Key technologies of railway intelligent customer service
-
摘要: 提出铁路智能客服总体框架,研究铁路客运服务专业的语义理解与处理,深度学习模型和用户画像等关键技术,解决铁路12306互联网售票系统(简称:12306系统)客服维护成本高、服务时间受限、培训成本高、线路忙、人为错误不利于控制等问题。研究表明,铁路智能客服可有效提高铁路12306系统客服接通率和客户满意度,提升铁路用户体验,树立铁路良好社会形象。
-
关键词:
- 12306互联网售票系统 /
- 智能客服 /
- 接通率 /
- 满意度
Abstract: This paper proposed the overall framework of railway intelligent customer service, researched on key technologies of railway passenger service professional semantic understanding and processing, deep learning model and user portrait, solved the problems of high maintenance cost, limited service time, high training cost, busy line, human errors which were not conducive to control of railway 12306 Internet ticketing and reservation system.Research results show that the railway intelligent customer service can effectively improve the connection rate and customer satisfaction of the railway 12306 system, enhance railway user experience, establish a good social image of railway. -
[1] 徐理军. 基于语义的智能客户服务系统的研究与设计[D]. 扬州:扬州大学, 2012:7-8. [2] 崔悦,宋齐军. 智能语音技术发展趋势及电信运营商应用浅析[J]. 邮电设计技术, 2016(12):6-11. [3] 朱明明. 语音情感识别方法及其应用研究[D]. 合肥:安徽大学, 2013:2-3. [4] 梁静,刘刚. 基于深度学习模型的语音特征提取方法研究[EB/OL].[2013-12-31].http://www.paper.edu.cn/releasepaper/contem/201312-1241. [5] 刘文举,聂帅,梁山,等. 基于深度学习语音分离技术的研究现状与进展[J]. 自动化学报, 2016(6):25-30. [6] Bo-Hsiang Tseng, Sheng-Syun Shen, Hung-Yi Lee,et al.Towards Machine Comprehension of Spoken Content:Initial TOEFL Listening Comprehension Test by Machine[Z]. New York, USA:Cornell University,arXiv preprint arXiv:1608.06378. [7] Bhuwan Dhingra, Hanxiao Liu, William W Cohen, et al.Gatedattention readers for text comprehension[Z]. New York, USA:Cornell University,arXiv preprint arXiv:1606.01549. [8] Alessandro Sordoni, Phillip Bachman, Yoshua Bengio. Iterative alternating neural attention for machine reading[Z]. New York, USA:Cornell University,arXivpreprint arXiv:1606.02245. [9] 杨燕. 面向电商领域的智能问答系统若干关键技术研究[D]. 上海:华东师范大学, 2016:14-18. [10] 费鹏. 用户画像构建技术研究[D]. 大连:大连理工大学, 2017:2-5. [11] 王斌,朱建军,谭雪. 智能推荐系统在铁路客运延伸服务中应用研究[J]. 铁路计算机应用, 2019, 28(3):33-36.
点击查看大图
下载:
