基于深度学习的藏文人名地名识别  

Tibetan Named Entity Recognition Based on Deep Learning

作　　者：朱亚军 拥措 ZHU Yajun;Yongtso(School of Information Science and Technology,Tibet University,Lhasa Tibet 850000,China)

机构地区：[1]西藏大学信息科学技术学院,西藏拉萨850000

出　　处：《信息与电脑》2022年第5期66-68,共3页Information & Computer

基　　金：西藏大学研究生高水平人才培养计划《藏文预训练语言模型研究》(项目编号:2019-GSP-S118)。

摘　　要：藏文人名地名识别主要是从藏文文本中识别出人名、地名等专有名词。藏文文本语法较为复杂,使用传统机器学习方法人工构建特征难度大。深度学习(DeepLearning,DL)可以对任务数据自动构建特征,使用BiLSTM提取藏文文本特征,采用CRF对标签序列准确解码。实验证明基于BiLSTM-CRF的藏文人名地名识别模型能够有效识别藏文文本中的人名、地名,并且F1值达82.1%。Tibetan named entity recognition mainly identify proper nouns such as person’s name and place name from Tibetan text. The syntax of Tibetan text is complex, and it is difficult to construct features manually by using traditional machine learning methods. Deep Learning(DL) can automatically construct features for task data, extract Tibetan text features with BiLSTM, and accurately decode the predicted tag sequence with CRF. Experiments show that the Tibetan named entity recognition model based on BiLSTM-CRF can accurately recognize people’s names and place names in Tibetan texts, and theF1 score is 82.1%.

关 键 词：藏文 命名实体识别 深度学习 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]