摘要
条件随机场(Conditional random field,CRF)模型是目前开展中文情感分析研究的一个热门工具。文章分析了CRF研究现状,给出了CRF适用于中文信息处理的理由,开展了基于CRF算法的比较研究:运用自然语言处理与中文计算2012年会议的公开评测结果,分别对CRF与隐马尔科夫模型和最大熵马尔可夫模型进行了比较研究,总结了CRF模型的特点。
Conditional Random Field (CRF) model is a popular tool to carry out research in Chinese sentiment analysis. This paper analyzes the research status of CRF, CRF applicable for Chinese information processing is given, and a comparative study based on CRF algorithm is carried out. Using natural language processing and Chinese to calculate public evaluation results of that meeting in 2012, CRF and hidden Markov model and maximum entropy Markov model are researched comparatively, the characteristics of CRF model is summarized.
作者
王茵
周学广
陆健
WANG Yin ZHOU Xueguang LU Jian(Computer Technology Institute of Navy, Beijing 100841 Department of Information Security, Naval University of Engineering, Wuhan 430033)
出处
《计算机与数字工程》
2017年第9期1703-1707,1730,共6页
Computer & Digital Engineering
基金
国家社会科学基金军事学项目(编号:14GJ003-152)资助
关键词
条件随机场
隐马尔科夫
最大熵马尔可夫
情感分析
中文信息处理
conditional random fields, hidden markov model, maximum entropy, emotion analysis, Chinese information processing