网络数据库访问中语义指向性算法优化

Optimization of semantic directivity algorithm for network database access

为解决数据库从高维单词空间映射至低维隐含语义空间中,无法有效实现数据库访问语义指向性分析的问题,提出基于主题模型的数据库访问语义指向性算法,建立PLSA主体模型并对其进行求解,通过PLSA主题模型获取理想的潜在语义主题,在数据库访问关键词上分布以及文档在潜在语义主题上的分布,将其应用于数据库访问语义指向性分析中,针对数据库表现出来的文本特征和结构特征建立PLSA主题模型,通过自适应不对称学习算法对不同的PLSA主题模型进行集成和优化,以实现数据库访问语义指向性分析,使数据库访问结果更加准确。仿真实验结果表明所提算法具有很高的数据库访问效率及精度。

To solve the problem that the database is mapped from high?dimensional word space to low?dimensional impliedsemantic space,and can not effectively implement semantic directivity analysis of database access,the database access seman?tic directivity algorithm based on theme model is proposed,the PLSA subject model is established and is solved,by which theideal latent semantic theme is obtained. The key distribution on the database access and document distribution on latent seman?tic subject are applied to the database access semantic directivity analysis to set up PLSA theme model according to the text fea?ture and structure feature showed by database. The different theme PLSA models are integrated and optimized by adaptive asym?metry learning algorithm to realize the semantic directivity analysis for database access and make the database access resultsmore accurate. The simulation results show that the proposed algorithm has high database access efficiency and accuracy.