Unstructured Construction Document Classification Model through Support Vector Machine (SVM)
Tarek Mahfouz · 2011
The dynamic nature of the construction industry yields enormous amount of documents that have to be stored, retrieved, and reused. Most of these documents are generated in an unstructured format. Therefore, in an attempt to provide a robust document classification methodology for the construction industry, the current research proposes an automated classifier model through Support Vector Machines (SVM). The adopted research methodology (1) gathered a corpus of documents including 300 correspondences, 150 meeting minutes, 25 claims, and 300 Differing Site Conditions (DSC) cases; (2) developed C++ algorithms which process unstructured documents into a readable format by the SVM algorithm; (4) developed 16 SVM automated classification models; and (5) tested and validated the developed models. The developed models under the current research attained higher accuracy, and better precision and recall than previous researches illustrated in the literature. The current research represents a continuation to previous researches performed within this realm.