The KRAKEN toolset is a comprehensive interface for knowledge acquisition that operates in conjunction with the Cyc knowledge base. The KRAKEN system is designed to allow subject-matter experts to make meaningful additions to an existing knowledge base, without the benefit of training in the areas of artificial intelligence, ontology development, or logical representation. Users interact with KRAKEN via a natural-language interface, which translates back and forth between English and the KB's logical representation language. A variety of specialized tools are available to guide users through the process of creating new concepts, stating facts about those concepts, and querying the knowledge base. KRAKEN has undergone two independent performance evaluations. In this paper we describe the general structure and several of the features of KRAKEN, focussing on key aspects of its functionality in light of the specific knowledge-formation and acquisition challenges they are intended to address.
[1]
Anthony R. Davis,et al.
Building and Maintaining a Semantically Adequate Lexicon Using Cyc
,
1999
.
[2]
Evelyne Viegas,et al.
Breadth and depth of semantic lexicons
,
1999
.
[3]
Nicola Guarino,et al.
Understanding top-level ontological distinctions
,
2001,
OIS@IJCAI.
[4]
Paul R. Cohen,et al.
Does Prior Knowledge Facilitate the Development of Knowledge-based Systems?
,
1999,
AAAI/IAAI.
[5]
Paul R. Cohen,et al.
Experimental Evaluation of Subject Matter Expert-oriented Knowledge Base Authoring Tools
,
2002
.