A rule induction-based knowledge system for retaining wall selection

摘要

Knowledge acquisition is the chief constraint in constructing a rule-based knowledge system (RBKS). Knowing how to use the automatic method to elicit required knowledge is a key issue in implementing a RBKS in the experience-oriented construction domain. Rule induction (RI) is one of the fastest means of extracting rule-based knowledge from previous cases. This paper presents a model (AMASS) that integrates a RI approach, a RBKS, and a database management system to support automated knowledge mining and usage. A typical experience-oriented problem in construction domain, selecting a suitable retaining wall system in construction planning stage, is employed to demonstrate how to implement the AMASS model.