基于潜在语义分析的藏文Web不良信息检索算法研究

Study on Filtering System of Harmful Information on Tibetan WEB Based on Latent Semantic Analysis

根据不良信息的特点对潜在语义分析(Latent Semantic Analysis LSA)进行了简化,并设计了基于简化的潜在语义分析(Latent Semantic Analysis Of a Simplified LSAS)的藏文Web不良信息检索算法.该检索算法能够对具有关于某个主题的特定倾向的文本进行过滤.该系统充分利用了领域知识,采用了潜在语义模式分析等技术.实验表明该系统具有查全率和查准率高,速度较快的特点.

The latent semantic analysis {LSA） was simplified based on the text network bad information. Latent Semantic Analysis Of a Simplified LSAS was designed according to the hiding of latent language righteousness analysis calculate way. The inspectional calculate way can originally carry on percolation to the text of particular tendency that has concerning a certain topic.. The system well made use of realm knowledge and adopted latent language righteousness mode to analyze an etc. technique, the experiment expressed that it checks whole rateses and checks a quasi - rate Gao, and the speed is quicker.