Identifying Historical Primary Sources in Social Media for Reliable Historical Reconstruction

摘要

Social media contains opinions and information in an enormous quantity. With its ubiquitous role in keeping up with current affairs, social media could be used to retrospectively research historical events by treating posts as primary sources, as in History and Journalism. This paper proposes a metric system and interface design for allowing users to sort through Twitter posts. The aim is to allow users to judge social media source reliability and reconstruct details of a historical event. After participants used the sample tool, it was concluded that the framework is a beneficial starting point to formalise a computational process for sorting and visualising Twitter data based on reliability. The study also showed that the presence of URLs in Tweets (backing evidence) and the author's number of followers (reputability) were most important for discerning credibility. This project can inspire future work to integrate this framework into social media applications for data extraction, user feed generation and interface design when presenting posts to users.