Abstract: For understanding the tourism image perception of World Cultural Heritage Sites of ancient villages, through collecting
1,704 tourists' online reviews about Hong village by ROST Content Mining software, this paper takes the methods of word frequency
analysis, emotion analysis, social network analysis and semantics analysis to extract the high- frequency words of Hong village's
tourism image, to analyze the emotional tendency and to construct the semantic network. The results showed that: visitors' tourism
image to Hong village is obviously comprehensive, and there is tourism shielding effect; visitors' tourism image to Hong village lacks of
deep-seated cultural image, and there are price sensitive visitors; visitors' tourism image to Hong village lacking of integrity, and there
is long tail phenomenon of the image perception; visitors have significant positive emotional tendency to Hong village, but being
worried about commercialization; the film and television works have little impact on the Hong village' tourism image perception .