Grammatical rules for specifying information for automated product data modeling

  • Authors:

  • Ghang Lee;Charles M. Eastman;Rafael Sacks;Shankant B. Navathe

  • Affiliations:

  • College of Architecture, Georgia Institute of Technology, 245 Fourth Street, Atlanta, GA 30332-0155, USA;College of Architecture, Georgia Institute of Technology, 245 Fourth Street, Atlanta, GA 30332-0155, USA;Faculty of Civil and Environmental Engineering Technion, Israel Institute of Technology;College of Computing, Georgia Institute of Technology, Atlanta, GA 30332-0155, USA

  • Venue:
  • Advanced Engineering Informatics
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a linguistic framework for developing a formal knowledge acquisition method. The framework is intended to empower domain experts to specify information required by activities in design, engineering, manufacturing, and maintenance processes. The longer-term goal of the framework is to (semi-)automatically derive a data model from product information specified by domain experts. The framework for information specification is named the Product Information Specification (PIS) framework. The linguistic framework categorizes terms ('tokens') required to define product information into six constituents, similar to the parts of speech in grammar, based on abstraction concepts of Knowledge Representation. Syntactic rules for combining these six constituents guarantee the consistency and the analyzability (computability) of the specified product information. A Context-Free Grammar (CFG) has been adopted for analyzing and defining the rules. The applicability and feasibility of the PIS framework has been demonstrated through a research project with the North American precast concrete industry. Examples in this paper are drawn from this project. The major contribution of the PIS framework is that it provides a basis for a knowledge acquisition method that can facilitate domain experts' direct participation in product modeling, can potentially increase the quality of the resultant model and decrease the product modeling time.