Skip to Main Content
The incidence of Gumblar attacks, which compromise websites by infecting them with a virus and direct operations, has been increasing rapidly in recent years. Existing measures to detect dangerous scripts using a Web browser cannot protect against all of the attacks, because it is difficult for everyone to find dangerous scripts. Therefore, the authors devised a mechanism named RPF (Request Policy Framework) that uses a highly accuracy white list obtained by using a Web browser cooperating with the DNS server. This paper reports the detailed mechanism of RPF, the prototype program, evaluation results of its function and performance, and the consideration of coverage by RPF.