| In the new media era,micro-blog has become an important platform for netizens to express their personal views and share external information.For tourism micro-blog published on the platform,it is of great significance to both scenic managers and users: for general users who love tourism,it can let them quickly understand about tourism.For the official micro-blog of tourist attractions,we can understand the user’s feelings and needs for tourist attractions and help to improve and enhance the service of tourist attractions by checking other users’ travel notes after they have published their feelings and exchanges with other users through micro-blog.Shanxi Province is rich in tourism resources and has many unique resources.But compared with the rich tourism resources,Shanxi tourism industry still has a lot of room for development in management,publicity and promotion.Through the research on hot topics of tourism micro-blog and related micro-blog returns,we can fully understand the users.Only by increasing diversification and individualization of tourism needs can we develop market competition advantages with unique selling points.This paper studies the application of Shanxi Tourism micro blog as follows:First,aiming at the problem that there are few data sets in Shanxi Tourism micro blog research,this paper determines the index of crawling data according to the needs of research content,uses data crawling software to crawl the data set independently,and carries out the data preprocessing process of word segmentation,de stop words,and text modeling,and processes the data set into a form that can be recognized and processed by computer for subsequent topic discovery The clustering process of.Secondly,on the basis of fully analyzing the current research status at home and abroad and the existing application of the microblog hot topic list,this paper proposes a tourism microblog topic discovery method based on improved K-means clustering algorithm,which uses hierarchical aggregation.The class algorithm AGNES(Agglomerative Nesting)obtains the cluster center,which is the initial class center point of K-means algorithm clustering.It solves the problem that the K-means algorithm initially selects the center point of the cluster and improves the stability of the clustering result.On this basis,the K-means algorithm is used to cluster again to overcome the problem that the results of hierarchical clustering can not be corrected.Through the combination of the above two steps,similar topics in the travel microblog text can be clustered to obtain the microblog topic.Finally,a method of calculating tourism topic heat by using document-topic probability distribution is proposed.On this basis,a related microblog return method including three-layer screening process is further proposed,which combines the correlation between microblog content and hot topics,user value,and the degree of attention of microblog,and so on,to correlate tourism hot topics.Finally,combined with the research results,the status quo of Shanxi tourism official blog and tourism industry development and some hidden problems are analyzed,and some reference suggestions are given for the problems.Through the application of actual data,it is proved that the improved K-means algorithm proposed in this paper is better than other single clustering algorithms in text clustering,and the accuracy of microblog returned by the related microblog return method proposed in this paper.It is also higher than the accuracy of the existing platform.Even if the Weibo users do not have relevant prior knowledge,they do not need to browse all relevant microblogs to get a general idea of the ins and outs of the topic,and get a better reading experience with a lower reading cost. |
