Anonymization Server for Multi-Institution Financial Model Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI models for financial product recommendation lack accuracy when trained solely on data from individual financial institutions, as they do not leverage the comprehensive financial transaction information held by multiple institutions, which could enhance analysis precision.
Innovation Solution
A model generation device and system that receives anonymized financial transaction information from multiple financial institution servers, generates a model for analyzing financial transactions, and outputs this model for analysis, ensuring customer privacy is protected through anonymization processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If financial transaction information is aggregated from multiple financial institutions to improve model accuracy, then measurement precision of financial transaction analysis is improved, but customer privacy protection deteriorates
Solution Approach 1:
An anonymization server acts as an intermediary between financial institutions and the model generation device. This server removes personally identifiable information from customer data before transmission, enabling multi-institutional data aggregation while protecting customer privacy. The anonymization server processes data from multiple financial institutions and outputs anonymized datasets that can be used for model training without exposing customer identities.
2Object-affected harmful factors
If customer information is anonymized to protect privacy, then customer privacy protection is improved, but information availability for model training deteriorates
Solution Approach 1:
The anonymization process extracts and removes personally identifiable information (PII) from customer datasets while retaining non-identifying financial transaction characteristics. This extraction of sensitive identifiers enables the preservation of useful analytical information needed for model training while eliminating privacy risks associated with identifiable customer data.
Data Source
AI summary
A model generation device according to the present disclosure comprises: an information receiving means that receives input of financial transaction information including customer information which has been anonymized in each of a plurality of financial institution servers; a model generation device that generates a model for analyzing financial transactions using the financial transaction information received from the plurality of financial institution servers; and an outputting means that outputs the model generated by the model generation means.