The text analysis method is used to study the tourist review text of Ctrip's Chengdu Research Base of Giant Panda Breeding. ROST ContentMining6 and UCINET 6 are used to conduct semantic network analysis, centrality analysis and emotion analysis on tourist online reviews. The research results are as follows: Social Semantic Network analysis "panda" and "Huahua" constitute the first-level core high-frequency words, "queue", "cute" and "Chengdu" are the second-level core words, and are interrelated with the first-level words. The positive emotion accounted for 75.15%, the neutral emotion accounted for 12.13%, and the negative emotion accounted for 12.72%.