Rumors

Do you remember the note that was circulated around in early December about sending mail to Santa@north.pole.org? The anti-business tone of the letter was a hint that something was fishy.

I thought about that note and a couple of other recent hoaxes over the past few months including the "Good Times Virus" and "Microsoft agrees to purchase the Vatican" and how much time, money, and bandwidth on the network was lost.

It would be a real service to the internet community if there was some way of controlling rumors.

It seems to me that the biggest problem is no one has time to investigate the validity of a rumor that they receive in the mail. In some cases the investigation would involve a simple phone call. In others, (like the good times virus), more in depth research might be necessary.

What if there were a place on the web that you could visit to check on the validity of a rumor? If you got a note from someone (who got it from a friend who got it from a friend...) and you wanted to see if there was any fact behind it (including, if possible, where the note originated), you could visit this web page.

The web page is great because it is a controlled information source. The owner of the page is the only one who can add or remove information. If the owner is careful and thorough about investigating the rumors, then the web page will gain a solid reputation for reliability.

What is the definition of a rumor?

Pick up Webster... but maybe it can be modified and tailored for the internet community.

What kinds of rumors exist?

• verified as true - emacs v19 has a security hole with mail functions
• verified as false - mail to santa@north.pole.org will cause SUN to donate...
• unverifiable - elvis was seen at the Latham McDonalds
• unverified - FCC considers fees for modem usage

What techniques can be used to classify a rumor?

Obvious manual ones like calling the company or organization that is the target of the rumor can work... or creating a test case that proves/disproves the rumor. Automatic ones might involve using the web or gopher or archie to track down information related to the rumor.

The problem with using the internet as a source of information is that the information is inherently unreliable. There is no editorial staff for the internet. And as more and more people join the community, the information garbage content will grow.

How do you determine a rumor is wide-spread on the internet?

Posting to widely read newsgroups is one way. How do you determine a newsgroup is widely read? Maybe a statistical sampling process like the random sampling that is used by pollsters. Polling on the internet... hmmm that's an interesting research question all by itself! In polls of "US citizens", demographics and location play a very important role... I wonder what the factors would be for internet users. How would you do this so that it is statistically significant?

How can you determine the impact a rumor has on the internet community?

Got me. Looking at papers and reports on nasty viruses that have affected the internet community would be a start. Figuring out how these people calculated the damage done to industry and academia would give some clues.

What techniques can be used to track down the source of a rumor?

I've hacked mailing systems in earlier days. The problem with internet mail is that it's so "OPEN". You only get a record of who sent YOU the note. You can't find out who sent the note to your sender.

It might be possible to build an automated system that creates a map of a rumor's trail by starting with your sender, asking politely for the name of the person who sent the note, and moving on along the trail. The problem is cooperation. If someone feels that they or a friend that sent them the rumor) did something wrong... the trail will dead end.

What is the difference between rumors generated on the internet, and ones generated by other forms of communication, such as oral, TV, radio, or newsprint?

I remember this game I played in grade school. It was called "Pass the Secret". We'd all sit around in a circle and the leader would whisper a secret into the ear of the person on his/her left. The secret would continue around the circle until it arrived back at the leader. It was always completely different from the original.

This doesn't happen on the internet. When you get a piece of mail from someone and you think it's noteworthy, you forward it to your friends and colleagues... nothing gets lost in the "translation".

What can be done about rumors on the internet?

The rumor web page is one idea. How would you ask this system about the rumor? Possibly a database engine could be used to index the text of rumors and provide an efficient query capability.