Supervised Model User Authentication via Transaction Details
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional user authentication methods in online transactions often require users to input intrusive or hard-to-remember information, hindering the transaction process and user experience.
Innovation Solution
A system and method that uses a supervised model, such as a random forest, to determine user identity based on transaction details without requiring username and password, by transforming input values into search terms, calculating inverse document frequencies, and ranking historical records to generate a confidence score for user authentication and personalized user interface customization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional user authentication methods requiring username and password input are used, then security verification is achieved, but user experience deteriorates due to intrusive and hard-to-remember information requirements
Solution Approach 1:
The system automatically performs authentication by matching transaction details against historical records without requiring active user participation. The supervised model independently verifies user identity by comparing current transaction data with stored historical patterns, eliminating the need for users to manually input credentials.
Solution Approach 2:
The patent replaces the mechanical interaction of manual password entry with an automated information processing system. A supervised machine learning model processes transaction details, transforms them into search terms, calculates inverse document frequencies, and ranks historical records to automatically determine user identity, substituting human cognitive effort with computational analysis.
2Speed
If supervised model processing is implemented for user authentication, then authentication speed is improved, but system complexity increases due to multiple processing steps including inverse document frequency calculation and historical record ranking
Solution Approach 1:
The system pre-processes and stores historical transaction records in a structured format with calculated inverse document frequencies before authentication is needed. This preliminary organization of data allows the supervised model to quickly retrieve and compare relevant records during actual authentication, reducing real-time processing complexity.
Solution Approach 2:
The patent introduces an intermediary layer of inverse document frequency calculation that bridges raw transaction data and historical record comparison. This intermediary metric transforms diverse transaction details into a standardized scoring system, simplifying the matching process between current and historical records.
Data Source
AI summary
Details associated with a transaction are received from a first computing device. An identity of a party to the transaction is determined. An address to which communications to the party can be directed and accessed via a second computing device is obtained, where the second computing device is associated with the party. Confirmation data specific to the transaction is generated. The confirmation data is provided, via the address, to the second computing device. Confirmation of receipt of the confirmation data is received from the party. A set of payment options is provided to the part. A selection of a payment option is received from the party, and notice that the transaction has been completed is provided to the first computing device.