Dealing with Fake Digital Personas: One particular issue in policing online social networks is the ease in which users can create fake profiles (digital personas) and pretend to be somebody they are not. There are a variety of reasons why people choose to use fake profiles online, ranging from perfectly innocent motives such as identity exploration, to deception in order to commit serious crimes, e.g. adults masquerading as children for the purpose of grooming. Using Natural Language Processing and Machine Learning techniques, author profiling and authorship attribution tools have been developed which are able to compare the language use of different digital personas. It will be shown how, using these methods, it is possible to indicate when the same person may be behind multiple digital personas, and predict key demographic information about a user (such as age and gender) with high accuracy.