On the self-similarity of intertextual structures in Wikipedia

  • Authors:
  • Alexander Mehler;Christian Stegbauer

  • Affiliations:
  • Goethe University Frankfurt, Frankfurt am Main, Germany;Goethe University Frankfurt, Frankfurt am Main, Germany

  • Venue:
  • Proceedings of the First ACM International Workshop on Hot Topics on Interdisciplinary Social Networks Research
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

In recent years, Wikipedia and other wikis (e.g., Wiktionary) have been explored as knowledge resources in natural language processing (NLP) and related disciplines to overcome the bottleneck problem of knowledge acquisition (see [10] for an overview). These approaches can be said to assume a sort of emergent semantics [1] according to which meaning constitution is a self-organized process of distributed cognition [5] among large groups of interacting agents who collectively generate and structure certain fields of knowledge [17]. Emergent semantics starts from a notion of many-to-many communication in which groups of agents locally interact to collectively shape and interpret a dynamically growing knowledge area. The output of this process is emergent in the sense of being adaptive in a self-organized manner instead of being predefined by system designers [17]. Because of its organization according to the wiki principle, rapid updates and encyclopedic broadness, Wikipedia is a prototypical source of building NLP systems based on emergent semantics. This is exemplified by the seminal work on Explicit Semantic Analysis (ESA) [3, 4] that explores Wikipedia articles as concepts in a semantic space to which linguistic signs (words, texts etc.) are mapped to get representations of their meaning. These meaning representations are emergent in the sense that they change with the underlying Wikipedia.