摘要
股票市场是一个特殊的知识领域,词汇极性是词汇所体现的与股票价格关联的特性。正确识别词汇的极性对提高信息处理系统的结果至关重要。该方法在股评文本中抽取词汇之间的逻辑关系作为极性约束条件,利用松弛标记法善于处理大量约束条件的特点,对每个词汇进行标记指派。实验结果表明该方法是股票市场词汇极性识别的一种有效解决方法。
Stock market is a special knowledge domain,and word polarity is the property of the word embodied in association with stock price.It is crucial to recognise correctly the word polarity for improving the result of information processing system.We first extract the logic relationship between words in stock comment documents as polarity constrains,and then assign polarity label to each word using the characteristic of relaxation labelling which is good at handling large scale of constrains.Experimental results show that it is an effective method in stock market word polarity recognition.
出处
《计算机应用与软件》
CSCD
2010年第10期162-164,174,共4页
Computer Applications and Software
关键词
股票市场
松弛标记法
词汇倾向性
Stock market
Relaxation labelling
Word orientation
