SoylentNews is people
SoylentNews
Encrypting emails can be tedious, difficult and very confusing. And even for those who have mastered the process, it's useless unless the intended recipient has the correct software to decode the message. A Georgia Institute of Technology researcher has created an easier method – one that sounds familiar to parents who try to outsmart their 8-year-old child. The new technique gets rid of the complicated, mathematically generated messages that are typical of encryption software. Instead, the method transforms specific emails into ones that are vague by leaving out key words.
"It's kind of like when mom and dad are talking about potential vacation spots while the kids are nearby," said Eric Gilbert, the Georgia Tech assistant professor who developed the software. "They can't say or spell 'Disney,' or the children will get too excited. So they use other words and the meaning is implied. Instead of 'Disney,' they could say 'have you bought tickets to the place yet.'"
Gilbert's Open Book system, a prototype that uses a Google Mail plug-in called Read Me, works the same way by substituting specific words with ambiguous ones. If the above example was an email conversation, the sender would write, "Have you bought tickets to Disney yet?" Open Book would change the message when it was sent. The other person would see, "Have you bought tickets to (place) yet?"
The process reduces the information disclosed to eavesdroppers or computer systems that monitor online communications, while taking advantage of common ground between the participants.
The system was presented at the ACM Conference on Human Factors in Computing Systems (CHI 2015) in Seoul, South Korea, April 18-23 (Open Book: A Socially Inspired Cloaking Technique that Uses Lexical Abstraction to Transform Messages) [PDF].
Original Submission
(Score: 2) by HiThere on Sunday July 05 2015, @07:24PM
A proper code is better than proper encryption...but MUCH more difficult to arrange. Even more difficult than one time pads. A proper code need to obscure the signs of the underlying grammar, as well as the individual words, reconfiguring the boundaries of where thoughts are demarcated. I.e., it essentially needs to be a language that is of a family previously unencountered. The WWII code talkers mentioned above were along this line, but the messages were theoretically breakable because the information on the language was publicly accessible. With good encryption, if you have the key the message is easily available. With a good code, the underlying message is available, but it can't really be translated into ordinary speech.
Please note: The code talkers were so successful because their language was not an IndoEuropean language. Some African languages would work equally well, though, e.g., clicks and be difficult to transcribe. And the code talkers didn't even talk straight Navajo, but a slangified version similar to the languages that teenagers create automatically to hide from their elders. This meant that the basic grammatical structures did not fit into the patterns that those trying to understand them expected. Hopi might have been an even better choice, but there were probably many fewer Hopi soldiers, as I believe that the base population is much smaller.
OTOH, a good one-time pad is also theoretically unbreakable, and usually much easier to arrange.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.