摘要
引入SCNF语法,据此进行概率CYK剖析,得到句子的最佳剖析树以消除歧义。此外通过重排SCNF语法规则和建立索引表对概率CYK算法进行改进,减少中间搜索步骤所需的时间以提高效率。文中给出一些例子作为比较,以验证这种方法的有效性。
Stochastic Chomsky normal form (SCNF) is introduced to parse sentences with probabilistic CYK algorithm .The optimal parsing trees of sentences are gained to eliminate ambiguous meanings. Besides, the search steps in probabilistic CYK algorithm is reduced and the probabilistic CYK algorithm is improved by rearranging grammar rules of SCNF and establishing the index table.Several experiments are given for comparison to validate the approach.
出处
《微计算机信息》
2010年第21期157-158,共2页
Control & Automation