||
本文为美国圣何塞州立大学(作者:Vyas Ajay Bhagwat)的硕士论文,共56页。
自然语言处理(NLP)要求对语言语义之间的复杂关系进行建模。虽然传统的机器学习技术被用于NLP,但是为会话构建的模型(称为聊天机器人)并不能真正通用。虽然聊天机器人是用传统机器学习技术设计的,但是深度学习使得NLP中的复杂性更容易建模,并且可以用来构建一个与人进行真正对话的聊天机器人。在这个项目中,我们将探讨用于构建聊天机器人的问题和技术,讨论了在哪些方面可以进行改进。我们分析了不同的体系结构来构建聊天机器人,并提出了一个混合模型,一部分基于检索,另一部分基于生成,该模型获得了最好的结果。
Natural Language Processing (NLP) requiresmodelling complex relationships between the semantics of the language. Whiletraditional machine learning techniques are used for NLP, the models built forconversations, called chatbots, are unable to be truly generic. While chatbotshave been made with traditional machine learning techniques, deep learning hasallowed the complexities within NLP to be easier to model and can be leveragedto build a chatbot which has a real conversation with a human. In this project,we explore the problems and techniques used to build chatbots and whereimprovements can be made. We analyze different architectures to build chatbotsand propose a hybrid model, partly retrieval-based and partly generation-basedwhich gives the best results.
更多精彩文章请关注公众号:
Archiver|手机版|科学网 ( 京ICP备07017567号-12 )
GMT+8, 2024-9-27 06:48
Powered by ScienceNet.cn
Copyright © 2007- 中国科学报社