| 摘 要: 随着自然语言处理技术的发展，对话机器人以节省人工和易于嵌入等特点受到业界青睐。为了满足在线教育的需求，本文提出的教育领域多轮对话机器人具备和用户针对知识点进行深入对话的能力。本文介绍了此对话机器人的实现全流程：采取用户模拟器生成教育领域语料，使用意图识别和槽位填充实现自然语言理解，通过对话状态追踪和对话策略设计多轮对话逻辑。这为学生提供了答疑解惑的新渠道，为线上教育的智能创新带来更多选择。 |
| 关键词: 教育；对话机器人；多轮对话 |
|
中图分类号: TP391
文献标识码: A
|
|
| Algorithm Design and Implementation of Multi-turn Chatbot in Education |
|
YAN Xiaoyu1,2, PENG Suting1,2
|
( 1. Dalian Neusoft University of Information, Dalian 116023, China;
2. Research Institute, Dalian Neusoft Education Technology Group Co . Limited, Dalian 116023, China )
yanxiaoyu@neuedu.com; pengsuting@neuedu.com
|
| Abstract: With the development of Natural Language Processing (NLP) technology, chatbots are favored for their advantages of labor saving and easy embedding. In order to meet the needs of online education, this paper proposes to design multi-turn chatbots with ability to conduct in-depth dialogue with users on knowledge points in the field of education. This paper introduces a lifecycle of realizing multi-turn chatbot: adopting user simulator to generate educational corpus, using intention recognition and slot filling to realize natural language understanding, and designing multi-turn dialogue logic through dialogue state tracking and dialogue policy. This chatbot provides students with a new way to solve their puzzles and brings more choices for intelligent innovation in online education. |
| Keywords: education; chatbot; multi-turn dialogue |
