| 引用本文: | 钟志农,刘方驰,吴烨,等.主动学习与自学习的中文命名实体识别.[J].国防科技大学学报,2014,36(4):82-88.[点击复制] |
| ZHONG Zhinong,LIU Fangchi,WU Ye,et al.Chinese named entity recognition combined active learning with self-training[J].Journal of National University of Defense Technology,2014,36(4):82-88[点击复制] |
|
| |
|
|
| 本文已被：浏览 9833次 下载 7319次 |
| 主动学习与自学习的中文命名实体识别 |
| 钟志农, 刘方驰, 吴烨, 伍江江 |
|
(国防科技大学 电子科学与工程学院， 湖南 长沙 410073)
|
| 摘要: |
| 命名实体识别是信息抽取中的一项基础性任务，如何利用丰富的未标注语料来提高实体识别的指标是该领域一个重要的研究方向。基于条件随机场提出一种将主动学习与自学习相结合的方法——SACRF，通过设置置信度函数和2-Gram频度阈值来选取样本，并采用人工与自动相结合的方式进行标注来扩展训练语料。实验表明，该方法在提高实体识别的精确率和召回率的同时，能够显著地降低人工标注的工作量。 |
| 关键词: 主动学习 自学习 条件随机场 命名实体识别 |
| DOI：10.11887/j.cn.201404015 |
| 投稿日期：2013-10-11 |
| 基金项目:国家高技术研究发展计划主题项目（2011AA120300）；湖南省自然科学基金资助项目（11JJ4028） |
|
| Chinese named entity recognition combined active learning with self-training |
| ZHONG Zhinong, LIU Fangchi, WU Ye, WU Jiangjiang |
|
(College of Electronic Science and Engineering, National University of Defense Technology, Changsha 410073, China)
|
| Abstract: |
| Named Entity Recognition (NER) is a basic task in information extraction, and it is an important research direction in this domain to use the abundant unlabeled corpus to improve the performance of NER system. An approach combining self-training with active learning based on CRF (SACRF) is proposed. It selected samples by setting the threshold of confidence and 2-Gram frequency, and expanded the training set by annotating the unlabeled corpus manually and automatically. The experiments revealed that this approach can not only improve the precision and recall of NER system, but also reduce the manually annotation efforts greatly. |
| Keywords: active learning self-training conditional random fields named entity recognition |
|
|
