Research On Smart Tourism Based On Data Mining

With the development of tourism,smart tourism is the inevitable trend of future tourism development.The construction of smart tourism should be user-oriented and fully consider the real scenes faced by tourists,such as ticket booking,quick access and intelligent navigation.Understanding the satisfaction of tourists with these factors can better accelerate the construction of smart tourism in scenic spots,and help scenic spots to understand the personalized needs of tourists,facilitate scenic spots to make timely improvements,and provide tourists with efficient and accurate services,thereby further stimulating consumption.Therefore,this thesis first establishes a prediction model for the passenger flow of Jiuzhaigou scenic spot based on the online search behavior of tourists and other factors affecting tourists ’ travel,so as to help the scenic spot to do a good job in the early warning of tourist flow.Secondly,based on the online comments of tourists,this thesis analyzes the tourist satisfaction of Jiuzhaigou scenic spot,and provides empirical suggestions for the scenic spot to improve the tourist satisfaction.Firstly,this thesis studies the prediction of passenger flow in Jiuzhaigou.According to other literature,the Baidu search index of tourists and the external factors affecting the passenger flow of Jiuzhaigou,such as holidays and weather,are used as the impact indicators of passenger flow prediction.Firstly,the preprocessing of Baidu search index includes : selecting and expanding the initial keywords,and using the time difference correlation analysis method and the random forest importance ranking to screen out the final keywords that have a predictive effect on the passenger flow of Jiuzhaigou.The search index of the final keywords is synthesized into a comprehensive index according to the weight.Secondly,descriptive statistical analysis is made on other influencing factors related to Jiuzhaigou tourism prediction.Finally,the gradient boosting regression tree model,XGBoost model and Light GBM model are used to predict and analyze the passenger flow of Jiuzhaigou,and the grid search algorithm is used to optimize the parameters of the prediction model.The results show that the optimized model has better prediction effect than the default model.While establishing the prediction model,the importance of variables is sorted.The results show that the variables of network search index,temperature and month are important for passenger flow prediction.Jiuzhaigou scenic spot should pay attention to these variables in time,do a good job in early warning of peak passenger flow in scenic spots and improve the corresponding preferential intensity in off-season.Secondly,this thesis studies tourist satisfaction based on online reviews.Firstly,based on the crawled online comment data of the same-course tourism Jiuzhaigou,the online comment data are labeled positive,medium and negative according to the positive,medium and bad reviews,and the online comment data are preprocessed by removing emoticons,word segmentation and stop words.Secondly,descriptive statistics are made on the comment data,including word frequency statistical analysis,word cloud visualization and semantic network analysis.Finally,the LDA topic analysis model is established,and the optimal number of topics is determined by visualization method.According to the topic feature words under each topic,the topics are summarized as : scenic spot service management,scenic spot scenery,and ticket purchase.Then,the IPA analysis model is established by calculating the emotional score and importance of each topic,and the tourist satisfaction is further studied according to the four quadrant diagram of the IPA model.The results show that tourists are satisfied with Jiuzhaigou ticket service and scenic spot scenery,but not satisfied with scenic spot service management.Therefore,it is suggested that Jiuzhaigou scenic spot should improve scenic spot service management to improve tourists ’ satisfaction,continue to maintain the advantages of ticket service,do a good job in natural disaster warning,and protect the natural scenery of scenic spots.In summary,this thesis takes Jiuzhaigou tourism data as the research object.Firstly,it uses Baidu search data of tourists combined with other external factors data that affect Jiuzhaigou passenger flow,such as holidays、weather,to establish a prediction model for Jiuzhaigou passenger flow and rank the importance of variables.Secondly,the tourist satisfaction of Jiuzhaigou scenic spot is studied and analyzed by using the online comment data of tourists.Through the comprehensive analysis of the above two aspects,this paper provides feasible suggestions for the intelligent development of Jiuzhaigou scenic spot,including : paying attention to network publicity,doing a good job in the prediction and early warning of tourist flow in scenic spots,paying attention to the relevant online comments of Jiuzhaigou scenic spot to improve tourist satisfaction.