Trait Analytic Profiling Search (TAPS) Sponsor: National Institute of Justice

Product Description: TAPS research has resulted in new technology to address one of the most difficult areas of investigation “the discovery of the hidden.” In addition, through our research and analysis, we have discovered a fundamental issue facing law enforcement and forensic labs around the country “the computational cost of hashing.” As we move forward and are required to filter out “known good” and/or “known bad” files in a more efficient manner the Fibonacci Hashing Method researched and developed under this effort provides promise to reducing backlog, providing high performance triage / filtering and would allow for the integration of stronger cryptographic hashing solutions such as SHA-512 while reducing the overall time required to hash.   Fibonacci Hashing:       Applying Fibonacci requires us to utilize secondary cryptographic hashes, such as SHA-256 or 512 to confirm detection once flagged by the Fibonacci. What this facilitates is the use of stronger hashes while improving performance. Since the Fibonacci hash acts as a filter to eliminate files that are not in the Known Hash Set this dramatically reduces the number SHA-256 or 512 that must be generated resulting in improved performance.   Advanced JPEG Artifact Detection:   The JPEG steganography detection module was developed as part of the TAPS project, and  utilizes machine learning techniques to unearth traits from jpeg images in order to detect the presence of hidden data in JPEG images. JPEG is a widely used format in image steganography due to its huge popularity as a multimedia carrier.      The above diagram depicts the improved performance resulting in the use of Fibonacci Hashing.  Status: Initial Software Release to Law Enforcement Available for commercial sale   Contact Information: Primary Investigator:  Chet Hosmer, chet@wetstonetech.com  Field Trial Coordinator: Geoff Barron, gbarron@allencorporation.com