Effective Anti-Spam Method with SMTP Session Abort
Grant-in-Aid for Scientific Research (B)
|Allocation Type||Single-year Grants |
|Research Institution||Okayama University |
YAMAI Nariyoshi Okayama University, Information Technology Center, Professor (90210319)
NAKAMURA Motonori OKAYAMA UNIVERSITY, National Institute of Informatics, Academic Network R&D Center, Project Professor (30268156)
MIYASHITA Takuya Tsuyama National College of Technology, Department of Computer and Information Engineering, Associate Professor (70304300)
OKAYAMA Kiyohiko OKAYAMA UNIVERSITY, Information Technology Center, Assistant Professor (20252588)
KAWANO Keita OKAYAMA UNIVERSITY, Information Technology Center, Assistant Professor (40397899)
|Project Period (FY)
2005 – 2007
Completed (Fiscal Year 2007)
|Budget Amount *help
¥6,250,000 (Direct Cost: ¥5,800,000、Indirect Cost: ¥450,000)
Fiscal Year 2007: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
Fiscal Year 2006: ¥1,700,000 (Direct Cost: ¥1,700,000)
Fiscal Year 2005: ¥2,600,000 (Direct Cost: ¥2,600,000)
|Keywords||e-mail / spam mail / SMTP / tempfailing / distributed collaborative spam database / Sender Policy Framework / spamメール / 情報システム / セキュア・ネットワーク|
This research project aims to develop an effective anti-spam method with minimal administration. This method introduces SMTP session abort function to obtain the header or the body of a message without actual delivery, in combination with existing techniques organically. This method has the following features.
1. Speeding up processing
We have developed a speeding up method of the existing tempfailing technique by introducing a secondary mail gateway (SMG). With this method, the sender would immediately resend the failed messages to the SMG after SMTP session abort, and hence delivery delay would be reduced.
2. Improving spam detection rate
Since most of non-resent messages destined to non-existing recipients are spam, we have improved the spam detection rate of the existing tempfailing method by registering them automatically into a distributed collaborative spam database. In addition, we have shown that delivery delay and the number ofReceived fields are possible criteria of spam detection according to delivery path analysis.
3. Reducing administrative cost
We have developed an advanced method of retransmission judgment using some header field or the message body obtained on the first delivery attempt. Since this method accepts a message from an untrusted SMTP client even in case of the second time delivery from a different SMTP client, the administrator does not have to keep such SMTP clients in the white list any more. In addition, this method can recover the non-rent messages in case of false positive, by showing their headers or bodies to users.
4. Suppressing bounce mails
5. Suppressing address leakage
We have adopted the existing method that receives all messages even to non-existing recipients and sends bounce mails only if sender authentication by Sender Policy Framework (SPF) succeeds. However, SPF has some drawback that it is difficult to authenticate forwarded messages. We have solved this drawback by tracing recipient addresses in the header.
Report (4 results)
Research Products (15 results)