基于NLP和知识图谱的旅游线路智能推荐系统构建

指导教师：牛宇　讲师

作　者：王莉漫

院（系）：旅游科学学院

专　业：电子商务（营销）

学　号：2015221137

完成时间：2019年5月7日

内容摘要(www.xing528.com)

目前旅游线路规划存在过度依赖经验和人力，且工作重复化与不规范化给效率提升带来瓶颈，也难于满足逐渐复杂的用户需求。本文采用基于知识图谱+对话机器人的方法应用在旅游线路规划领域，不仅可以实现个性化推荐，还可以对接到例如携程的资源接口，根据实时的车票、机票、酒店查询，帮助用户计算出费用供参考，并直接辅助用户下单。文章以四川省旅游线路规划为例，着重于对旅游知识图谱的构建提出新的思路，以及如何把知识图谱与对话机器人对接，实现一个前端对话中既懂用户，后端又有庞大数据库支撑的旅游线路智能推荐系统。

关键词：知识图谱；自动问答系统；旅游线路推荐；自然语言处理

Abstract

At present，tourism route planning relies too much on experience and manpower，and work repetition and non-standardization bring bottlenecks to effi ciency improvement，and it is diffi cult to meet the increasingly complex user needs.This paper uses the method of knowledge map + dialogue robot to apply in the fi eld of tourism route planning，which can not only realize personalized recommendation，but also provide resource interfaces to such as Ctrip.Real-time ticket，ticket，Hotel inquiries，help users calculate the cost for reference，and directly assist users to place orders.Taking the tourism route planning of Sichuan Province as an example，this paper puts forward new ideas for the construction of tourism knowledge map，and how to connect the knowledge map with the dialogue robot to realize an intelligent recommendation system for tourism routes which understands both users and huge databases in the front-end dialogue.

Key words：Knowledge Graph，Automatic Question and Answering System，Travel Route Recommendation，Natural Language Processing