FRAMEWORK FOR AUTOMATED INFORMATION EXTRACTION FROM CONSTRUCTION CORRESPONDENCE LETTERS

Abstract

Information transfer in construction industry has always been immense mainly in the form of letters, emails or field study notes etc. But the process used predominantly, which is Manual Content Analysis (MCA) or manual method to deal with information extraction has never been quite pleasant due to number of inefficiencies including time-consuming, error prone, costly etc. To refine the approach and enhance the system, automated information extraction was explored and for this a code was prepared consisting of particular libraries and operations. This procedure based on the Natural Language Processing (NLP) technique using Artificial Intelligence (AI) assisted in extracting the useful information out from massive data, in our case the letters. The positive productivity by using this Automated Content Analysis (ACA) can be observed as it has eliminated all the issues that are associated with or may arise as a result of MCA process by making the Content Analysis (CA) process fast, and less troublesome.