Research On The Construction Of Tourism Knowledge Graph And Information Service Based On User Generated Content

With the rapid development of the tourism industry in recent years,tourism information services are facing problems such as weak levels of service content and lack of knowledge extraction.Tourism information services urgently need a large amount of information to meet user personalized and diversified service needs.Due to the popularity of the tourism content platform,tourists can publish a large amount of content on the tourism content platform.The tourism user-generated content covers information on multiple aspects of tourism activities such as "food,lodging,travel,and entertainment",and can provide data support for tourism information services.However,user-generated content contains unstructured data such as text and pictures.Traditional database technologies cannot effectively organize this information,nor can they extract valuable knowledge from mass information.As a knowledge organization technology,knowledge graph can describe various entities existing in the real world,as well as the relationships between entities.A large number of knowledge triples are connected to form a large knowledge network.The use of machine learning,deep learning and other technologies can carry out knowledge fusion and reasoning,thereby expanding the knowledge network and mining more relationships between entities.Based on knowledge,tourism information services can provide richer and deeper services.Based on the analysis of the existing tourism information service problems,this paper constructs a tourism information service framework based on knowledge graph.The framework includes five layers: data layer,knowledge extraction layer,knowledge construction layer,knowledge fusion layer and service application layer.The difficulty of implementing the whole framework lies in how to construct the tourism UGC knowledge graph and integrate the tourism UGC knowledge graph and general knowledge graph.Based on the actual needs of tourism UGC organizations,this paper first proposes a tourism UGC knowledge graph construction model.The model includes four steps: data acquisition,knowledge extraction,ontology construction and related links.Then,through in-depth analysis of the entity characteristics of the UGC knowledge graph and the general knowledge graph,it is found that the fusion of these two types of knowledge graphs requires the simultaneous use of many features such as entity structure information,text description,and attribute values to find equivalent entities,so this paper proposes a JASE network to achieve the fusion of UGC knowledge graph and general knowledge graph,and prove the superiority of the method through comparative experiments.Finally,we developed a service platform based on the constructed tourism knowledge graph using Baidu Map API.The case on the platform can highlight the advantages of the convenience and diversity of tourism information services based on knowledge graphs.