Extracting Value from Automated Classification Tools: the Role of Manual Involvement and Controlled Vocabularies
http://argus-acia.com/white_papers/classification.html

Hagedorn, Kat
DLib Magazine
2001

Abstract:

Automated classification tools can't solve today's large-scale web and intranet indexing challenges alone. Neither can humans. But solutions that integrate human expertise with software products such as Interwoven's Metatagger and Autonomy's Categorizer can provide real value and savings. After a brief introduction to automated classification, this white paper discusses the benefits and limitations of manual, automated, and hybrid approaches. It explores the opportunities for leveraging controlled vocabularies and thesauri to produce more effective indexing solutions.