View source for 20240046095. NEURAL EMBEDDINGS OF TRANSACTION DATA simplified abstract (Capital One Services, LLC)

=NEURAL EMBEDDINGS OF TRANSACTION DATA=


==Organization Name==

[[:Category:Capital One Services, LLC|Capital One Services, LLC]]


[[Category:Capital One Services, LLC]]

==Inventor(s)==

[[:Category:Christopher Bruss of Washington DC (US)|Christopher Bruss of Washington DC (US)]][[Category:Christopher Bruss of Washington DC (US)]]

[[:Category:Keegan Hines of Washington DC (US)|Keegan Hines of Washington DC (US)]][[Category:Keegan Hines of Washington DC (US)]]

==NEURAL EMBEDDINGS OF TRANSACTION DATA - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240046095 titled 'NEURAL EMBEDDINGS OF TRANSACTION DATA

==Simplified Explanation==

The patent application describes systems, methods, and computer program products for providing neural embeddings of transaction data. It involves creating a network graph of transaction data that defines relationships between transactions, merchants, and accounts. A neural network is trained using positive and negative entity pairs from training data. An embedding function encodes transaction data for a new transaction, and an embeddings layer determines a vector representation for the new transaction. The similarity between vectors for transactions is then calculated, and the new transaction is determined to be related to another transaction based on this similarity.

* Neural embeddings of transaction data are provided using a network graph and a trained neural network.
* Relationships between transactions, merchants, and accounts are defined in the network graph.
* The neural network is trained using positive and negative entity pairs from training data.
* Transaction data for a new transaction is encoded using an embedding function.
* An embeddings layer determines a vector representation for the new transaction.
* The similarity between vectors for transactions is calculated.
* Relatedness between the new transaction and another transaction is determined based on the similarity.

Potential Applications:
* Fraud detection: The technology can be used to identify potentially fraudulent transactions by analyzing the relationships between transactions, merchants, and accounts.
* Customer behavior analysis: By analyzing transaction data and determining relationships between transactions, the technology can provide insights into customer behavior and preferences.
* Recommender systems: The neural embeddings can be used to recommend relevant products or services based on transaction data and similarities between transactions.

Problems Solved:
* Complex transaction data analysis: The technology simplifies the analysis of transaction data by representing it as neural embeddings and determining relationships between transactions.
* Fraud detection challenges: By considering the relationships between transactions, merchants, and accounts, the technology can overcome challenges in detecting fraudulent transactions that may not be apparent through traditional methods.

Benefits:
* Improved fraud detection accuracy: By considering the relationships between transactions, the technology can enhance fraud detection accuracy and reduce false positives.
* Enhanced customer experience: By analyzing transaction data and providing personalized recommendations, the technology can improve the overall customer experience.
* Efficient data analysis: The use of neural embeddings and a trained neural network allows for efficient analysis of large volumes of transaction data.


==Original Abstract Submitted==

systems, methods, and computer program products to provide neural embeddings of transaction data. a network graph of transaction data based on a plurality of transactions may be received. the network graph of transaction data may define relationships between the transactions, each transaction associated with at least a merchant and an account. a neural network may be trained based on training data comprising a plurality of positive entity pairs and a plurality of negative entity pairs. an embedding function may then encode transaction data for a first new transaction. an embeddings layer of the neural network may determine a vector for the first new transaction based on the encoded transaction data for the first new transaction. a similarity between the vectors for the transactions may be determined. the first new transaction may be determined to be related to the second transaction based on the similarity.



[[Category:G06N3/08]]
[[Category:G06N20/00]]
[[Category:G06Q20/16]]
[[Category:G06F16/25]]