Blockchain storage space optimization method based on knowledge granularity-neighborhood rough entropy feature selection algorithm
Han Yang, Xueqing Zhao, Hao Liu, Xin Shi, Sunan Ge, Guigang Zhang, Yun Wang, Yibing Chen · 2023
Blockchain is a distributed database with tamper-evident and decentralized data structure. However, due to the huge amount of data in blockchain, the storage costs are is high. In an effort to minimize data storage expenses, it is necessary to choose the most pertinent data features for preservation on the blockchain. Therefore, an effective data feature selection method can quickly and accurately select the most representative data features, which can reduce the redundancy of blockchain upload data features and effectively confirm rights. This paper puts forth a feature selection algorithm (i.e. FS_NGRE) based upon the neighborhood granularity rough entropy and firefly algorithm (abbreviated to FA) to select representative data features, effectively decreasing the cost of storing data on the blockchain. Furthermore, the public datasets are employed for simulated testing. The results of these trials demonstrates that the proposed feature selection algorithm surpassed alternative feature selection strategies in the matter of accuracy, facilitating the selection of more significant features while diminishing the scale of the feature set. The proposed methodology can serve as an effective means for verifying the integrity of data stored on the blockchain.