engleski

Domain Specific Honeytokens Based on Natural Language Processing - A Conceptual Model

This paper presents the idea and conceptual model for keyword modelling by using Natural Language Processing (NLP) in a specific domain. The paper shows that keywords for Honeytokens (HTs) derived from a specific website domain can be generated automatically by extracting concepts of interest from security related or other domain specific texts. A conceptual model for generating such tokens is presented and guidelines for implementation are given. It is argued that such domain specific HTs are a better form of deception technology, that provides a harder challenge for detection from automated attacks and thus improves early detection and incident response procedures in modern complex systems.

honeytokens ; honeynets ; information security ; natural language processing ; domain modelling ; keyword generation

nije evidentirano