Computational Methods for Web History

In light of the exponential growth in digital data characterizing the 21st century, future historians of our time will have to rely on born-digital materials as primary sources for establishing historical facts. Yet born-digital materials challenge historians’ well-established source criticism techniques used for establishing facts based on the authenticity, authorship and authority of documents, for they are ephemeral, immaterial, fragile and easy to manipulate. For example, the content of websites can be easily modified, tweets are frequently deleted, the number of social media comments and likes can be artificially boosted through click farms, and dubious sources spreading misinformation can be disguised as reliable news organizations. With the commercialization of the web, more than ever before, web data is primarily proprietary, and therefore subjected to platforms’ policies and constraints.