A preliminary approach to classify work descriptions in construction projects

Construction projects have associated big amounts of, usually, unstructured data generated from several sources during the construction process. To be able to store these data in a structured database is a desired goal, for example, to assist project managers in the task of making better decisions based on these data. One of the most important document in this context is the bill of quantities, which is a collection of work descriptions grouped into chapters and sub-chapters. The use of natural language in the descriptions makes bill of quantities be irregular and present ambiguity despite using a highly technical language. In this paper, as a first step in the task of mapping bill of quantities to a structured schemata, we present a preliminary approach to automatically classify Work Descriptions extracted from bill of quantities. Different predefined prices databases are used to build a hierarchical knowledge structure together with expert knowledge and then, a methodology based on the use of ordered weighted aggregation operators is proposed to accomplish the classification task.