融合NER和Apriori算法的游记文本关联知识挖掘及推荐服务研究

Research on Travel Text Association Knowledge Mining and Recommendation Service Based on Fusion of Named Entity Recognition and Apriori Algorithm

[目的/意义]为挖掘旅游平台游记文本蕴含知识,协助旅游者高效获取符合需求的信息和知识,为制定旅游计划提供科学决策的信息支持。[方法/过程]首先面向用户需求提出基于布尔矩阵和集合逻辑改进Apriori算法的思路;然后融合命名实体识别实现了游记文本关联知识挖掘及聚合,构建了基于关联知识挖掘的个性化推荐服务模式,并以携程网中杭州相关的游记文本进行了实证研究。[结果/结论]研究发现,融合命名实体识别和改进的Apriori算法,能够有效挖掘游记文本蕴含知识,实验结果验证了在算法性能及结果上要优于传统Apriori算法,并能够根据挖掘结果向用户提供个性化推荐服务,协助旅行者科学、高效地制定旅游计划。研究结论丰富了游记文本挖掘的方法论,为旅游平台服务推荐优化提供了新的思路。

[Purpose/Significance]In order to explore the knowledge contained in tourism platform travelogue texts and assist travelers in efficiently obtaining information and knowledge that meets their needs,this study aims to provide information support for scientific decision-making in travel planning.[Methods/Process]Firstly,based on Boolean matrix and set logic,the study proposed an improved Apriori algorithm to meet user needs.Then,integrated a named entity recognition to implement travelogue text association knowledge mining and aggregation.Based on association knowledge mining,a personalized recommendation service model was constructed,and empirical research was conducted on Hangzhou-related travelogue texts in Ctrip.[Results/Conclusions]The study finds that the integration of named entity recognition and improved Apriori algorithm can effectively mine the knowledge contained in travelogue texts.The experimental results verifies that the performance and mining results are better than traditional Apriori algorithm.Based on the mining results,personalizes recommendation services can be provided to users,assisting travelers in scientifically and efficiently planning their trips.The research enriches the methodology of travelogue text mining and provides new ideas for optimizing tourism platform service recommendations.