【摘要】 基于自然语言句法歧义消解常用的一种概率模型—概率上下文无关文法,融入上下文相关的概率信息,提出一种新的歧义消解计 算模式,该模式经测试可以有效地提高句法分析中歧义消解的正确率。更多还原

【Abstract】 In natural language parsing, probabilistic context-free grammar (PCFG) is a common model used in disambiguation. In this paper, we combine PCFG with context-sensitive probabilistic information, and propose a new disambiguation method. This method can effectively improve the correctness of syntactic disambiguation. ;更多还原