Studies on Editing Patterns in Large-scale Wikis

In this chapter, we present a complementary pair of experiments on the way users edit Wikipedia. First, we look at the macro scale, quantitatively categorising the types of edits users made. Then, we look at the micro scale, qualitatively studying the rationale and process of individual editors. These experiments form part of a broader project looking into the potentially beneficial relationships between open hypermedia, the study of interconnected documents; Semantic Web, the study of interconnectable data; and ‘wikis’, web-based communal editing systems. Wikipedia is a communally-edited encyclopædia with over two and half million articles in the English version. Each article is a document with prose about an encyclopædic subject, usually supplemented with illustrations. Almost all articles are placed into at least one ad-hoc category, and linking between articles is common. Hypermedia is a long-standing field of research into the ways in which documents can expand beyond the limitations of paper, generally in terms of greater cross-referencing and composition (reuse) capability. Bush’s As We May Think [3] introduces the hypothetical early hypertext machine, the ‘memex’, and defines the “essential feature” of it as “the process of tying two items together”. This linking between documents is the common feature of hypertext systems, upon which other improvements are built. As well as simple binary (two endpoint) links, hypertext systems have been developed with features including n-ary links (multiple documents linked to multiple other documents), typed links (links which indicate something about why or how documents are related), generic links (links whose endpoints are determined by matching criteria of the document content, such as particular words), and composite documents, which are formed by combining a set of other, linked, documents. Open