Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!

Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


11.7. EVALUATION

A framework for evaluating knowledge-sharing technology (software, ontologies, and documentation) has been presented by Gómez-Pérez (1994). Evaluation means to carry out a technical judgment of the ontologies, their software environment, and documentation with respect to a frame of reference (in our case, the requirements specification document) during each phase and between phases of their life cycle. Evaluation subsumes the terms Verification and Validation. Verification refers to the technical process that guarantees the correctness of an ontology, its associated software environments, and documentation with respect to a frame of reference during each phase and between phases of their life cycle. Validation guarantees that the ontologies, the software environment, and documentation correspond to the system that they are supposed to represent. Based on the experience of verifying Ontolingua ontologies, a set of guidelines and methods of how to look for incompleteness, inconsistencies, and redundancies have been also presented (Gómez-Pérez, 1996).

The output proposed by METHONTOLOGY for this phase is an evaluation document, in which the ontologist will describe how the ontology has been evaluated, the techniques used, the kind of errors found in each phase, and the sources of knowledge used in the evaluation.

11.8. DOCUMENTATION

There are not widely accepted guidelines on how to document ontologies. In many cases, the only documentation available is in the code of the ontology, the natural language text attached to formal definitions, and papers published in conference proceedings and journals setting out important questions of the ontology already built. This problem is the result of a vicious circle: almost anyone documents ontologies, as there are no guidelines on documentation, there are no guidelines on how to document ontologies because of the absence of methodologies to build ontologies, and there are no standard methodologies to build ontologies because ontologists do not write the steps they take to build ontologies during the entire ontology development process.

METHONTOLOGY seeks to break this circle by including the documentation phase as an activity to be carried out throughout the entire ontology development process. Indeed, after the specification phase, you get a requirements specification document; after the knowledge acquisition phase, a knowledge acquisition document; after conceptualization, a conceptual model document that includes a set of intermediate representations that describe the application domain; after integration, an integration document; after implementation, the implementation document; and during evaluation, an evaluation document.

12. INTERLINGUAS

This section presents two interlinguas (KIF and PIF) used to interchange knowledge and processes between heterogeneous and independent applications.

12.1. KIF

One of the work groups created by the ARPA Knowledge-Sharing Effort (Neches et al., 1991) was the Interlingua Working Group at Stanford, led by Fikes and Genesereth. Their goal was to solve the problem of the heterogeneity of knowledge representation languages. In order to interchange knowledge between heterogeneous programs, they realized that a formal language was needed, which, like an interlingua, allowed knowledge in a given representation language to be expressed in another. The interlingua had to:

  1. Be a language with declarative semantics and independent of any interpreter.
  2. Be a language with sufficient expressive power to represent the declarative knowledge contained in typical applications system knowledge bases.
  3. Have a structure that enabled semiautomatic translations into and out of typical representation languages.

The result was KIF (Genesereth and Fikes, 1992), a prefix version of first-order predicate calculus, with extensions to improve its expressiveness, such as: definition of terms, representation of knowledge about knowledge, reifying functions and relations, specifying sets, and nonmonotonic reasoning.

The basis for KIF semantics is a correlation between the terms and sentences of the language and a conceptualization of the world. Every term denotes an object in the universe of discourse associated with the conceptualization, and every sentence is either true or false. The semantics of KIF tell us the meaning of its complex expressions. We can unambiguously determine the referent of any term and the truth or falsity of any sentence.

The world is conceptualized in KIF in different universes of discourse. In fact, different users have different universes of discourse. Each universe of discourse is the set of all objects presumed or hypothesized to exist in the world. Objects can be concrete or abstract, primitive or composite, real or fictional. Relationships between objects are conceptualized by functions and relations of variable arity. Conceptually, relations and functions are sets of finite lists of objects. The difference between them is that the function associates a unique object (called value) for each combination of possible arguments, and the relation does not.

KIF also provides a standard vocabulary for dealing with: lists (listof, single, first, rest, last, etc.); sets (set, individual, union, intersection, etc.); numerical calculus and properties of numbers (cos, expt, log, integer, >, etc.); logical operations (negation, conjunction, disjunction, equation, inequality, etc.), rules, constraints, quantified expressions, metarules, programs, etc. More information about KIF can be found at http://logic.stanford.edu/kif.html.


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.