摘要
提出一种基于有穷自动机的自动识别方法,结合部队番号的构成特征,对汉语组块识别方法进行改进,提出了一种适合计算机处理的递归型部队番号组块规则,并构造了对应的LR(0)文法,实现了部队番号的有效识别。实验结果表明:该方法对作战文书中部队番号识别的召回率、精确率均在95%以上,这种识别方法是可行的。
An automatic recognition method based on finite automata was proposed.Based on the formation characteristics of the unit designations,this paper improved the recognition method of Chinese block,defined a block rule to unit designations which is suitable for computer processing,constructed the corresponding LR(0)grammar,and realized the effective recognition of the unit designations.Experimental results show that the recall and precision rate of the recognition method are all above 95%,and the method is feasible.
作者
唐坚
刘海燕
TANG Jian;LIU Haiyan(Academy of Army Armored Force,Beijing 100072,China;The No.31632nd Troop of PLA,Kunming 650000,China)
出处
《兵器装备工程学报》
CAS
北大核心
2020年第2期143-147,共5页
Journal of Ordnance Equipment Engineering
关键词
作战文书
部队番号
组块规则
有穷自动机
识别
operation documents
unit designation
block rule
finite automata
recognition