Research On Tourist Satisfaction In Red Tourism Classic Scenic Area Of Jiangxi Province Based On Machine Learning

The report of the 20 th National Congress of the Communist Party of China pointed out that it is necessary to "make good use of red tourism resources and carry out indepth promotion and education of socialist core values".Red tourism,as a highly distinctive tourism product with Chinese characteristics,has always been valued by the country and has introduced multiple policies to promote its development.As a carrier of red tourism activities,red scenic spots play an important role in promoting the development of red tourism and inheriting red genes.The prosperity and sustainable development of red scenic spots are crucial for tourist satisfaction,which not only directly affects tourists’ willingness to revisit,but also affects their word-of-mouth dissemination.Online comments,as tourists’ true views on scenic spots,reflect their subjective impression and satisfaction with them.Therefore,using online comments as a data source to deeply explore the influencing factors of tourist satisfaction is not only conducive to the improvement and upgrading of scenic spots themselves,but also provides suggestions for consumers’ purchasing decisions.Jiangxi Province,as the first place of red tourism,has a profound revolutionary tradition and glorious history,so this study uses the network text of 11 classic red tourism scenic spots in Jiangxi Province as the data source to analyze the tourist satisfaction of red tourist attraction in Jiangxi Province.This study first used Python software to construct a crawler program to capture a total of 9943 tourist comments from two major tourism platforms,Ctrip and Mafengwo,and preprocessed the data;Secondly,construct an emotion analysis model,select five emotion analysis models: Convolutional Neural Network(CNN),Long Short Term Memory Network(LSTM),Bidirectional Long Short Term Memory Network(Bi LSTM),Bidirectional Long Short Term Memory Network fused with Convolutional Neural Network(CNN-Bi LSTM),and Bidirectional Long Short Term Memory Network based on attention mechanism(Bi LSTM ATT)for model training,and select the optimal emotion analysis model;Thirdly,the influencing factors of tourists’ satisfaction are constructed through feature word analysis,semantic network analysis and LDA theme mining;Finally,the optimal performance sentiment analysis model is used to analyze the satisfaction status of various factors,and then a fuzzy comprehensive evaluation method is used to evaluate the overall satisfaction of tourists.Research has found that:(1)By comparing the evaluation indicators of five sentiment analysis models,it was found that the CNN model has the worst classification performance,while the Bi LSTM-ATT model has the best classification performance.(2)In the analysis of feature testimony,tourists pay the most attention to the tourism resources,education,and overall evaluation of scenic spots;In the analysis of semantic network,it is found that its core node is revolution,and the words related to it are tourism resources,history,humanity and education;The LDA model excavated a total of 5 influencing factors,namely infrastructure and services,scenic area education,tourism resource endowment,historical and cultural,and overall evaluation.(3)In the satisfaction analysis of the five factors,historical and cultural satisfaction is the highest,while infrastructure and service satisfaction is the lowest.The overall evaluation of tourist satisfaction is 0.6307,indicating overall satisfaction.