Tiered prediction mechanism in collaborative e-commerce

Many medium and small enterprises have implemented their collaborative e-commerce oil ASP (Application Set-vice Provider) platform. On this platform, the enterprises can group spontaneously into some alliances according to their business requirements. The entities in the e-commerce platform can be divided into four levels: host, enterprise, enterprise alliance and global ASP, according to their positions in the architecture. Many surveys show that the aforethought malicious intrusion with the purpose of stealing business secrets begins at a single host and then breaks into other hosts on the same level. To prevent such intrusion, a tiered intrusion prediction mechanism for B2B e-commerce platform was proposed in this paper. The four prediction levels of this mechanism were defined according to the scope of the disaster. Finally, this intrusion prediction system was implemented with C/C++ programming language on Linux platform.