例句與造句

1. It process documents not only based on latent semantic analysis , but also based on text multilevel dependency structure . the method first analysis the latent semantic structure of texts , make single value decomposition on text - matrix , reconstruct the semantic matrix ; then a method based on text multilevel dependency structure is adopted , deeply analysis the content of the semantic matrix , abstract the important sentences to generate abstraction and make up the shortage of latent semantic analysis on structure and syntax
   首先通過對文本進(jìn)行潛在語義分析,對文本矩陣進(jìn)行相應(yīng)的奇異值分解,重構(gòu)語義矩陣;然后采用基于篇章多級依存結(jié)構(gòu)的文摘分析方法,對重構(gòu)的語義矩陣表示的文本內(nèi)容進(jìn)行深入的分析,抽取重要的句子生成文摘,這樣就彌補(bǔ)了潛在語義分析在詞法和句法分析上的不足;同時過濾和去除了語義噪音,縮小了問題的規(guī)模。
2. It's difficult to find semantic matrix in a sentence. 用semantic matrix造句挺難的