Font Size: a A A

Design Of Voice Interaction Systerm Based On Intelligent Terminal

Posted on:2019-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:Y W LiuFull Text:PDF
GTID:2348330545955777Subject:Logistics Engineering
Abstract/Summary:PDF Full Text Request
Voice interaction has drawn lots of attentions during last decade.With the development of science and technology,people want to communicate with computer through natural language,so the research process of voice interaction has been pushed to a new level,and intelligent chatbot is the product which is new born under such historical background.Intelligent chatbot is a kind of Human-computer dialogue intelligent system which can freely communicate with people through natural language.With the development of information society,intelligentialize becomes more and more important for us.In the paper,we introduced the research background of voice interaction,and the research actuality of key points.Besides,introduced the deep neural networks that employed in this paper,including recurrent neural network,long-short term memory network,and the seq2seq model.The study of natural language can be divided into two areas:specific field and open field.For the specific domain of Natural Language Processing,this paper uses the template matching method based on regular expression.As for the semantic understanding of the chatting robot in the open field,the seq2seq model based on recurrent neural network is employed in this paper.The latter part of this paper discusses the application of model we designed,functional realization and architecture optimization.The client uses the MVP framework to organize the code structure,and thus better managing UI changes.Finally,the client shows the use of various functional modules.In summary,this paper completes the speech interaction system based on intelligent terminals,which achieves intelligent dialogue,logistics order inquiry,Baidu search and other functions.This systerm has a good practical significance.
Keywords/Search Tags:voice interaction, natural language processing, deep learning, seq2seq modle
PDF Full Text Request
Related items