Innovative Data Anomaly Detection and Transformation for Analysis Applications

The Navy has invested significant resources in constructing, maintaining and updating various information systems/databases that contain data/information that needs to be manually filtered to detect/remove anomalies prior to transforming it to perform useful analyses. The proposed approach will automate this process by: 1. Parsing the data/information needed and detecting its source 2. Allowing users to define a customizable template to extract/map data 3. Applying a logic system to the data/information records to automatically detect anomalies and repair records 4. Allowing users to define/select a customizable template and format repaired data/information to make it suitable for additional analyses The approach will utilize data/information on part failures; flight hours; number of aircraft landings; maintenance hours per repair; maintenance procedure content; part costs; part nomenclature; historical information on"mishaps", and other disparate data elements as part of a software tool that extracts, transforms and detects/repairs dataset anomalies in preparation for Naval Aviation Maintenance & Supply Model (NAMSM) analyses, eliminating the need for most manual intervention. Due to the data/information-independence and user-customizable capabilities of the approach, it will have broad applicability throughout the DoD, other government agencies and numerous commercial/industrial market segments not limited to reliability, availability and maintainability analyses.