Enhancing the Creation of Detection Rules for Malicious Software through Ontologies and Crowdsourcing

摘要: The analysis of malicious software (malware) is one the hardest open problems in computer security, since there a huge and varied number samples produced daily. In addition,modern programs have automatic mutation capabilities. Through behavior existing malware, we are able to understand new variants develop protection methods. Ontologies can be used model those behaviors, enabling experts define classes rules that represent complex behaviors. this paper, an ontology architecture built during our previous studies as starting point inspire development crowdsource-based framework platform. objective work explore crowdsourcing mechanisms collaboratively evolve ontologies, which users propose increasingly identify potential programs. With user-friendly platform, expect leverage could other malware systems, well quickly respond variants. Eight domain evaluated platform with goal validating identifying platforms potentials limitations.