Re: ID of anonymous posters via word analysis?
Just reading this list I'm sure it would be fairly clear that word analysis could be used to identify posters. Reread a few posts on the cypherpunks list. Note who spells out "government" and who abbreviates to "gov't". Some people consistently use one or the other. Count who uses "though" and who uses "tho". Also look at who refers to "anonymous posters" and who talks about "nyms". I think you will notice some definent patterns. Other possible word favoritisms: cypherpunks/c-punks cryptography/crypto cipher/encryption America/USA England/UK baud/bps DigiCash/digital cash Internet/"the net" information/info Mail/E-mail/Net-mail Just looking at the above list I'm sure some of you will realize how much you favor using certain terms, others probably without noticing it. Sorting by subject is possible too. Notice that there is only a certain group of users who consistently discuss DCnets. Another group consistently mentions the IRS, and taxes. A different group typically discusses anonymnity and anonymous postings. Others tend to avoid certain topics. Think about your own postings and realize what topics interest you most. I don't think it would be too hard to establish a "text fingerprint" of people based on what words they use. Maybe when I have some time I'll write a program to do it and see how many different patterns/styles I can identify. P.S. Also note the variations in text markings to express emphasis. Note who CAPITALIZES, *stars* _underscores_ or Capitalizes The First Letters.
Matthew J Ghio writes:
Just reading this list I'm sure it would be fairly clear that word analysis could be used to identify posters.
Though I agree with some other contributors that iron-clad identification may require substantial amounts of material, I think Mr. Ghio's point is correct to the extent that a party can satisfy itself informally that a particular anonymous post is from some well-known identity. Though the evidence may be useless in a legal sense, that's not a problem in some contexts. For example, if Bob Scum is posting anonymously some risky notes to a particular mailing list or newsgroup, it may be quite unfortunate for Bob if mere suspicion arises that the notes are from him. That suspicion need not be based on admissable-in-court evidence; if it's noted by someone that both Bob and the anonymous author routinely use the word "copacetic", things could heat up for poor Bob. If the risky notes involve some socially unacceptable topics like drug use or pornography, the fact that Bob can't actually be convicted is unimportant. -- Mike McNally
participants (2)
-
m5@vail.tivoli.com -
Matthew J Ghio