基于底表的多层扫描术语自动标注算法  

Term Auto-tagging Algorithm Based on Base Glossary

作　　者：郑泽芝[1] 敖婷[1,2] 

机构地区：[1]厦门大学人文学院,福建厦门361005 [2]福建卫生职业技术学院

出　　处：《厦门大学学报（自然科学版）》2011年第3期546-552,共7页Journal of Xiamen University：Natural Science

基　　金：国家社会科学基金项目(10BYY041);国家语言资源监测与研究中心资助项目(E080106-02)

摘　　要：以学科教材术语语料库建设为目标,实现了一种基于底表的多层扫描术语自动标注算法.该算法首先采用预测性规则模板扫描文本中未登录术语,并进行标注;其次采用最大匹配方法识别出每个可能的候选术语,把每个候选术语看作术语的定位点,扫描其上下文语境,分别调用单位术语规则模板、例外规则、部件规则、部件例外规则、例外校正规则等对扫描结果进行判断,确定候选术语的身份,并进行标注.该方法以规则的预测和限定功能为辅,充分利用了底表术语信息,取得了较高的标注准确率和召回率,开放测试F-指数达到了84%左右.A multi-scanning algorithm based on base glossary was designed for constructing a teaching material term corpus.Firstly,the term auto-tagging process scans and labels terms which match the prediction templates;secondly,finds out every candidate terms with the maximum matching algorithm based on a base glossary,and takes every candidate term as an anchor point,scans the context of the anchor point,calls the unit-term templates,exceptional-correct rules,term component and component exceptional rules in tern to judge whether the candidate string is a term or not.Together with the prediction and limited function of rules,this method makes full use of the information of terms in base glossary and achieves a higher precision and recall rate.The F-index reached about 84% in open test.

关 键 词：术语 自动标注 规则 术语部件 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]