WebKB: a Manageable Approach for Knowledge Repositories


The problem. Only knowledge bases may store all sorts of facts, rules and categories, and permit knowledge inferencing, sharing and retrieval. Databases can only store predefined kinds of information, and the semantic content of free or structured documents cannot yet be extracted automatically (hence, precise/semantic IR systems, e.g. question-answering systems, cannot use documents as input). However, representing knowledge is a difficult manual task that requires training.

The solution. Large-scale knowledge bases, first initialized with default knowledge and constraints, can guide the entering of knowledge representations, and cross-link and cross-check them for correctness, precision, completeness or cooperation purposes.

The advantages of WebKB. Our knowledge base server can exploit very large KBs, has been initialized with the biggest general KB after CYC (e.g. WordNet 1.7, TAP and many top-level ontologies are integrated), has intuitive expressive and high-level input/output formats (but may also use the RDF/XML format), can exploit the KB to generate forms-to-fill to guide and ease knowledge entering, and is the only system permitting users to update a shared KB without lexical/semantic conflicts nor redundancies, and without obliging the users to agree with each other. The main WebKB server (accessible at www.webkb.org) can be freely used by any person or Web agent.

Kinds of applications. WebKB is useful when information of various kinds need to be interlinked or compared to each other, or retrieved in a question-answering system. This is the case for some corporate memories, structured catalogs, or more generally for "knowledge" repositories. WebKB is suited as a support for brokering systems that should permit people to provide information, comment on it or refine it. As a support for national/domain Yellow-Pages where people and firms would represent their products or services, or comment on products they have used, WebKB would be a particularly useful tool to store and organize the various information, and permit people to find and compare products.

Reference. Ph. Martin. Knowledge Representation, Sharing and Retrieval on the Web.
Chapter of a book titled "Web Intelligence", Springer-Verlag (2002). http://www.webkb.org/doc/papers/wi02/