| 基于多层术语度的一体化术语抽取研究 |
| |
| 引用本文： | 章成志.基于多层术语度的一体化术语抽取研究[J].情报学报,2011,28(3). |
| |
| 作者姓名： | 章成志 |
| |
| 作者单位： | 南京理工大学信息管理系,南京,210094;中国科学技术信息研究所,北京,100038 |
| |
| 基金项目： | 国家自然科学基金项目，教育部人文社会科学研究一般项目，"十一五"冶国家科技支撑计划重点项目 |
| |
| 摘 要： | 以往的术语抽取研究大多将语言学方法和统计方法分别进行单独的处理,并且只考虑候选术语本身的术语度,而没有考虑候选术语所在句子的术语度对术语抽取性能的影响.本文将语言学方法与统计方法进行并行融合,综合考虑候选术语及其所在语句的术语度,进行基于多层术语度的一体化术语抽取.该研究有两个特色:首先,采用条件随机场模型,能有效融合语言学方法和统计方法,实验结果表明了基于一体化策略的术语抽取方法的有效性;其次,通过语料库比较方法,提出基于多层术语度的术语抽取方法,该方法能抽取多字术语,实验结果表明了利用多层术语度进行术语抽取的有效性.
|
| 关 键 词： | 术语抽取 多层术语度 一体化策略 条件随机场 |
Using Integration Strategy and Multi-level Termhood to Extract Terminology |
| |
| Zhang Chengzhi.Using Integration Strategy and Multi-level Termhood to Extract Terminology[J].Journal of the China Society for Scientific andTechnical Information,2011,28(3). |
| |
| Authors: | Zhang Chengzhi |
| |
| Institution: | Zhang Chengzhi~(1,2) (1.Department of Information Management,Nanjing University of Science & Technology,Nanjing 210094,2.Institute of Scientific & Technical Information of China,Beijing 100038) |
| |
| Abstract: | In most previous studies on terminology extraction,linguistics methods and statistical methods were used through independent process respectively.At the same time,without considering termhood of sentence which includes the terminology candidate,only the termhood of the terminology candidate was considered.In this paper,a method based on multi-level termhood is proposed.The method uses integration strategy which ensembles the linguistics methods and statistical methods.The proposed method uses the termhood o... |
| |
| Keywords: | terminology extraction multi-level termhood integration strategy conditional random fields |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
