Guilin Tourism Image Perception Based on Web Text Analysis

Abstract

Under the background of Guilin Municipal Government's proposal to build international tourism by 2020, this paper studies Guilin's tourism destination perception image, and puts forward the development direction of Guilin's tourism image in the future. In this paper, Octopus is used to customize the text review data from some popular tourism websites. A total of 14094 valid text review data are collected. Firstly, the data collected from Guilin tourism texts are pre-processed to remove duplication, word segmentation, stop words and invalid symbols. Subsequently, we use R to do word frequency statistics, draw word clouds, construct a three-tier Bayesian probability model, and use ROSTCM6 text review data to analyse emotional orientation. The results show that 78% of the tourists hold a positive attitude towards Guilin tourism. The factors affecting tourists experience are mainly divided into four aspects: natural scenery, tourism service, characteristic tourism products and commercialization degree, among which natural scenery is the most important factor. In the last, we put forward corresponding countermeasures from four aspects: increasing the propaganda of scenic spots, intensifying the development of tourism characteristic products, improving the quality of tourism service and strengthening the construction of commercial streets.