中文地址识别算法研究及在医院的应用
详细信息 查看官网全文
- 英文论文题名：The Study And Applications In Hospital Of Chinese Address Recognition Algorithms
- 论文作者：邓兴文 ; 吴庆斌
- 英文论文作者：Deng Xing-wen ; Wu Qing-bin ; Guangzhou Overseas Chinese Hospital Information Sector
- 年：2016
- 作者机构：暨南大学附属第一医院;
- 论文关键词：地址识别 ; 正则表达式 ; 文本相似度 ; 命名实体识别
- 英文论文关键词：address recognition ; regular expression ; text similarity ; named entity recognition
- 会议召开时间：2016-04-08
- 会议录名称：广东省医学装备学会2016年度学术年会暨第四届理事会第一次全体理事会议资料汇编
- 语种：中文
- 分类号：R197.32;TP391.1
- 学会代码：GDKL
- 会议名称：广东省医学装备学会2016年度学术年会暨第四届理事会第一次全体理事会议
- 会议地点：中国广东江门
- 主办单位：广东省医学装备学会
- 学会名称：广东省科学技术协会科技交流部
- 页数：5
- 文件大小：1101k
- 原文格式：O
- 会议级别：地方
摘要
中文地址分析技术在本地搜索服务已经得到广泛应用,其实其在医院管理特别是医院病人来源性分析、病案地址标准化上报也存在重要的作用。该技术就是把自然语言描述的地址转化成机器可识别可定位的信息。基于正则表达式匹配无语义的方法,识别效果较差,全文的最大相似度计算方法虽然效果好,但是需要占用大量的计算资源和完整的基础资料,而利用分词和基于命名实体识别的思想来对地址进行分析,使得整体效果有了明显改善。
Chinese address analysis technology has been widely used in local search services,and actually it is also useful in hospital management,especially in the analysis of the source of patients and the standardization report of medical record address.This technology address just intend to make machine can recognize location information from address text described by natural language,method based on Regular expression non-semanticmatching has a poor recognition effect,method based on maximum string similarity has great effect,but need a lot of compute resource and complete basic data,while analyzing address using segment and the idea based onnamed entity recognition can make improvement in overall effect.
Chinese address analysis technology has been widely used in local search services,and actually it is also useful in hospital management,especially in the analysis of the source of patients and the standardization report of medical record address.This technology address just intend to make machine can recognize location information from address text described by natural language,method based on Regular expression non-semanticmatching has a poor recognition effect,method based on maximum string similarity has great effect,but need a lot of compute resource and complete basic data,while analyzing address using segment and the idea based onnamed entity recognition can make improvement in overall effect.
引文
[1]谭侃侃.基于规则的中文地址分词与匹配[D].山东科技大学,2011
[2]赵阳阳,王亮,仇阿根.地址要素识别机制的地名地址分词算法[J].测绘科学,2013,38(5):74-76
[3]徐聪,张丰,杜震洪等.基于哈希和双数组trie树的多层次地址匹配算法[J].浙江大学学报(理学版),2014,41(2):217:222
[2]赵阳阳,王亮,仇阿根.地址要素识别机制的地名地址分词算法[J].测绘科学,2013,38(5):74-76
[3]徐聪,张丰,杜震洪等.基于哈希和双数组trie树的多层次地址匹配算法[J].浙江大学学报(理学版),2014,41(2):217:222