Good news about bad words

by Lithium Guru on 05-04-2009 06:44 PM - last edited on 05-04-2009 06:44 PM

One of the things we learned in our Community Health Index (CHI) research was how vitally important community ambiance is. Without a measure of civility, the atmosphere becomes toxic and participation drops like a rock.

 

Moderators are a community's first line of defense, but even great moderators can't be everywhere at once. Enter the smut filter: a handy list of words that you never want to see in your community. The mechanism is simplicity itself: when a community member clicks Post, the Lithium platform checks the contents of the post against the smut filter and flags any offenders.

 

The protection of a smut filter has long been available in English. Starting soon, it will become available in all of the languages that we support.

 

The process of assembling smut filters in 17 languages has been a challenge. We gave our translation service the list of English no-nos and turned them loose. We wanted not only literal translations, but also the local equivalent.

 

Some of the translation teams really got into it, sending us lists that were three and four times longer than the originals. Others were more circumspect. As a final check, we asked folks here at the office who are fluent in some of our supported languages to take the smut filters for a test drive. We knew that we had a winner when our own DavidGR, a native of Madrid, reported back that the Spanish smut filter made him blush.

 

So how do you smut-proof your multilingual community? That's an interesting question. What's to stop a community member from using another language to circumvent the filter? Do we make the filters cumulative and risk bumping out perfectly good English words that mean something entirely different in another language?

 

For now at least, we're leaving it up to each community. We've built in the ability to bulk-load filter terms and will make our list of naughty words available through our Community Success Managers (CSM).

 

Just, please, don't ask me what some of those words mean. Like David, I blush.

Comments
by Esteemed Contributor on 05-05-2009 06:44 AM

Susan,

 

I can only imagine the complexities of building the smut filter in each language, and then you throw in the curve ball of having to defend against other language terms for a given phrase being introduced in another language community.   While our community is presently English only, our members are truly global and I occasionally see phrases or even whole posts in alternate languages.  Luckily,  we have a few multilingual moderators, and have on occasion, made use of Google machine translation to understand and craft a somewhat appropriate response.

 

Perhaps this begs the question  if a member is able to slip a naughty word past the smut filter through use of another language, does this adversely impact the community if the majority of members don't recognize the term or phrase?

 

Thanks for sharing some of your challenges with your readers - it certainly adds some appreciation for a feature many of us take for granted on a daily basis.  It works so well we don't even think about it !

 

Any news on the new sarcasm filter that I heard might be in development?

 

Mark

by Lithium Guru on 05-05-2009 10:40 AM

lol -- good one, Mark. Since sarcasm is strictly in the eye of the beholder, I'd expect that one on a cold day in, um, a very hot place.

 

You do raise a good if-a-tree-falls-in-the-forest question. Is a word or phrase offensive if no one knows what it means? Don't think so. Which is part of the reason we're not trying to do anything fancy with multiple smut filters. Community managers know what they need. We're just making some comprehensive lists available if needed.

 

...Susan

by Frequent Advisor on 05-05-2009 01:51 PM

what about leet talk?

by Lithium Guru on 05-05-2009 02:48 PM

l33t?No filters for that. Do we need one?

by Frequent Advisor on 05-06-2009 12:29 PM

1337...using numbers to replace letters... like h3110[Hello]

there are others but these are the main ones

 

1=I

7=T

3=E

0=o

6=G

4=A

8=B

5=S

|] or |}=D

thats about it for me... but there may be more

 

by Frequent Advisor on 05-06-2009 12:32 PM

yes i believe lithium need leet filters it will add an extra "firewall" to your filter.

by Frequent Advisor on 05-08-2009 07:42 PM

"Leet" speak is annoying, but there is an infinite amount of combinations that can be made while using it. Anyway, this was a very interesting blog post, Susan! :smileyhappy:

Post a Comment
Type the characters you see in the picture above.Type the words you hear.
Announcements

Announcements

The Lithosphere: Your place to exchange ideas and share experiences about online community in the enterprise.

Getting Started

Here are a few ways to maximize your experience on the community:

  1. 1
    Choose your preferences
  2. 2
    Read our guidelines
  3. 3
    Check out the Help FAQs
About the Author
  • Michael Wu is the Principal Scientist of Analytics at Lithium Technologies Inc. Michael received his Ph.D. from UC Berkeley's Biophysics graduate program. His graduate research focuses on modeling the human brain, specifically the visual cortex, with techniques from math, statistics, and machine learning. Michael has been a DOE (US Dept. of Energy) fellow during his graduate career and was awarded 4 years of full fellowship plus stipend under the Computational Science Graduate Fellowship. During his fellowship tenure, he has also served at the Los Alamos National Lab, conducting cutting edge research in machine learning and face recognition. Currently, Michael is applying similar data-driven methodologies to investigate and understand the complex dynamics within online communities. Prior to his graduate research, Michael received his undergraduate degree from UC Berkeley triple majoring in Applied Math, Physics, and Molecular & Cell Biology.
  • Paul is the Director of Customer Marketing at Lithium Technologies, responsible for customer engagement marketing, social media and Lithium's own community, the Lithosphere.
Top Kudoed Authors
User Kudos Count
7