基于混合主题语义特征提取的旅游吸引物地理画像

Geographic Portrait of Urban Tourism Attractions Based on Hybrid Theme Semantic Feature Extraction

通过刻画城市旅游吸引物地理画像,可为城市旅游吸引物建设和游客出行提供参考,但现有相关研究较少考虑到旅游吸引物的空间位置和配套设施等因素。该文提出从地理空间和地理语义两个角度构建城市旅游吸引物地理画像的方法,首先基于兴趣点(POI)数据通过隐式狄利克雷分布(LDA)模型提取主题语义特征,然后通过空间聚类分析旅游吸引物的空间格局,最后计算旅游吸引物集群的主题多样性,据此衡量游客对不同旅游目的地的满足度。以武汉市为研究区进行实验,结果表明:该文方法能客观地刻画旅游吸引物的配套设施建设情况,定量计算旅游吸引物集群的主题混合度和优势度,很好地融合了地理语义特征和地理空间关系,可为景区规划建设和游客选择旅游目的地提供参考。

The geographic portrait of urban tourism attractions can provide references for the construction of urban tourism attractions and tourists′ travel,but the geographic factors such as the spatial location and supporting facilities of tourism attractions are seldom taken into account in existing related works.This paper proposes a method to construct the geographical portrait of urban tourism attractions from both geospatial and geosemantic perspectives.Firstly,thematic semantic features are extracted based on point of interest(POI) data through the latent Dirichlet allocation(LDA) model,then the spatial pattern of tourism attractions is analyzed through spatial clustering,and finally the thematic diversity of tourism attraction clusters is calculated,which can be used to measure tourists′ satisfaction with different tourist destinations.The experiment results in main urban area of Wuhan show that the proposed method can objectively portray the building of supporting facilities of tourism attractions,quantitatively calculate the thematic mixture and dominance of tourism attraction clusters,which well integrates geosemantic features and geospatial relationships.In addition,these analysis results can provide references for scenic spot planning and tourists′ selection of tourism destinations.