Payment Prediction System Using Remote Data Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current customer prediction techniques are limited by the source data used, preventing accurate payment-related predictions, as they primarily rely on internal data and cannot effectively incorporate data from other business entities.
Innovation Solution
A machine learning system that collects and processes data from multiple independent sources, including telephone numbers and webpage URLs of other providers, to form input data, which is then used to make payment-related predictions for customers, enhancing prediction accuracy by incorporating remotely sourced information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current customer prediction techniques use only internal activity data, then the system complexity remains low, but the prediction accuracy for payment-related behaviors is insufficient
Solution Approach 1:
The patent combines internal activity data from the communication service provider with external data from multiple independent sources (telecom providers, payment processors, web analytics services) to create a comprehensive dataset for payment-related customer predictions, thereby improving prediction accuracy while managing complexity through systematic integration
Solution Approach 2:
The system processes multiple types of data from various sources (call detail records, text messages, webpage URLs, IP addresses) through a unified machine learning framework that handles diverse data formats and sources, enabling the system to make accurate payment predictions using multi-functional data processing capabilities
2Measurement precision
If data from multiple independent sources is collected and processed, then payment-related prediction accuracy improves, but the data processing complexity increases
Solution Approach 1:
The patent segments the data processing into distinct stages: collecting data from multiple independent sources, matching external data to internal customer records using identifiers like phone numbers and IP addresses, processing the matched data through machine learning algorithms, and generating predictions. This segmentation reduces processing complexity by organizing complex operations into manageable steps
Solution Approach 2:
The system uses matching identifiers (telephone numbers, IP addresses, device identifiers) as intermediaries to connect external data from independent sources with internal customer records. These intermediaries enable accurate data integration without requiring direct complex connections between all data sources and the prediction system
Data Source
AI summary
As described herein, a machine learning system, method, and computer program are provided for making payment related customer predictions using remotely sourced data. A system of a communication service provider (CSP) identifies a customer of the CSP. Additionally, the system collects data from a plurality of data sources independent from the CSP, the data including telephone numbers and/or webpage URLs of other services providers that are associated with making payments. Further, the system processes the collected data to form input data indicating which of the telephone numbers were contacted by the customer and/or webpage URLs were accessed by the customer. Still yet, the system processes the input data using at least one machine learning algorithm to make at least one payment related prediction for the customer. Moreover, the system outputs the at least one payment related prediction made for the customer.