摘要
为研究哈萨克语的句法分析,研究自然语言处理的统计方法的句法分析。根据概率上下无关文法模型和Chart算法的特点,将概率引入Chart算法,提出一种PChart算法。利用PChart算法对哈萨克语进行句法分析,解决句法分析中哈萨克文法的获取和歧义消解问题,实现一个基于PChart算法的哈萨克语句法分析器,识别正确的哈萨克语句子。实验结果表明,该分析器在准确率、召回率等方面的性能都有所提高。
To research the syntactic analysis of Kazakh language in natural language,the syntactic analysis of statistical methods was researched.Based on the characteristics of probabilistic context free grammar and Chart algorithm,the probability was introduced to Chart algorithm and the probabilistic Chart algorithm was proposed.Probabilistic Chart algorithm was used to do the Kazakh syntactic analysis,Kazakh grammar acquisition and disambiguation issues in syntactic analysis were solved,a Kazakh language syntax parser based on probabilistic Chart algorithm was carried out,and the correct result of Kazakh phrase analysis sentence was recognized.Experimental results show that the accuracy and the recall rate of the Kazakh syntactic parser with probabilistic Chart are improved.
出处
《计算机工程与设计》
北大核心
2016年第3期832-836,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(61063025
61063062)
