This paper mainly describes the Bots and Gender Profiling task. Specifically, given a Twitter information, determine whether its author is a social bot or a human. In case of human, identify her/his gender. On the whole, this paper regards this as a classification task, using the RoBERTa pre-training language model to extract the emotional semantic features of the tweet, and identify the duplication rate of its content, and combine a variety of statistical features (including words+character n-grams, emoji and the forwarding rate of tweets), jointly recognize the human and bot categories.