The invention discloses a method for censoring and processing text contents issued by a user and a device thereof. The method comprises the following steps: receiving the text contents issued by the user and judging user information according to a list rule database; if the user information neither belongs to a white list or a white rule nor a black list or a black rule, calculating a first similarity of a first characteristic vector of the text contents of the user and a second characteristic vector of pre-established spam sample contents, and judging whether the text contents issued by the user are qualified contents according to the first similarity, if the text contents are the qualified contents, publishing the text contents issued by the user; or sending the text contents issued by the user for manual censoring. The method and the device can help censor and filter the user information and the text contents issued by the user without total manual censoring of the information issued by the user, thus greatly avoid the manual censoring time and saving the human resources and correspondingly enhancing the censoring efficiency.