Estimating age and gender in instagram using face recognition: advantages, bias and issues.
Diego Couto de Las Casas · Americanae (AECID Library) · 2016
Studies in social computing often take into account personal attributes of users of specific online services in order to better understand their behavior. As these attributes are often unavailable for researchers and developers, recent efforts have been devoted to estimate them by combining other sources of information. Besides offering insights to how users relate to online platforms, such attribute estimation methodologies can also contribute to understanding how exposed is user information to third parties. In this master thesis I propose to study the use of face recognition technologies to estimate the age and the gender of the users of a popular, image-based online social network, Instagram. This approach is inspired by the increasing wealth of information from image data in online social networks, as well as recent advances in face recognition.