摘要
本文着重介绍利用“后邻字符树”的方法在领域语料库中生成字符树库,用于自动文摘候选句子选取时提高精度,介绍了后邻字符树的构造。
A method of generating Rear Charater Trees (RCT) in domanial corpus is presented in this paper.The construction of RCT,the generation and optimization of the RCT database as well as the computation of sentenee weight are discussed in detail.
出处
《中文信息学报》
CSCD
北大核心
1999年第2期16-23,共8页
Journal of Chinese Information Processing