Calculation Method Of Hot-toponym And Its Co-occurrence In Text Written By Tourists

Tourism destination spatial structure study is an important research content in tourism geography.With the development of computer and internet,the development of the tourism destination is no longer confined to a relatively closed environment.Research on spatial structure is not limited to the official data and the analysis of location factors.The destination data which can be taken advantage of has increasingly rich because of the birth of the “crowdsourcing” model.The location based service including weibo,check-in,photos,point of interest has enriched the research for more and more researchers.In this paper,basic data is the crowdsourcing tourist text.The research purpose is to build a complete set of ideas and methods.It is possible to explore the spatial structure of tourism destination in the tourism text collection.Taking Jiuzhaigou as an example to use the proposed ideas and methods and packaged application tools,completed from data preparation,process design to a complete process of classification analysis,the results prove the feasibility and value method.The main research ideas and methods presented in this paper are as follows.In the calculation method,based on the problem that a large number of temporal and spatial behavior information of tourist has to mine contains in crowdsourcing tourism texts,based on the previous work,I put forward the innovation of weight allocation,frequency and co-occurrence matrix combination calculation method of toponym in text set based on social network analysis.And then I join the network analysis methods in the study of city to dig crowdsourcing geographic information in tourism texts.On the train of thought,this paper makes a systematic study on the spatial structure information of tourism destination,and puts forward the research on the application of tourism text to complete the internal relationship of tourism destination and its relationship with the outside.The calculation results were stored in the form of co-occurrence matrixes and triples,which made extracting the subset and was based on type of place names very convenient.The results of the analysis achieve the combination of statistical charts,tables and visualization,try to conduct a scientific way of performance,as far as possible to make data and information to effective knowledge,and try to play the greatest value of tourism text research.Taking Jiuzhaigou as a research destination to do empirical research,this paper makes the realization of the automatic extraction of the toponym of tourist attractions in Jiuzhaigou,the quantification and the visualization of the co-occurrence relationship between the toponym and the specific landscapes.Study on the connection mode of Jiuzhaigou with outside realized three kinds of co-occurrence relationship visualization.It is concluded that the toponym and their co-occurrence can clearly demonstrate the cognitive structure of the group of authors to the relationship between the internal and external space of Jiuzhaigou tourism destination.For example,the contact strength between Jiuzhaigou tourism and the tourist city of inside or outside the province such as Shanghai,Chengdu,Beijing,the contact Jiuzhaigou with the city that have the support services function such as Songpan,Maoxian,and Wenchuan,and the relation characteristics with the tourism in Sichuan province such as HuangLong,Dujiangyan,Mountain Emei etc..Jiuzhaigou internal contact models research implemented the overall and layered analysis of all kinds of landscape.In conclusion,the network connection degree of landscape and landscape attention level presents three models,but totality present a certain positive correlation.Jiuzhaigou exists obvious characteristics in internal partitions,classification,agglomeration.There is a difference among the three scenic region;The development of the landscape types is also different;Space exists the core guide and agglomeration phenomenon,specifically embodies three core zone,two big group.The results showed that this method could meet different research needs of tourism geography and related disciplines,and the crowdsourcing tourism text data sources were particularly suitable for mining spatial perception characteristics and structure the specific tourist groups.