US 12,081,684 B2
System and method for fact verification using blockchain and machine learning technologies
Helene Khaykovich, Tel Aviv (IL); Shravan Kumar Parunandula, Hyderabad (IN); Pranav Deshpande, Nagpur (IN); and Jacob Mendel, Givat Brenner (IL)
Assigned to JPMORGAN CHASE BANK, N.A., New York, NY (US)
Filed by JPMorgan Chase Bank, N.A., New York, NY (US)
Filed on Jun. 30, 2022, as Appl. No. 17/810,014.
Claims priority of application No. 202211028699 (IN), filed on May 18, 2022.
Prior Publication US 2023/0379180 A1, Nov. 23, 2023
Int. Cl. H04L 9/00 (2022.01); H04L 9/32 (2006.01)
CPC H04L 9/50 (2022.05) [H04L 9/3218 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for performing fact verification, the method comprising:
performing, using a processor and a memory:
receiving a document for verification on a computer network;
storing the document for verification on a private Inter Planetary File System (IPFS) network;
performing, on the computer network, a hashing operation on the document and generating a document hash, the document hash including a document identifier (ID);
transmitting, from the computer network to a blockchain, the document ID, the blockchain being facilitated by a plurality of computers;
receiving the document ID from the blockchain and retrieving the document for verification corresponding to the document ID from the IPFS network;
identifying and extracting, on the computer network, at least two named entities from the document for verification and associated information as metadata;
sending, to the blockchain, the document ID along with a metadata ID corresponding to the extracted metadata;
identifying, off the blockchain, identifiers of relevant documents corresponding to each of the at least two named entities, the relevant documents being authenticated documents including one or more of the at least two named entities;
identifying a predetermined number of relevant documents among the identified relevant documents based on a number of the at least two named entities present and a number of occurrences for each of the named entities;
transmitting the document ID along with the identifiers of the predetermined number of relevant documents to the computer network;
performing verification of the document by executing one or more machine learning algorithms to determine whether the document to be verified is supported by the predetermined number of relevant documents; and
providing and storing a verdict for the document to be verified with supporting evidence from the predetermined number of relevant documents.