Audience-Centric Taxonomy: Using Taxonomies to Support Heterogeneous User Communities

Abstract

Controlled vocabularies enhance precision and recall but sometimes they achieve this at the expense of imposing a prescribed terminology and a homogeneous worldview upon a heterogeneous user community. Folksonomies allow end-users the freedom to describe content any way they want, but in doing so they create meta noise which diminishes precision and recall. This paper presents an alternative model called audience-centric taxonomy, which blends the best practices of top-down controlled vocabularies with the bottom-up approach of folksonomy. The result is a semantically rich and well structured vocabulary that can adapt how it presents itself to different end-user communities ensuring each audience sees the language and worldview that it prefers. The paper describes how the National Library Board, Singapore intends to utilize audience centric taxonomy to provide enhanced information access to its multilingual, multi-cultural user community.