Construction And Application Of Knowledge Graph Of Tourist Attractions In Hebei Province

In the environment of internet and big data,search engine has gradually become the main way for people to understand the world.But with the explosion of data in the tourism sector,it often takes a lot of time and effort to sift through it several times to get the information they need,therefore,it is very meaningful to provide a complete system that can help the user to understand the scenic spot information quickly and accurately.knowledge graph can greatly improve the efficiency of search engine.Therefore,according to the demand of tourism industry in Hebei Province,this paper constructs the knowledge graph of tourist attractions by integrating multi-source heterogeneous data,on this basis,the information visualization system based on knowledge graph is designed and implemented.The main work is reflected in the following areas:(1)Acquisition and preprocessing of corpus.First of all,through the analysis and comparison of relevant websites,select the data source website;then combine phenomenjs and selenium to build a crawler framework,and achieve the crawling of the initial corpus through the framework;finally,use nlpir tool to realize the preprocessing work of Chinese word segmentation and part of speech tagging of the initial corpus.(2)Research on the construction of knowledge graph of tourism attractions in Hebei Province.Firstly,the Bi LSMT-CRF named entity recognition model and SVM relationship prediction model are trained to realize the extraction of scenic spot attribute and attribute relationship;then,Levenshtein Distance algorithm and word2 vec model are used to fuse attributes by calculating the similarity of attributes,and the attribute values are fused by learning sorting method and multi value merging method;and then,according to the actual characteristics of Hebei Province's tourist attractions,the system is implemented an improved ontology construction method is proposed,that is,the circulation idea is integrated into the seven step method,and the iterative updating process is added.According to the improved seven step method,the ontology construction is realized in protégé ontology editing software;finally,all knowledge is imported into neo4 j map database for storage to complete the construction of knowledge graph of tourist attractions.(3)The application of knowledge graph of tourist attractions.Based on theconstructed knowledge graph,this paper designs and implements an application system integrating scenic spot information management,scenic spot information search and search result visualization.The system is directly connected with the map database,which can quickly and efficiently complete the scenic spot information search and the operation of adding,deleting,modifying and checking the scenic spot information;and through D3 technology,the scenic spot information in the knowledge graph can be fully and intuitively displayed to users.This study will help users search information more efficiently and quickly,and understand the scenic spots directly and concisely,and then promote the development of tourism in Hebei Province.It is also helpful to the further research of domain knowledge graph construction.