Graph-based malware distributors detection

摘要: Search engines are currently facing a problem of websites that distribute malware. In this paper we present novel efficient algorithm learns to detect such kind spam. We have used bipartite graph with two types nodes, each representing layer in the graph: web-sites and file hostings (FH), connected edges fact can be downloaded from hosting via link on web-site. The performance spam detection method was verified using set ground truth labels: manual assessments antivirus analysts automatically generated obtained companies. demonstrate proposed is able new malware even before best known solutions them.