of nodes each comprise contract addresses, contract code, and balances.

3. The method of claim 1 wherein applying the one or more pattern recognition algorithms comprises using one or more graph matching engines to recognize the one or more identified sub-graphs whose topology and activities resemble templates stored in a fraud pattern knowledge base, the fraud pattern knowledge base comprising codified signatures of known deceptive contract graphs and cluster profiles.

4. The method of claim 1 wherein executing the one or more filtering contracts comprises deploying one or more targeted smart contracts configured to scrutinize one or more of transactions, logic flows, and user behaviors associated with blockchain account addresses and smart contract addresses in the one or more identified sub-graphs to gather forensic evidence support packets.

5. The method of claim 4 wherein executing the one or more filtering contracts comprises deploying one or more server-side software modules on a server, the one or more server-side software modules being configured to analyze at least one of the transaction flows, logic patterns, and behavioral characteristics without deployment to the blockchain network.

6. The method of claim 1 wherein determining risk scores comprises analyzing at least one of discrepancy types, deception indicators, suspicious behaviors, regulatory violations, and fraud harms identified by the one or more filtering contracts.

7. The method of claim 6 further comprising confirming high-likelihood frauds using at least one of rule-based scoring and supervised models.

8. The method of claim 1 further comprising:

continuously monitoring the blockchain network for new smart contracts and state changes by ingesting new blocks and transactions from the blockchain network using a bulletin board messaging framework;

publishing messages containing contract details and transaction data to one or more topics managed by a bulletin board server; and

subscribing to the one or more topics to receive updates for updating the graph representation.

9. The method of claim 8 wherein continuously monitoring the blockchain network comprises:

registering one or more variables with a global variable name system;

updating one or more variable values responsive to one or more identified blockchain state changes; and

broadcasting updates to the one or more variables through the bulletin board messaging framework to subscribers of the one or more topics.

10. The method of claim 1 wherein the one or more filtering contracts are configured to perform security functions selected from the group consisting of verifying sender and receiver addresses, enforcing time limits, enforcing quantity restrictions, filtering suspicious transactions, and performing virus filter functions for filtering out suspicious transactions.

11. The method of claim 1 wherein the one or more filtering contracts are stored in and executed by one or more software modules comprised by a sever positioned in communication with the blockchain network.

12. A system for blockchain address and smart contract risk assessment, comprising:

a processor;

a network communication device positioned in communication with the processor and operable to communicate across a network with a blockchain network; and

a non-transitory computer-readable storage system having software stored thereon that, when executed by the processor, is operable to instantiate:

a graph generator module configured to continuously construct one or more relationship maps between a plurality of blockchain addresses on the blockchain network and a plurality of smart contracts deployed on the blockchain network based on transactional and informational flows by ingesting new blockchain data from the blockchain network;

a knowledge base having stored therein at least one of fraud detection patterns, schemas of known deceptive contract topologies, and regulatory policies governing transactions to guide anomaly detection;

a pattern matcher module configured to execute one or more graph pattern matching algorithms operable to recognize one or more sub-graphs comprising a subset of blockchain addresses and a subset of smart contracts within the one or more relationship maps, where the topology and transaction history of the one or more sub-graphs resembles one or more fraud templates in the knowledge base;

a filtering engine module comprising one or more targeted filtering contracts configured to generate one or more filtered results by analyze at least one of transactions, logic flows, and user behaviors associated with the subset blockchain addresses and the subset of smart contracts comprised by a sub-graph of the one or more sub-graphs to gather forensic evidence; and

a fraud classifier module configured to evaluate the one or more filtered results using at least one of rule-based scoring and supervised models to determine a fraud likelihood and severity levels for triggering responses.

13. The system of claim 12 wherein the software is further operable to, when executed by the processor, instantiate a case management module configured to track at least one of detected fraud instances, risk levels, affected assets, response actions, and evidence packets to enable investigations, data analysis, auditing and regulatory reporting requirements.

14. The system of claim 12 wherein the software is further operable to, when executed by the processor, instantiate a connector module comprising connectors for at least one of blockchain networks, bulletin board servers, and a global variable name system.

15. The system of claim 12 wherein the filtering engine module is further configured to:

gather forensic evidence comprising at least one of transactions, data consistency, logic flows, and user behaviors within the one or more sub-graphs; and

quantify a fraud risk score for each sub-graph of the one or more sub-graphs along one or more severity dimensions.

16. The system of claim 12 wherein the software is further operable to, when executed by the processor, instantiate a response trigger module configured to trigger automated on-chain actions on the blockchain network to limit detected fraud, the on-chain actions comprising blockchain transactions selected from the group consisting of asset freezes, contract blocking, and transaction reversal based on severity thresholds and configurable platform policies.

17. The system of claim 12 wherein the software is further operable to, when executed by the processor, instantiate a reporting and alerting module comprising a plurality of integration hooks operable to pass data to external systems for alerts, investigations, preventative screening, and regulatory reporting requirements.

18. A system for generating risk assessment scores for blockchain addresses and smart contracts, comprising:

means for constructing a graph representation comprising a plurality of nodes, each node of the plurality of nodes representing one of a blockchain address or a smart contract address on a blockchain, and a plurality of edges, each edge of the plurality of edges representing a transaction between an externally-owned blockchain account on a blockchain network and a smart contract address on the blockchain network;

means for applying one or more pattern recognition algorithms to the graph representation to identify one or more identified sub-graphs within the graph representation comprising one or more groups of blockchain account addresses and smart contract addresses having relationship structures indicative of potential fraudulent activity;

means for executing one or more filtering contracts configured to generate analysis results by analyzing at least one of transaction flows, logic patterns, and behavioral characteristics of blockchain account addresses and smart contract addresses within each identified sub-graph of the one or more identified sub-graphs; and

means for determining a risk score for each blockchain account address and smart contract address comprised by an identified sub-graph of the one or more identified sub-graphs based on the analysis results from the filtering contracts, the risk score quantifying probabilistic fraud risk along a plurality of dimensions.

19. The system of claim 18 wherein the means for applying the one or more pattern recognition algorithms are operable to use one or more graph matching engines to recognize the one or more identified sub-graphs whose topology and activities resemble templates stored in a fraud pattern knowledge base, the fraud pattern knowledge base comprising codified signatures of known deceptive contract graphs and cluster profiles.

Assignments (2)

Continuity (10)

References Cited (5)