Federated Learning Provider Matching Using a Coordinator
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In federated learning scenarios involving multiple data providers, manual screening for data providers that meet user requirements is inefficient and subjective, leading to resource waste and inaccurate model outcomes due to complex data structures and service differences in the financial industry.
Innovation Solution
A federated learning method utilizing a coordinator to store and compare data fields, enabling efficient screening of data providers by sending required data fields to the coordinator for matching, reducing unnecessary data transmission and leveraging encryption to protect privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual screening is used to select data providers in federated learning, then flexibility and adaptability are maintained, but efficiency and objectivity deteriorate due to time-consuming processes and subjective judgment
Solution Approach 1:
The patent introduces a coordinator as an intermediary component that automatically matches data providers with data users based on data field requirements. The coordinator receives matching requests, compares required data fields with available data fields from multiple data providers, and returns matched results. This intermediary mechanism eliminates manual screening while maintaining objective and accurate matching, resolving the contradiction between operational flexibility and selection efficiency.
2Reliability
If all data providers are contacted for federated learning, then comprehensive data coverage is achieved, but resource waste increases due to unnecessary data transmission and processing
Solution Approach 1:
The patent implements preliminary action by having the coordinator pre-compare data field requirements with available data fields before initiating federated learning processes. Data providers that do not meet the required data fields are filtered out in advance, preventing unnecessary data transmission and computational resource consumption. This preliminary filtering ensures that only relevant data providers are selected, maintaining comprehensive data coverage while eliminating resource waste from contacting unsuitable providers.
3Measurement precision
If data fields are transmitted for every matching request, then accurate matching is achieved, but data transmission overhead and privacy risks increase
Solution Approach 1:
The patent extracts only the essential data field information (field names and types) from the complete data sets for matching purposes. Instead of transmitting actual data values, the system compares extracted data field metadata between data users and data providers through the coordinator. This extraction approach maintains matching accuracy by preserving the structural characteristics of data fields while significantly reducing data transmission overhead and minimizing privacy risks associated with data exposure.
Data Source
AI summary
A method includes: sending, by one or more computers, in response to the number of data providers for federated learning being greater than a first threshold, a data field required for the federated learning to a coordinator, the coordinator comprising a computer; receiving, by one or more computers, from the coordinator, information about one or more data providers comprising the required data field, for determining the data providers comprising the required data field as the remaining data providers, wherein the coordinator stores a data field of each data provider; and performing, by one or more computers, federated learning-based modeling with each of the remaining data providers.