Author Profiling: Predicting Gender from Document

As the Internet is aging, a massive amount of data is being created on the web, out of which mostly is text. Therefore, authorship of the contents and prediction of characteristics of the author is becoming a new domain of data analytics making Author Profiling a research area with huge scope of possibilities and outcomes. The ability to describe the features or traits of an author has a key application in many security and forensic areas. The PAN labs provide a platform for scholars by organizing author profiling tasks, for example, language, gender prediction, etc. In this paper, we are attempting to predict gender of a particular author, for which we have considered English dataset of PAN 2017.