Line item transaction inference using merchant data and website interaction
AssigneeCapital One Services, LLC
InventorsKevin Osborn, Francis Ogbennah, Anita Eradla
Life of this patent
Estimating the term…
Abstract
Aspects described herein relate to methods, systems, and apparatuses that infer which items were purchased in a transaction based on correlating a transaction record for the purchase, such as credit card receipt, with browsing records received from multiple users of vendor websites that identify individual costs of items.
Claims20
A computer implemented method for determining purchased items in a transaction based on a transaction record and website data, the method comprising:
- receiving, via a network from a plurality of client computers, a plurality of browsing records indicating that a plurality of websites were displayed via web browsers run on the plurality of client computers;
- analyzing the plurality of browsing records to determine a plurality of individual costs respectively of a plurality of items from a plurality of vendors;
- receiving, via the network, the transaction record identifying:
- one of the plurality of vendors, and a total cost of the transaction;
- correlating the total cost of the transaction with the plurality of individual costs to identify a subset of the plurality of items included in the transaction;
- and
- updating, based on identifying the subset, one or more databases that track purchases of the plurality of items.
A server comprising:
- at least one computer processor;
- and
- computer memory comprising computer-executable instructions that when executed by the at least one computer processor, cause the server to:
- receive, via a network from a plurality of client computers, a plurality of browsing records indicating that a plurality of websites were displayed via web browsers run on the plurality of client computers;
- analyze the plurality of browsing records to determine a plurality of individual costs respectively of a plurality of items from a plurality of vendors;
- retrieve, from one or more databases, a plurality of transaction records each identifying:
- one of the plurality of vendors, and a total cost of a respective transaction of a plurality of transactions;
- and
- for each respective transaction of the plurality of transactions:
- correlate the total cost of the respective transaction with the plurality of individual costs to identify a subset of the plurality of items included in the respective transaction, and
- update the one or more databases to associate the subset with the respective transaction.
A non-transitory machine-readable medium having instructions stored thereon that, when executed by one or more processors, cause the one or more processors to:
- receive, via a network from a plurality of client computers, a plurality of browsing records indicating that:
- a plurality of websites were displayed via web browsers run on the plurality of client computers, and the plurality of web sites offered a plurality of items for sale from a plurality of vendors;
- receive, via the network, a plurality of transaction records associated with a user account, wherein each of the plurality of transaction records identify:
- one of the plurality of vendors, and a total cost of one of a plurality of transactions;
- retrieve, from one or more databases, a classifier model;
- evaluate the classifier model with information from the plurality of browsing records and information from the plurality of transaction records as inputs to determine a likelihood that each of the plurality of items was included in each of the plurality of transactions;
- generate, based on the likelihood, a hierarchical data structure comprising:
- a first list including the plurality of transactions, and for each transaction of the plurality of transactions, a subordinate list including a respective subset of the plurality of items associated with the respective transaction;
- receive, via the network from a particular client computer, a query identifying the user account;
- and
- transmit, in response to the query, the hierarchical data structure to the particular client computer.
Dependencies are inferred from the claim text.