An Approach for the Automatic Classification of Work Descriptions in Construction Projects

Project success is directly related to reliable access to accurate project data. During the construction project life cycle, an enormous amount of documents with largely interesting information are exchanged, which makes the decision-making process very complex to project managers. One of the most important documents in this context is the Bill of Quantities BoQ, which is a collection of Work Descriptions describing the nature of the different works needed to be done to achieve the project goal. As BoQs are differently built, practitioners have trouble when trying to insert them in a common database storage system, because both the linguistic descriptions and the structure of each BoQ document may be defined in a different way. In this article, we focus on this problem and present an approach to automatically extract the information from the work descriptions and to appropriately place it in a predefined hierarchical structure, independently of its original format. Our approach is based on a multicriteria aggregation process based on the use of learned information and expert knowledge. The methodology has been validated with a broad set of work descriptions. Results show that the proposal is able to classify them with success.