针对海运货物邮件实体识别中存在识别精度不高、实体边界确定困难的问题,提出一种结合深度学习与规则匹配的识别方法。其中:深度学习方法是在BiLSTM-CRF(Bidirectional Long Short Term Memory-Conditional Random Field)模型的基础上...针对海运货物邮件实体识别中存在识别精度不高、实体边界确定困难的问题,提出一种结合深度学习与规则匹配的识别方法。其中:深度学习方法是在BiLSTM-CRF(Bidirectional Long Short Term Memory-Conditional Random Field)模型的基础上添加词的字符级特征,并融入多头注意力机制以捕获邮件文本中长距离依赖;规则匹配方法则根据领域实体特点制定规则来完成识别。根据货物邮件特点将语料进行标注并划分为:货物名称、货物重量、装卸港口、受载期和佣金五个类别。在自建语料中设置多组对比实验,实验表明所提方法在海运货物邮件实体识别的F1值达到79.3%。展开更多
At 7:30 Beijing time on Ocotber17,2016,a LM-2F launch vehicle soared up from the Jiuquan Satellite Launch Center,successfully putting the Shenzhou 11 spaceship and its crew into space.Two astronauts,JING Haipeng and ...At 7:30 Beijing time on Ocotber17,2016,a LM-2F launch vehicle soared up from the Jiuquan Satellite Launch Center,successfully putting the Shenzhou 11 spaceship and its crew into space.Two astronauts,JING Haipeng and CHEN Dong will carry out the 6th manned spaceflight mission of the country.展开更多
文摘针对海运货物邮件实体识别中存在识别精度不高、实体边界确定困难的问题,提出一种结合深度学习与规则匹配的识别方法。其中:深度学习方法是在BiLSTM-CRF(Bidirectional Long Short Term Memory-Conditional Random Field)模型的基础上添加词的字符级特征,并融入多头注意力机制以捕获邮件文本中长距离依赖;规则匹配方法则根据领域实体特点制定规则来完成识别。根据货物邮件特点将语料进行标注并划分为:货物名称、货物重量、装卸港口、受载期和佣金五个类别。在自建语料中设置多组对比实验,实验表明所提方法在海运货物邮件实体识别的F1值达到79.3%。
文摘At 7:30 Beijing time on Ocotber17,2016,a LM-2F launch vehicle soared up from the Jiuquan Satellite Launch Center,successfully putting the Shenzhou 11 spaceship and its crew into space.Two astronauts,JING Haipeng and CHEN Dong will carry out the 6th manned spaceflight mission of the country.