The NASA Engineering Safety Center (NESC) has commissioned our group to develop a tool to help the NESC cluster NASA documents and identify recurring anomalies.
This development project has a distinctive research component: Developing new algorithms/approaches to successfully mine for recurring anomalies. We are developing a family of novel methods to mine text documents and identify recurring anomalies across reports, and then integrating the algorithms into a web-based search platform.
The ReADS Team at IDU is developing the most efficient and effective algorithms and data mining techniques to enable researchers and engineers to obtain the vital information they need.
( Fort details, see our Publications Page )
The Recurring Anomaly Detection System tool is widely applicable:
The identified customers are NESC data mining and subject experts.
NESC has identified this tool as one of three applications to be used by NESC subject experts for text mining. NESC has also committed to evaluating the tool at regular intervals and providing feedback to direct the development to ensure a valuable tool in their repertoire.
Originally intended for NESC orbiter CARS (Corrective Action Reporting System) reports, the web based tool is being expanded to be used for ISS PRACA (Problem Reporting And Corrective Action) documents
The tool can also be expanded to mine other similar report data bases:
* NASA Center PRACA documents
* FAA Safety reports
* US Highway Safety Reports
* US Rail Safety Reports
Ashok Srivastava, Ph.D.