Abstract:
As an important part of ecosystem, cultural ecosystem services (CES) are considered as the core of ecosystem protection and human well-being.However, since its characteristic of subjectivity, it is often difficult to evaluate CES quantitatively.The types and differences of cultural services in each urban park often recorded in social media comment data, and can be used as a data source for ecosystem cultural service perception. Based on the web crawler, the comment data of urban parks in the main urban area of Xian city was acquired. With the help of natural language processing, topic model was carried out to mine the potential themes in the data. Based on the probability of text topic, the cultural services of urban parks are classified, so as to analyze the cultural services of ecosystem. The results indicated that the cultural ecosystem services of urban parks can be divided into five types: aesthetic, cultural, education, recreation and therapeutic. Cultural ecosystem services show great diversity in the spatial dimension, and the service categories of some urban parks are similar. The scale and characteristics of urban parks determine the cultural ecosystem service perception level of the tourists. There is a gap in the number of comments in different urban parks, and there is a difference in popularity in different urban parks.
