Using Machine Learning to Improve Public Reporting on U.S. Government Contracts
The U.S. Government procures more than $500 billion annually in goods and services on public contracts, which it classifies using a hierarchical product and service taxonomy. Classification serves several purposes including transparency in the use of taxpayer funding, reporting, tracing and segmenting government expenditures, budgeting and forecasting. Government acquisition personnel have historically performed these classifications manually, resulting in a process that is time-consuming, error-prone and that limits visibility into government purchases. We improve classification by leveraging natural language processing and machine learning techniques to generate a descriptive manual and a predictive classifier. Using 4 million historical data records on governmental purchases, we fit a machine learning classifier and demonstrate (a) superior performance when explicitly modeling the hierarchical structure of information domains through the use of top-down strategies; and (b) the effectiveness of character-level convolutional neural networks when textual inputs are terse and contain irregularities such as abnormal character combinations and misspellings, common in government contracts. Our machine learning models are embedded in multiple software applications, including a web application we developed, used by government personnel and other contracting professionals. Our innovative approach demonstrates a significant potential for sharply reducing the time spent by public officials in categorizing procurement transactions.
Team:
William A. Muir, U.S. Air Force Installation Contracting Center
Roger H. Westermeyer, U.S. Air Force Installation Contracting Center
Daniel Reich, Naval Postgraduate School