Research on Strategies of Digital Media in Shaping and Communicating City Image--Text Mining and Sentiment Analysis Based on Image Communication

Shaping and spreading a good city image can effectively improve the comprehensive strength of the city, promote investment attraction, promote the development of the tourism industry, and help social civilization and harmony. The research utilizes the TextRank algorithm for text mining of city images. Construct an emotional analysis model for city images, using the calculation of emotional tendencies, to make a comprehensive evaluation of the city’s image communication. For example analysis of Hangzhou city, the overlap rate of high-frequency words between city officials and tourists on Hangzhou’s city image reaches 43.3%. The two perceptions of the imagery of Hangzhou are both cross and different, and the overall satisfaction score of tourists in Hangzhou is 3.84 points. Most of the tourists in Hangzhou commented positively, and the percentage of positive, neutral, and negative sentiment comments accounted for 81.17%, 4.50%, and 14.33% of the overall, respectively.