摘要
因分布式数据库多区块链的数据种类繁多、基数大,导致目标信息查找较为困难。为解决上述问题,提出一种结合支持向量机划分的多区块链属性连接算法。利用支持向量机(Support Vector Machine, SVM)的非线性特点,将数据点划分为线性可分和不可分两种,将多区块链上各点的分布看作二维平面,计算平面上边缘点与中线点的距离关系,结合最近向量到最远向量的标准值,得到平面上每个数据点与中心点的偏离系数,采用对偶方程求解最佳划分中线,实现多区块链数据划分。设计多区块链数据连接池,定义数据的语义关键词属性并计算信任度以及信任度连接支持度,根据数据支持度查找实现属性区块链连接。实验数据证明,经连接后用户查询目标消耗时间短,查询代价低于4GB,证实了所提方法提高了查询效率,可保证多区块链间有效连通。
In this paper,an algorithm for multi-blockchain attribute connection algorithm based on Support Vector Machine partitioning was put forward to solve the problem of difficult information retrieval in distributed databases.This algorithm first utilized the nonlinear characteristics of Support Vector Machine(SVM)to divide the data points into linearly separable problems and linearly inseparable problems.The distribution of points on the multi-blockchain was regarded as a two-dimensional plane.Then,the distance relationship between the edge points and midline points on the plane was calculated.Based on the standard values of the nearest vector to the farthest vector,the deviation coefficient of each data point from the center point on the plane was calculated.Next,the dual equation was used to solve for the optimal midline,thus achieving the partitioning of multi-blockchain data.Moreover,a multi-blockchain data connection pool was designed.Finally,the semantic keyword attributes of data were defined.Meanwhile,the trust degree and connection support degree were calculated.According to the data support degree,the attribute blockchain connection was realized.Experimental results prove that after connection,the consumption time for users to query targets is very short,and the query cost is less than 4GB,which confirms that the proposed method improves query efficiency and ensures effective connectivity between multi-blockchains.
作者
李之玲
李柯景
LI Zhi-ling;LI Ke-jing(School of Computer Science and Technology,Changchun University,Changchun Jilin 130022,China)
出处
《计算机仿真》
2025年第1期390-394,共5页
Computer Simulation
基金
吉林省教育厅科学研究项目(JJKH20230678KJ)。
关键词
分布式数据库
多区块链
属性连接
支持向量机
偏离系数
Distributed database
Multi-blockchain
Attribute connection
Support vector machine
Deviation coefficient
