Abstract: As one of the most popular Internet applications, it is relatively easy for tagging systems to become the targets of malicious attackers. In a tagging system, malicious users always annotate resources with tag spam for some reason, thereby misleading normal users from searching and sharing those resources. Existing work has already shown that the defense effect of tag spam would directly influence the quality of services provided by tagging systems. Based on this background, through investigating some existing research efforts, this paper gives the formal descriptions of attack styles, and summarizes and extracts the existing three types of methods to defend against tag spam: detection-based method, demotion-based method, and prevention-based method. As to each type of method, this paper further describes corresponding implementing algorithms in detail, and analyzes their advantages and disadvantages, such as efficiency and availability. At the end of this paper, the authors also exhibit the future research directions of defending against tag spam. According to the description of countermeasures against tag spam, this paper can help readers or beginners to get familiar with the situation of this new research field and help researchers to think out more and more novel, practical and effective countermeasures against tag spam.