基于构成模式的部队机构名识别

Unit Organization Names Recognition Based on Composition Models

部队机构名的识别是军事信息处理领域的一项重要课题,也是命名实体识别研究的重要内容之一。部队机构名识别的难点主要在于其结构的复杂性,这些复杂的结构是有规律可循的,并可以将其总结成若干的固定模式。该研究正是通过总结部队机构名的构成模式,从而制定出基于构成模式的识别方法。设计并实现一个基于构成模式的部队机构名自动识别程序,通过真实语料对系统进行评测,取得较好的效果。

Unit organization names recognition is an important task in unit information processing, and it is also an important part of Named Entity Recognition（NER）. The major difficulty of unit organization names recognition lies in the complexity of its structure. These complex struc- tures have their own rules, and they can be summed up into a number of fixed patterns. This study summarizes the composition models of unit organization names, in order to develop the reorganization meshed. Designs and implements a program using for automatic recogniz- ing unit organization names. Through the evaluation, the method achieves a good result.