命名实体识别是自然语言处理必不可少的重要部分,其中组织机构名识别占了很大的比例.提出了基于词频统计的组织机构名识别方法.训练数据主要通过百度百科词条整理得到.训练时,利用百度百科词条名在词条文本中的频数统计进行机构构成词的词频统计.在此基础上,构建了数学模型,实现了组织机构名识别算法.该识别算法集成到了中文分词中,取得了较好的识别结果,可以满足一定的实际应用需求.%Named entity recognition(NER)is a very important part of natural language processing.And,organization name recognition accounts for a large proportion of NER.This paper proposed the recognition algorithm based on word frequency statistics.Training data mainly came from Baidu encyclopedia entries.In the process of training,organization names were split into a number of words and all the words frequency were computed.On the basis of computation of words frequency,it established the mathematical model and implemented the algorithm of organization name recognition.The algorithm was integrated into Chinese word segmentation system,it achieved good result and met some application requirements.
展开▼
