Open source intelligence (OSINT) in a colombian context and sentiment analysis

Abstract

Open source intelligence (OSINT) is used to obtain and analyze information related to adversaries, so it can support risk assessments aimed to prevent damages against critical assets. This paper presents a research about different OSINT technologies and how these can be used to perform cyber intelligence tasks. One of the key components in the operation of OSINT tools are the “transforms”, which are used to establish relations between entities of information from queries to different open sources. A set of transforms addressed to the Colombian context are presented, which were implemented and contributed to the community allowing to the law enforcement agencies to develop information gathering process from Colombian open sources. Additionally, this paper shows the implementation of three machine learning models used to perform sentiment analysis over the information obtained from an adversary. Sentiment analysis can be extremely useful to understand the motivation that an adversary can have and, in this way, define proper cyber defense strategies. Finally, some challenges related to the application of OSINT techniques are identified and described.

Keywords

• cyberintelligence
• open source intelligence
• adversary profiling
• machine learning
• sentiment analysis
• data science