DFLER: Drone Flight Log Entity Recognizer to support forensic investigation on drone device
Swardiantara Silalahi, Tohari Ahmad, Hudan Studiawan
Abstract
DFLER is an open-source CLI-based tool developed using Python and supported by a fine-tuned BERT model to perform named entity recognition on drone flight log data, specifically the log messages. This model is hosted on the HuggingFace platform to make it publicly available and accessible. The tool expects decrypted DJI flight log files as input and generates a forensic report in a PDF containing a forensic timeline highlighting parts of the mentioned entities in the log messages. The generated file is an attachment to a complete forensic report and helps forensic investigators pinpoint critical events on the constructed forensic timeline.
Topics & Concepts
TimelineDroneComputer sciencePython (programming language)Open sourceMerge (version control)Forensic scienceInformation retrievalOperating systemSoftwareBiologyGeographyStatisticsMathematicsArchaeologyGeneticsDigital and Cyber ForensicsData Quality and ManagementAnomaly Detection Techniques and Applications