Abstract: The tourism image of characteristic villages is an important basis to test and measure the tourism destination of characteristic villages. Taking Wuling Mountain Area in Hunan Province as an example, this paper uses Python web crawler technology to obtain travel notes and comments from websites such as wasp nest and Ctrip, and uses Rost content mining software to analyze the text with high-frequency feature words, semantic network and emotion, so as to study the tourists' perception of the tourism image of the scenic spot.The research conclusions are as follows: (1) 162 high-frequency words are extracted from 128 travel notes, and the words ranked in the top 60 indicate that tourists focus on "village landscape", the semantic network matrix of high-frequency words takes the village as the core, and other characteristic words diverge around; (2) Tourists' cognitive image of characteristic village tourism mainly includes village image, tourism attractions, village facilities and services; (3) Tourists' perception of the village tourism image in Wuling Mountain Area of Hunan Province is mainly positive emotion, and the negative emotion mainly comes from the imperfection of the village's road traffic, health environment and other infrastructure.

Key words: characteristic village tourism, perceived image, network text analysis, Wuling Mountain Area