Socialsfy: Supporting Citizen Science In Social Network Studies

This paper presents a system capable of classifying posts from social media platforms like Twitter, Instagram, and Facebook. It was developed to support citizen science initiatives involving humans categorizing social media posts using mobile devices. These initiatives use the crowd's knowledge approach to classify data gathered from social networks. This can then be used to study the impact of those networks on society, namely on subjects such as social media politics, misinformation, and privacy. The availability of humanly classified data is also key to training machine learning algorithms for the same purposes, which is the ultimate purpose of the underlying research initiative. The system consists of a web application for researchers to set up studies and a mobile application for citizens to classify social media posts within the scope of those studies. It supports gamification strategies to encourage citizen participation. The system's requirements, architecture, and deployment are described in the paper. The strategy for its future validation is also described.