- 《Using Search-Logs to Improve Query Tagging》,google论文https://static.googleusercontent.com/media/research.google.com/zh-CN//pubs/archive/38276.pdf
根据搜索查询语料来标注词性标签。基于统计的方法来做。
搜索:budget rent a car
搜索结果:结果名字不符的词性。
根据统计:基础结果π(t|w) 和 上下文中的结果φ(t|w,s)
- 《FLORS: Fast and Simple Domain Adaptation for Part-of-Speech Tagging》http://www.aclweb.org/anthology/Q14-1002
word 特征:左邻居个数、右邻居个数、0-1后缀特征、0-1形状特征
分布特征:Distributional features. We follow a long tradition of older (Finch and Chater, 1992; Schütze, 1993; Schütze, 1995) and newer (Huang and Yates, 2009) work on creating distributional features for POS tagging based on local left and right neighbors.
形状特征:数字、大小写、ed、ing结尾