摘要
近年来,由于网络新闻数据的快速增长,新闻文本中出现了很多新意的词语,如何能够准确识别实体,并提出了基于新词的新闻命名实体识别方法。该方法首先利用网络资源来获得含有新词的词典,并与条件随机场相结合构建实体识别模型,然后提取新闻实体。实验结果表明,该方法在提取新闻实体方面取得较好的效果。
In recent years,due to the rapid growth of online news data,there have been many new words in news texts,how to accu-rately identify entities,and news named entity recognition method based on new words has been proposed.The method first usesthe network resources to obtain a dictionary containing new words,and constructs entity recognition model combined with condi-tional random fields,and then extracts news entities.The experimental results show that this method achieves better results in ex-tracting news entities.
作者
李娟
虞金中
LI Juan, YU Jin-zhong (School Of Computer Science, Southwest Petroleum University, Chengdu 610500 China)
出处
《电脑知识与技术》
2018年第8期153-154,共2页
Computer Knowledge and Technology
基金
国家自然科学青年基金项目(项目编号:61503312)