TExtractor: An OSINT Tool to Extract and Analyse Audio/Video Content

摘要

Hacking, data breaches, and information loss are a growing concern for organizations. Aware of the escalation of cyber threats, organizations are looking for ways to detect and mitigate cyberattack scenarios. Cyber intelligence, that is, knowledge produced through data and information on cyber threats and its actors is one of the means explored for this purpose. OSINT (Open Source INTelligence) is one of the areas of data collection for the production of cyber intelligence. In this paper we propose an OSINT tool (TExtractor) to facilitate the process of obtaining information about cyber threats. The TExtractor tool consists of extracting text from video/audio in open sources and searching for keywords linked to the activities of malicious actors. To support the development of TExtractor, we conducted a study to measure the effectiveness of different text extraction tools in audio/video sources. The results are presented in the paper and show that a tool like TExtractor can detect references to cyberattacks on audio/video sources in real time, with an accuracy between 60% and 70%.