PROJECT TITLE: Learner Data Management
Project completed December 2018
Collaborators:
- PI: Dr. Christan Grant, Assistant Professor, Computer Science Department, University of Oklahoma
- Co-PI: Dr. Dean Hougen, Associate Professor, Computer Science Department, University of Oklahoma
- Students: Keerti Banweer, Santosh Maddula, Chenguag Xu
- Industry Partner: Computer Systems Designers
- FAA Sponsor and Technical Monitor: Shawn Mansfield, Technical Training (AJI-2)
Abstract:
Learner data for different sets of FAA trainings is stored across different systems at different locations. In the 1st phase of our project, we explored and analyzed existing database management systems in the academy. We looked in detail at NTD, TRAX, CEDAR, and the Training Database Management systems. In the start of the year, we started with identifying and creating an open-source document storage and search system to support a safe and efficient airspace system. With that motivation, we determined that the Comprehensive Knowledge Archive Network (CKAN) system was suitable to move forward with our work. Using CKAN, we installed and created a document storage website. We continued our analysis of the existing database to be able to understand the different training systems used across the FAA. We also analyzed the course documents required for training air traffic controllers for the FAA. We analyzed the existing schema of the database management system and the relationships between the different systems for updating and transferring data. This helped us analyze what types of information are stored and maintained for trainees for the FAA. There have been multiple updates in the requirements of the project which created a diversion in our analysis and approach for accomplishment of the project goals. With the help of project coordinators, we learned about the developing system CEDAR and Training Database Management system. These discussions were very informative and helped us in understanding the most recent changes in the FAA database management system. These discussions changed the directions of our plans, as we learned about the database systems in development and that it will affect our plans. This report details our analysis and our inputs on the existing FAA training databases. This report also includes the details of our schedule and milestones.
Expected Project Outcomes:
The goal of this project is to empower FAA management with data-driven insights into connections between training and performance, reduce the data management burden, and improve technical training and human. To accomplish this goal we will create a prototype system for an intelligent data enclave that integrates existing FAA data sources, accommodates new data sources as they are developed, allows appropriate access to data by a wide variety of system users, and incorporates flexible analytics that produce insights into the data to encourage data-driven decision making. The intelligent data enclave thus will provide an ecosystem for secure data management and knowledge extraction, as well as a flexible and adaptable interface for future use.
VALUE OF RESEARCH:
This project will focus on the application of twenty-first century text analytics, knowledge extraction, and machine learning techniques to integrate data from existing FAA databases and transform it into useable information for more efficient and effective management of Aviation Safety.