Semantic Filters in Intelligence Analysis

It is a challenge to identify the relevant pieces for further intelligence analysis among a big chunk of data. Filters have been built to provide such a function in almost all the network traffic capture and analysis tools as well as signature-based intrusion detection systems. However, most filters only work on strings of words, numbers, and/or other symbols. This paper proposes a type of context-aware and semantically relevant filters. This proposal is built on the findings in ontological semantics [1]. A detailed case study is used to show the effectiveness and efficiency of this proposal. The result of this research indicates that a good filter for intelligence analysis should incorporate relevant linguistic theories, which can explain one major aspect of human intelligence at another level.