Inspired by its biological counterpart, we'll introduce the concept of digital DNA, as a mechanism to characterize online user behaviors on social media and the Web. We'll model online users' actions as digital DNA sequences, introducing a strikingly novel, simple, and effective approach to model and analyze different types of online accounts. The high flexibility featured by digital DNA sequences, makes this modeling technique well suited for different scenarios, with the potential to open up new directions for research. It also opens up the possibility to draw upon decades of research and development in bioinformatics. Among the possible applications, it will be presented a series of experiments in which digital DNA has been successfully used to detect Twitter spambots. In particular, a spambot detection techniques able to provide an effective mean to detect novel, evolving social spambots on Twitter. An extensive investigation, leveraging a wide experimental campaign, shows that neither Twitter, nor humans, nor cutting-edge applications are currently capable of accurately detecting such novel social spambots. Instead, this digital fingerprinting detection technique, based on the DNA-like sequentialization approach, succeeds in revealing the novel spambots, in an unsupervised fashion and only exploiting the timeline data, thus being both effective and efficient.
