Collaborative Concept Extraction from Documents.
Keiichi Nakata, Angi Voß, Marcus Juhnke, Thomas Kreifelts · 1998
A group of individuals who share the same interest or a task, would profit from making use of the knowledge possessed by the group. It is then essential that such a body of knowledge, or "community knowledge", be captured in an effective manner. This paper describes the notion of Concept Index, which aims to index important concepts described in a collection of documents belonging to a group, and provide user-friendly cross-references among them to aid concept-oriented document space navigation. Unlike approaches relying primarily on automatic concept extraction tools, the Concept Index relies on users in identifying important concepts by marking keywords and phrases that interest them. Once the concepts are extracted in this manner, they are then enhanced by automated tools and more importantly by users who inspect them. We argue that with an appropriate support for users provided by a system, this interactive process optimises the index generated, and enhances collaboration between t...