引用本文: | 张磊,王斌,靖红芳,吴丽辉.中文网页搜索日志中的特殊命名实体挖掘[J].哈尔滨工业大学学报,2011,43(5):119.DOI:10.11918/j.issn.0367-6234.2011.05.024 |
| ZHANG Lei,WANG Bin,JING Hong-fang,WU Li-hui.Mining special named entities from Chinese Web search query logs[J].Journal of Harbin Institute of Technology,2011,43(5):119.DOI:10.11918/j.issn.0367-6234.2011.05.024 |
|
摘要: |
利用少量具有类别信息的种子词,结合特征选择技术来提取每个类别的特征信息;再利用这些特征信息,结合文本分类等数据挖掘技术来提取特殊命名实体.过程中只有构造种子词的环节需要人工辅助,其他环节均实现自动处理.实验证明,该系统和方法能够从查询日志中挖掘出高质量的命名实体列表,6个类别上识别结果的平均P@500达到了77%.系统的自动化程度和识别的效果均达到实用的要求. |
关键词: 特殊命名实体 数据挖掘 信息检索 网页搜索日志 |
DOI:10.11918/j.issn.0367-6234.2011.05.024 |
分类号:TP393.092 |
基金项目:国家自然科学基金资助项目(60603094);国家重点基础研究发展计划(973)资助项目(2007CB311103);国家高技术研究发展计划(863)资助项目(2006AA010105);北京市自然科学基金资助项目(4082030) |
|
Mining special named entities from Chinese Web search query logs |
ZHANG Lei1, WANG Bin1, JING Hong-fang1, WU Li-hui2
|
1.Institute of Computing Technology,Chinese Academy of Sciences,100190 Beijing,China;2.Informationization Division,General Office of Chinese Academy of Sciences,100864 Beijing,China
|
Abstract: |
Recognizing special Named Entities from Chinese Web search query logs is very useful for practice.In this paper,a few seed words and feature selection techniques are used to extract features from the query logs first,and then the features are combined into a text classifier for NE recognition.Experiments on a handred milllion queries to mine six types of Special Named Entities show that it is able to find high quality Sepecial Named Entities from Chinese Web search query logs.The average P@500 of the six classes achieves 77%. |
Key words: special named entities data mining information retrieval Web search query logs |