计算机技术与发展2018，Vol.28Issue(4)：109-113,118,6.DOI:10.3969/j.issn.1673-629X.2018.04.023
基于HMM与RBF混合模型的情感智能聊天系统
Emotional Intelligence Chat System Based on HMM and RBF Mixture Model
摘要
Abstract
The existing intelligence chat system mainly communicate by text,which cannot accurately answer the question,shift the topic, and obtain all relevant answers.Thus,the further study,improvement and upgrading of emotional intelligence chat system can not only promote its development,but also provide people with a more humanized,intelligent service.In this paper,we propose a HMM and RBF mixture model to build the human emotional speech libraries which is dynamically updated based on Flex technology,catching the most real emotions of human experience.At the same time,this system,which uses the corpus annotation scheme,ensures the completeness of corpus by means of a corpus quality monitoring method of the tagging criterion,error correction mechanism,supplementary learning.Us-ers can chat with text or voice by this system which generates text chat record,breaking the existing limitations of using text chat.关键词
隐马尔可夫模型/径向基函数/Flex技术/情感智能/语料库收集Key words
HMM(hidden Markov model)/RBF(radial basis function)/Flex technology/emotional intelligence/corpus collection分类
信息技术与安全科学引用本文复制引用
闫丹阳,姜梅,耿秀丽,闫伟..基于HMM与RBF混合模型的情感智能聊天系统[J].计算机技术与发展,2018,28(4):109-113,118,6.基金项目
教育部留学回国基金 （）
山东省高等学校科技计划项目(J15LN26) （J15LN26）
山东师范大学大学生创新创业训练计划项目 （）
