@article{4614,
  author = {Wei Zeng},
  title = {A Rough Set Based Parallel Verification Model for Accounting Mining Systems},
  journal = {Journal of Data Processing},
  year = {2025},
  volume = {15},
  number = {4},
  doi = {https://doi.org/10.6025/jdp/2025/15/4/151-158},
  url = {https://www.dline.info/jdp/fulltext/v15n4/jdpv15n4_1.pdf},
  abstract = {The paper proposes a cloud based data integrity verification algorithm for accounting informationization
systems, leveraging rough set based data mining techniques. With the integration of cloud computing into financial
management, data security and integrity have become critical concerns due to risks like data breaches
and accidental loss as illustrated by real world incidents involving Huawei and Tencent Cloud. Traditional
verification methods, such as HMAC or RSA based Provable Data Possession (PDP), suffer from high computational
and communication overheads, especially under dynamic data operations.
To address these limitations, the author introduces a parallel verification algorithm that supports multi user
batch validation, significantly reducing communication costs and verification time. By employing rough set
theory, the system performs efficient feature selection and attribute reduction, enhancing computational
efficiency without compromising data integrity. The verification process uses BLS signatures with homomorphic
properties, allowing multiple files to be verified at the communication cost of a single file.
Experimental results show that the proposed algorithm outperforms conventional single point verification
methods, especially as user scale and data volume increase. The system also offers tiered audit levels that
balance detection accuracy, verification frequency, and cost. Implemented in a Java/SQL Server
environment, the approach demonstrates improved scalability and suitability for accounting cloud environments,
offering a practical solution for secure, efficient financial data management in the era of digital
transformation.},
}