|
Technology Infusion of Text-Mining for Problem Trending into Software Change Reports at JSC
|
|
Point of Contact
|
Larry Grissom
larry.w.grissom@nasa.gov
|
|
Dates
|
June 2008 - November 2008
|
|
Problem
|
International Space Station generates ~1400 Software Change Reports (SCR) annually. Line analysts and management both need reports on trends and recurring problems, and FT 4-14 calls for increased trend analysis. But there is no trending tool currently used on SCRs. The SCRs? most useful data for trending are the phrases that describe problems, written in English text. For useful trending, these descriptions need to be extracted from the text (?semantically tagged?) and to be organized hierarchically and graphed.
|
|
Objective
|
? Fewer man-hours to find previous problems that are similar to on-orbit anomalies.
? Provide the governing IPT with the metrics they?re requested.
? Use text-mining to trend the recurring software problems which are documented in SCRs.
|
|
Results
|
No results are available at this time. Please check back again.
|
|
Keywords
|
Text mining, data visualization, problem reports, natural language processing, trend analysis
|
|
Categories
|
Issue & Risk Tracking
|
|
