The Design Of Characteristic Tourist Route Of Hunan Province Based On Latent Dirichlet Allocation Model

In recent years,with the development of the economy and the material level,the pursuit of material levels has gradually faded away from the focus of life,and spiritual culture has gradually entered people's life.Tourism is a typical way of enjoying spiritual culture.It is already part of the daily life of most people to enjoy freedom and share joy during the journey.With the further upgrading of the concept of consumption,the tourism market has also quietly changed,and individualization and quality have become new competition points,which have been valued by the tourism industry.There are many tourist attractions in Hunan province,so Hunan province have absolute advantages in tourism resources.At the same time,the infrastructure of Hunan tourism area is perfect and the transportation is convenient.These are indispensable conditions for the development of tourism in Hunan.However,there are still some problems that restrict the development of tourism in Hunan Province.Firstly,the tourism brand has no recognition and own characteristics.Secondly,the tourism development between regions is uneven,the core areas,such as Changsha and Zhangjiajie,whose development of tourism is significantly better than other margin cities.The purpose of this thesis is to rationally integrating the resources of tourist attractions in Hunan Province,and to driving the development of marginal cities.Chosing 91 3A-level scenic spots in Hunan Province and getting their base information as sample data to analyze the resources of the attraction.Firstly,the text preprocessing operations such as text segmentation are carried out on 91 text samples.Secondly,based on the extracted high frequency vocabulary,analyzing the overall characteristics of Hunan province attractions.Next,after the text representation based on the vector space model,the unstructured text had represented as structured text data,which can recognized by the computer.Then,after K-Means clustering based on the vector space model,this thesis has combined with the characteristics of the algorithm of text clustering and analyzed the reasons for the poor clustering effect.Based on this,this thesis has tried to do KMeans clustering based on Latent Dirichlet Allocation,and to compare the effect of two clustering methods,the results show that the text clustering based on the LDA model is better than traditional clustering method.Finally,nased on the resule of cluster,the 91 scenic spots were divided into 5 categories.Based on the clustering results,the geographical distribution of various scenic spots in Hunan is analyzed,and the characteristics of tourism attractions in Hunan Province are further analyzed,The geographical location of each scenic spot is combined,the scenic resources are fully integrated,and the clustering result label is used as the line theme to achieve regional linkage.Adhering to the tenet of driving the core cities to develop the edge cities,we have designed five special tourist routes.