Automated Classification Methods: Supporting the Implementation of Pull Techniques for Information Flow Management

Carlos H. Caldas¹ & Lucio Soibelman²

Abstract

One of the requisites for implementing lean construction processes is the management of information flows through the networks of cooperating project organizations. Information flows about directives, criteria, prerequisites, availability, commitments, and resources are essential to production control and work structuring. Since a large percentage of these project information is generated in text format, methods for managing the information contained in these types of documents becomes essential to improve work flow reliability. Information management systems have been used for this purpose. One limitation of the text-based information management aspects in current systems is the reliance on push methods. Push systems schedule the release of information based on demand. On the other hand, pull systems release information based on system status. For that reason, the implementation of pull information systems is an essential requirement of lean construction delivery systems. This paper describes a methodology to support the implementation of pull techniques in construction management information systems based on automated text classification methods.

Keywords

Construction management, information flows, information management, machine learning, pull systems, text/data mining.

Files

Full Text

 

Reference

Caldas, C. H. & Soibelman, L. 2002. Automated Classification Methods: Supporting the Implementation of Pull Techniques for Information Flow Management, 10th Annual Conference of the International Group for Lean Construction , -. doi.org/

Download: BibTeX | RIS Format