A cross-border multi-currency transaction automatic accounting method based on deep learning
By using deep learning technology to clean, aggregate, and compensate cross-border transaction data, the problem of inefficient accounting for many-to-many entangled transactions has been solved, achieving high accuracy and automated cross-border transaction accounting.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 深圳欧税通技术有限公司
- Filing Date
- 2026-04-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing cross-border transaction automatic accounting technologies suffer from computational inefficiency, redundant accounting entries, and an inability to perceive the overall semantics of the set in many-to-many entangled transaction scenarios, making it difficult to meet the real-time accounting needs of large enterprises.
A deep learning-based approach is adopted to clean multi-source transaction data through normalization operators, construct feature tensors, extract deep semantic features using feature embedding architecture, aggregate transaction entities using entangled state deconstruction logic, establish a nonlinear joint correlation model, and generate accounting vouchers through dynamic exchange rate residual compensation.
It enables efficient and automated accounting for cross-border multi-currency transactions, reduces redundant outstanding accounts, improves accounting accuracy, reduces manual verification workload, and meets the real-time accounting needs of large enterprises.
Smart Images
Figure CN122415220A_ABST