1. Removal from public accessibility. A number of e-mail addresses were posted on publicly accessible Web sites for two weeks, then removed. The goal was to determine whether removing the address from public view would have an effect on the overall amount of spam received.
2. Posting in "human-readable" form. Some Internet users posting their addresses in public places have altered the form of their e-mail address in such a way that another user can still easily reach them, but an automated tool would not recognize them. For example, a user with e-mail address example@domain.com could post his address as "example at domain dot com." We tested the effectiveness of this practice by posting addresses on the Web and on USENET newsgroups in this "human-readable" form.
3. Posting in HTML-obscured form. Tech-savvy Internet users have sometimes used special codes in HTML -- Hypertext Markup Language, used to construct Web pages -- to post their addresses in a way that Web browsers can interpret, but that is an obstacle to automated spam tools. In HTML, the letter "e" can be written "e" and the "@" symbol "@." So, the address "example@domain.com" could be written "exampl e@domain.com."[3] We tested the effectiveness of this practice by posting addresses on the Web and on USENET in this HTML-obscured form.
4. Changing personal preferences on a Web site. Many Web sites provide users with the opportunity to alter their personal preference so that they no longer receive e-mail communication from that site. Some Internet users, however, have been concerned that changing those preferences will have little effect on the amount of spam received, believing that once an address is "out," there is little they can do about it. We tested the effectiveness of changing one's personal preferences by returning to Web sites to which we'd submitted e-mail addresses and changing the addresses' associated preferences to request no further e-mail communication. We tried this in two separate ways. For certain addresses, we would "opt-in" to certain kinds of communication, then log back in and immediately change our preferences to "opt-out." For another set of addresses, we allowed at least two weeks to elapse before changing preferences. In both cases, we allowed a two-week "grace period" for our changes to take effect before classifying received e-mails as spam.
Source: cdt.org
2. Posting in "human-readable" form. Some Internet users posting their addresses in public places have altered the form of their e-mail address in such a way that another user can still easily reach them, but an automated tool would not recognize them. For example, a user with e-mail address example@domain.com could post his address as "example at domain dot com." We tested the effectiveness of this practice by posting addresses on the Web and on USENET newsgroups in this "human-readable" form.
3. Posting in HTML-obscured form. Tech-savvy Internet users have sometimes used special codes in HTML -- Hypertext Markup Language, used to construct Web pages -- to post their addresses in a way that Web browsers can interpret, but that is an obstacle to automated spam tools. In HTML, the letter "e" can be written "e" and the "@" symbol "@." So, the address "example@domain.com" could be written "exampl e@domain.com."[3] We tested the effectiveness of this practice by posting addresses on the Web and on USENET in this HTML-obscured form.
4. Changing personal preferences on a Web site. Many Web sites provide users with the opportunity to alter their personal preference so that they no longer receive e-mail communication from that site. Some Internet users, however, have been concerned that changing those preferences will have little effect on the amount of spam received, believing that once an address is "out," there is little they can do about it. We tested the effectiveness of changing one's personal preferences by returning to Web sites to which we'd submitted e-mail addresses and changing the addresses' associated preferences to request no further e-mail communication. We tried this in two separate ways. For certain addresses, we would "opt-in" to certain kinds of communication, then log back in and immediately change our preferences to "opt-out." For another set of addresses, we allowed at least two weeks to elapse before changing preferences. In both cases, we allowed a two-week "grace period" for our changes to take effect before classifying received e-mails as spam.
Source: cdt.org