Slashdot is affiliated with Sourceforge, and its “Terms of Use” link points to the Sourceforge “Terms of Use” site. The rules that the user abides by, according to the Sourceforge terms, are basically that the user must (1) provide true and current information for registration, (2) be responsible for his or her own content used in Slashdot, (3) not post illegal content, (4) “not sell, resell, or offer for any commercial purposes” any portion of Slashdot, and (5) not “abuse site resources or attempt to gain unauthorized entry to the site or site resources.” http://web.sourceforge.com/terms-use
Because the Sourceforge “Terms of Use” does not specifically explain the rules that Slashdot members must follow, I extracted rules (actually social norms) from the Slashdot FAQ in the “Help and Preferences” section. Because Slashdot strives to provide quality content (comments) by filtering out “bad comments,” the implied rules are to (1) not invoke flaming, (2) not be a troll (the troll definition as posted by Slashdot is in Example 3), and (3) stay on-topic. Most of the information on filtering out “bad comments” and trolls are in the “Comments and Moderation” FAQ. I found a few implied rules in the “Accounts” FAQ which are to (1) post under a false identity, (2) not flood the forums with offtopic, flamebait, or troll comments, and (3) not use programs that attempt to disrupt Slashdot servers. http://slashdot.org/faq
2. Three Examples of Broken Rules
Example 1:
For this first example, I extracted the implied rule of not posting under a false identity from the Slashdot Accounts FAQ. http://slashdot.org/faq/accounts.shtml#ac600
Someone is posting under a false identity, or an account designed to look like someone elseI performed a search for users using the keywords “commander taco” which resulted in a list of 90 users. I performed another user search using the keywords “cmdrtaco” which again resulted in a list of 90 users. (90 might actually be the maximum number of results returned from the search routine used by Slashdot) Many of the users that seemed to have malicious purposes did not post anything to their account, submit stories, or post comments. They may have been caught by moderation as described in the above excerpt from the Slashdot FAQ. Possibly because of the Slashdot moderation system, I was not able to find any instances of comments made by imposters to troll a thread of comments.
There are many accounts in the system that were created for malicious purposes. There are dozens of variations on names like 'CmdrTaco' and 'Hemos' as well as heads of many corporations, writers, etc. We don't tamper with existing accounts though, so there's nothing we can do about it. Over the last few years, numerous restrictions have been placed on accounts to make this difficult (for example the system won't let you create an account named 'CmdrTaco' because there already is one listed) but that doesn't prevent any of the hundreds of existing ninnies from doing what ninnies do. In most cases, these folks are caught by moderation, and they eventually get karma that is low enough to make it obvious that they are impostors.
Answered by: CmdrTaco
Last Modified: 6/19/00
False CmdrTaco identity found from "commander taco" search. No activity on account.
False CmdrTaco identity found from "cmdrtaco" search. No activity on account.
Example 2:
The example 2 implied rule of not posting bad comments to invoke flame wars or enrage other members was extracted from the “Comments and Moderation” FAQ. http://slashdot.org/faq/com-mod.shtml#cm1600
What is a good comment? A bad comment?One of the goals of Slashdot is to “Promote quality, discourage crap” as indicated in the “Comments and Moderation” FAQ. One method that Slashdot uses to filter out bad or off-topic comments as described in the above excerpt is to allow moderators to score comments. The scores are based on a scale of -1 to 5 with 5 being the most positive. Slashdot readers can set a threshold, which allows them to view comments that have a certain score level.
A good comment says something interesting or insightful. It has a link to a relevant piece of information that will add something to the discussion. It might not be Shakespeare, but it's not Beavis and Butthead. It's not off topic or flamey. It doesn't call someone names. It doesn't personally attack someone because of a disagreement of opinion.
Some of my favorite "bad" or off-topic comments are things like "Slashdot sucks!" and "This isn't news for nerds!" and "Moderate this XXX!" Any of these may be true, but they're probably off topic!
Answered by: CmdrTaco
Last Modified: 6/12/00
Read Slashdot at a threshold of 3 and behold the quality of the comments you read. Certainly you aren't reading a wild and freewheeling discussion anymore, but you are reading many valid points from many intelligent people. I am actually pretty amazed.The forum for the “Intel CPU Privilege Escalation Exploit” story had 3 groups of bad comments that went off-topic. They were in reference to the appearance of Joanna Rutkowska, a security specialist who developed Blue Pill. The parent comments for the three groups are
Answered by: CmdrTaco
Last Modified: 6/26/00
- But more importantly... (Score:5, Funny)
- wow (Score:1, Offtopic)
- Joanna Rutkowska (Score:1, Offtopic)
The “Joanna Rutkowska” group. All comments have a score of 2 and under.Example 3:
The implied rule for example 3, which is to not post offtopic, flamebait, or troll comments was extracted from the “Comments and Moderation” FAQ.
What sorts of anti-troll filters exist?
A handful of filters have been put into place to try to make sure that people don't abuse the system. The most important is that the same person can't post more than once every 120 seconds. Also, if a single user is moderated down several times in a short time frame, a temporary ban will be imposed on that user... a cooling off period if you will. It lasts for 72 hours, or more for users who have posted a ton.
The vast majority of you will never encounter any of these troll filters. If you do encounter one unfairly, let us know so we can fix it. This stuff is fairly beta code, so there are bound to be problems.
Answered by: CmdrTaco
Last Modified: 1/27/02
Moderation Labels for Comments
Offtopic -- A comment which has nothing to do with the story it's linked to (song lyrics, obscene ascii art, comments about another topic entirely) is Offtopic.
Flamebait -- Flamebait refers to comments whose sole purpose is to insult and enrage. If someone is not-so-subtly picking a fight (racial insults are a dead giveaway), it's Flamebait.
Troll -- A Troll is similar to Flamebait, but slightly more refined. This is a prank comment intended to provoke indignant (or just confused) responses. A Troll might mix up vital facts or otherwise distort reality, to make other readers react with helpful "corrections." Trolling is the online equivalent of intentionally dialing wrong numbers just to waste other people's time.
Answered by: timothy
Last Modified: 2/17/03
As mentioned in the second example, Slashdot strives to “Promote quality, discourage crap.” This includes filtering out trolls and comments that are meant to “insult and enrage” other members. Comments that are meant to “insult and enrage” or “confuse” readers, are specifically labeled “offtopic,” “flamebait,” or “troll” as defined in the Comments and Moderation FAQ. These labels are used to inform other readers of users who are potential trolls or troublemakers. However, many comments labeled as “troll” are posted anonymously as “Anonymous Coward.”
Users may also filter out comments made by trolls or troublemakers by setting their threshold to 3 or 4. Offtopic, flamebait, and troll comments are usually given scores of -1.
I found two instances of “troll” comments in the “Plethora of New User Space Filesystems For Mac OS X” and “Decent DVD-Ripping Solution for Linux” stories. In the Mac OS X Filesystems discussion, the “troll” comment was meant to insult Macintosh users. The “troll” comment in the Linux discussion was meant to insult Linux users by mentioning the faults of using Linux. The “Anonymous Coward” is probably a Macintosh user as suggested by the comment “Be serious and get a Mac.” This post is interesting because it appears that the user intends to provoke Linux users to post defensive comments to his or her insults. Actually, one member at the risk of obtaining a “-1, Offtopic” rating replied to the “troll” comment.
Troll comment from “Plethora of New User Space Filesystems For Mac OS X”
Troll comment from “Decent DVD-Ripping Solution for Linux”
3. Responses by Site Admins and Other Users
For all three examples, I believe that using moderators to help identify trolls and troublemakers is very effective. Using the scoring and threshold systems to filter out distracting and bothersome comments helps to promote quality content. Using moderators allows the site administrators to not have to be concerned with having to remove accounts created by trolls and troublemakers. Trolls and troublemakers are ignored by using the threshold system to filter their comments. Because they are not acknowledged, they tend to not add content after an initial post.
By having site administrators remove false identity accounts and bad content, the offenders may overrun the site, causing it to shutdown. Forums might be over-flooded with bad content. Keeping false identity accounts and bad content allows moderators to identify and filter out offenders and bad content for other members.
4. Relate Examples to Session 7 Readings
According to the Cosely et. al. article, Slashdot is a member-maintained community that uses oversight (using members to review contributions) to moderate posts or comments. As mentioned in the article, members read stories (actually, story submissions by other members) posted by editors and post comments on them. Any Slashdot member may obtain moderator access according to the rules posted in the “Slashdot Moderation” document to score comments. These scores are used to filter out “bad comments.” There is also a Metamoderation system that allows other moderators to moderate moderations. According to Cosely et. al. Slashdot’s distributed moderation scheme hides “low-quality contributions,” allowing quality contributions to “float to the top.” Oversight appears to improve the quality of the viewable content, but I have found groups of off-topic posts that were visible to members with a threshold of 4. However, I’m not certain of reducing antisocial behavior. I believe that many members contribute to Slashdot because they are allowed to post anonymously more than using oversight mechanisms. Being made aware of oversight as in the Slashdot “Comments and Moderation” FAQ, does seem to discourage bad contributions. These bad contributions are ignored and offenders don’t seem to make additional contributions.
Relating the Slashdot community to the Kollock and Smith article, the overall individual goal and overall group goal of the Slashdot member is to filter out bad comments. I believe that despite not having a set of well defined rules, Slashdot is able to successfully filter out bad comments because it uses members as moderators to score comments and identify bad comments and trolls. This allows Slashdot to set the same goal for the individual and group to encourage cooperation to achieve the goal of promoting quality content.
The Gazan – Rogue User article characterizes an online community rogue user. The Slashdot community appears to have identified rogue users as trolls, and has developed a moderation system to identify rogue users and filter out their comments. Flooding was mentioned in the Accounts FAQ, where site administrators will ban a user from Slashdot if he or she floods the forums. Moderators will score comments with abusive language negatively and flag them as flamebait or troll to help Slashdot members to filter these comments. As mentioned in Example 1, members who create “sock puppets” usually have malicious intentions. In addition to the rogue user characteristics mentioned in the Gazan article, I found that a group of rogue users challenged the authority of the site administrators with name-calling and derogatory statements. These members did not post anonymously indicating that they wanted the “rewards of recognition.”
