Alternatives to Author-centric Knowledge Organization

Rama C. Hoetzlein

Abstract


The digital era of the document begins with the hyper-textural, yet still author-centric model of HTML. Since then, we find both the migration of large-scale traditional document collections to online resources (Siemens 2006), and the development of community-driven resources such as Wikipedia. In these secondary resources, the authorship of meta-data and factual reference knowledge is extended to a larger, global community. However, overlapping opinion often leads to issues of authoritative control (Viegas 2004). While the traditional document is typically an individual, or small-scale collaboration, the social aspect of the document becomes much more critical in global online systems. Different strategies are needed to maintain or track authorship in documents, or knowledge resources, with thousands of authors.

Quanta is presented as an alternative to author-centric documents using a custom built, non-relational database of sentences. While encyclopedias have traditionally represented facts as a body of words in an article, Quanta represents knowledge on an atomic level, explicitly maintaining each word. This granularity introduces new ways to represent and interact with a text, or a collection of facts, and presents the possibility of assigning authorship to individual sentences.

By taking the sentence as the unit of knowledge, and representing it in machine-readable form, written language is connected to the database in ways similar to semantic networks (extended in this work with hypergraphs). All words in a sentence are hyperlinked to the text, and to larger bodies of knowledge. As Quanta makes several assumptions about language, it may be viewed as a supplement to the document, yet at a much deeper level than meta-data. By introducing filtering and visualization tools, an atomic description of language allows for overlapping, related, and even inconsistent, distinctly authored facts to co-exist in a single system.

Keywords


Quanta, knowledge environment, representation of text, authorship, information studies, encyclopedias, language-based systems

Full Text:

PDF HTML


New Knowledge Environments
© University of Victoria