Objective: To understand the emotion state and emotion trajectory of adolescents during the transition stage from high school to college by analyzing text on social network platform, and to provide reference for carring out mental education for adolescents. Methods: From September to November 2018 and September to October 2019, 534 undergraduates were recruited through online and offline advertisement, and their Tencent Tweets posted during the period from last year of high school to first year of college were crawled and stored with the time stamps. Total number of words, the number of Tencent Tweets, frequency of positive emotion words and negative emotion words were obtained by using jieba (a word segmentation program) and an emotion dictionary. The degree of activity in Tencent Tweets during the transition stage were described. The trend of the frequency of negative and positive emotion words were analyzed by using hierarchical linear model. The degree of activity and trend of monthly variation between high school and colleges were also compared. Results: After entering college, adolescents degree of activity in Tencent Tweets were much higher. Total number of words, number of Tencent Tweets, number of positive emotion words and negative emotion words were significantly higher than those in high school (Wilcoxon Z =-18.93, -18.03, -18.40, -13.60, P <0.01). Total number of words, number of Tencent Tweets, number of positive emotion words and negative emotion words posted by female adolescents were significantly higher than male adolescents during first year of college (Mann-Whitney U =26 499.00, 28 581.00, 24 636.00, 28 010.50 , P <0.05). There were more positive emotion words than negative emotion words in adolescents Tencent Tweets, the monthly average frequency of positive emotion words (0.029) was 2.40 times of frequency of negative emotion words (0.012). The frequency of positive emotion words and negative emotion words decreased significantly with each month in the last year of high school ( β =-6.01, 3.55; t=-3.40, -2.86, P <0.01). Conclusion Adolescents were less active in using Tencent Tweets in last year of high school, but their positive and negative emotional fluctuation were more intense in Tencent Tweets. It is noteworthy to pay attention to the emotional state of adolescents in last year of high school.