政务领域本体术语的自动抽取  

Automatic Domain-specific Term Extraction in Administrative-domain Ontology

作　　者：翟笃风[1] 刘柏嵩[2] 

机构地区：[1]宁波大学商学院,宁波315211 [2]宁波大学网络中心,宁波315211

出　　处：《现代图书情报技术》2010年第4期59-65,共7页New Technology of Library and Information Service

基　　金：国家社会科学基金项目"领域本体的自动构建和应用研究"(项目编号:08CTQ014)的研究成果之一

摘　　要：提出一种新的政务本体术语自动抽取的方法。首先通过中文分词技术和单字合并法提取政务文本中的词作为候选术语;通过C-value求解法和TF-IDF算法对候选术语进行过滤抽取,从而实现政务领域术语的自动抽取。通过实验比较,发现该方法在不影响领域术语抽取召回率的同时可以提高抽取术语的正确率。This paper introduces anew method to extract the administrative - domain Ontology term automatically. Firstly, some words that are representative of the candidate terms should be extracted through the technology of word segmentation and the characters merger method. Secondly, the candidate terms are filtered by the way of C - value method and TF -IDF algorithm to achieve the automatic domain- specific term extraction in administrative- domain Ontology. Finally, the experiment shows that this method can improve the accuracy of the extracted terms and do not affect the recall - rate.

关 键 词：政务领域本体 术语 单字合并法 C—value TF—IDF算法 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]