Automatic Classification of Construction Work Codes in Bill of Quantities of National Roadway Based on Text Analysis

Abstract

The construction work classification code is a crucial index for consistently collecting unit cost data from the bill of quantities (BOQ). Appropriate classification codes assigned to detailed construction work item descriptions are required to verify project cost and for quality control. However, the codification system is complicated and time consuming for estimators to follow. This study proposes a framework to recognize the text of detailed construction work item descriptions in the BOQ and automatically assign the most similar work classification code. The automatic assignment algorithm was designed to score the similarity of the tokenized words in the work item description based on whether they contain the same word. The framework was experimented with using the national roadway BOQ; this BOQ is used for the procurement process in South Korea. It transformed work item descriptions from unstandardized BOQs into standardized data with more than 91.12% accuracy. The results of this study can be used as a prerequisite step for standardizing construction life cycle cost data and the automatic generation of BOQs.