The SpamResponder will go into the spam folder of a provided email and respond to each email as if you yourself had responded. Those who sent the emails will then need to spend time on reading and respondng to each email sent back and forth to them. This means they will have significantly less time to spend preying on real people.

Why is this an important project:

  • "A New York technology industry research firm confirmed the complaints of consumers, analysts, office workers and lawmakers by blaming the unsolicited e-mail known as spam for nearly US$20 billion in lost time and expenses worldwide." (1)
  • "a recent U.S. Federal Trade Commission test that resulted in more than 3,300 spam messages being sent to 250 e-mail addresses posted on Web pages, chat rooms, news groups and directories after six weeks. The first message came just nine minutes after one address had been posted in a chat room." (1)

Inspiration for this project:

How it will work:

Step 1: The user labels emails as SpamResponder to show the program which ones to target
Step 2: The program reads through the email and records the email it came from, the subject, and the body
Step 3: The body is analysed to obtain the sender's name and keywords from the email
Step 4: The keywords are used to construct a response, it doesn't need to be perfect English as the scammers have no idea if English is not your first language or not.
Step 5: The reply email is sent and the because the email chain is labeled it will continue to respond as long as possible

Challenges I ran into

We were especially challenged by the Google APIs used. Multiple credentials were needed as well as various authorizations all with different formats but similar names. We also had great difficulty getting the bot to send a reply email in the same thread. Even after following all directions in the documentation and ensuring it complied with the RFC 2822 standards it wouldn't thread the emails.

Accomplishments that I'm proud of

We are very proud of the program's ability to locate and respond to emails.

What's next for SpamResponder

The next steps are to give the django backend a pretty face and host the entire project on a domain.

Works Cited and Consulted:

  1. Data on negative effects of spam
  2. Why spam emails have poor grammar
  3. Gmail API: How to thread emails 4.Gmail API: Message Structure
  4. Django Web Development with Python Introduction
Share this project: