摘要
人机会话系统是指运用自然语言理解技术实现的人机交流的智能系统,也就是所谓的“聊天机器人”。提出了一个基于特征词的静态个人信息模型,结合句子生成的需要,把描述个人信息的特征词分为特征属性和特征动词两种类型;与此同时,给出了一种句模与规则相结合的句子生成方法,并重点讨论了疑问句型的特征属性;最后设计、实现了一个基于“个人介绍”主题的人机会话系统,该系统具有简单的逻辑推理和上下文相关问题的处理能力,能够以多种方式提问和应答个人基本信息方面的问题,并已经在自行设计的聊天室中成功地运行。
Man-machine conversation system is an intelligence system which uses natural language understanding technology to realize the man-machine conversation, called Chat Robot. A static individual information pattern is introduced based on character, which is divided into two types ofcharacteristic attribute and characteristic verb by the requirement of sentence-building, At sametime, an sentence building way combined with sentence patterns and rules is given, and the characteristic attributes of question sentence is emphasized on, Finally, an Man-machine conversation system on individual-introduce theme is designed and implementesd, which has the characters of simple reasoning, context problem treating, has the ability of asking and answering questions about individual information in many ways, and successfully run on the chatting room designed by ourselves.
出处
《计算机工程与设计》
CSCD
北大核心
2005年第9期2533-2536,共4页
Computer Engineering and Design
关键词
自然语言理解
人机会话系统
聊天机器人
主题会话
个人信息模型
natural language understanding
man-machine conversation system
chat robot
theme conversation
individual information pattem