Recruiting (Updated, now with survey option) How do you feel about algorithms analyzing your Bluelight data?

Hello again! Please see the original post in this thread because we've updated the study to include a survey option. This survey is going to address the same kinds of questions we asked in the interview, but should be much easier and quicker to complete (about 10 min). For anybody who's interested in the subject but didn't want to do a one-on-one interview, I hope this format will make it more attractive.

Here's the link to take part in the survey: https://lehigh.co1.qualtrics.com/jfe/form/SV_8caTvv8sIiqBHOC
 
Analyzing this data show who are you,where are you from,your sex....all.This is like to have fullprofile of somebody else.Personally me don't care.I have nothing to hide....and i am old and sick and just don't care.Not much time for me I hope
 
Analyzing this data show who are you,where are you from,your sex....all.This is like to have fullprofile of somebody else.Personally me don't care.I have nothing to hide....and i am old and sick and just don't care.Not much time for me I hope
I appreciate your concern, it is important to be careful about information you give out. However, this survey will not ask you for any demographic or personal details whatsoever. I have also confirmed that Qualtrics will not record IP addresses, browser information, or any other data that could be used to build a profile of somebody.

You've made me realize I should clarify an important distinction from the interview. The survey will not ask for your Bluelight or other usernames, and will not do any sort of analysis with your data on Bluelight or elsewhere. I will update the original post to make this clear.
 
Bumping this thread one more time because we've gotten very few responses so far. It's a very easy questionnaire to get through, and gives you the opportunity to share your opinion on an important and potentially threatening emerging technology. Even if you think the application to Bluelight is insignificant or unlikely, this research will provide valuable context and criticism to the trend of harvesting and analyzing "public" internet content with algorithms that find (correctly or not) information that wouldn't normally be possible for human observers. This is the "human" part of Human-Computer Interaction, and it's something the programmers and app developers need to know about!

Finally, even if you've taken part in the interview already, you're more than welcome to complete the survey as well. I'm always happy to address hesitation or provide more information, either here or privately, so please reach out!
 
Bumping this as this kind of research is extremely important. There is zero potential privacy risk to taking the survey and very little potential risk in participating in the interview. I have personally done both. The survey is just asking about your thoughts on online privacy and stylometric algorithms.

Law enforcement and government surveillance agencies already have this kind of data available. Participating in the study will allow Bluelighters to also have this data available to us. We will be able to learn about the possible risks of stylometric technology in identifying people, and take precautions to protect ourselves.
 
Last edited:
tried to schedule the interview but i got an error saying something like "the booking profile is not available"
 
Wow @Skeebs this is pretty amazing! I knew this was a possibility and even happening in other Uni’s in the UK, on a smaller scale obviously.

I can predict the reaction you’ll get from bluelighter’s. Looking forward to the comments on this one.
That sucks about the same as Facebook and the rest. Thanks a lot for your hard work.
 
I just had a brilliant idea. If you wanna see if this thing works: run it on that Jabberwocky account and then cross reference the posts with currently (reincarnated) active members. How's that for a "control" test! :ROFLMAO: Proof of concept no? But the results have to be made public of course!

Actually not that dumb of a suggestion nor my simply taking the piss (although that was the initial intention I'll admit).

Think about this:

There's more than a few members who have been around for decades and under not such ideal circumstances that may have been posting under different usernames and who may have, by now, cleaned up their act somewhat and are posting with a new account. Guess my point being: to what extent would substance abuse, or even specific substances, obfuscate somebody's written communication style (Tweakers need not respond here i.e. AI not required for analysis).
My question had to do with Tweakers. As I have not know of any real Meth in about 15 yrs or more, what exactly is being sold as such? And it’s really cheap too. I still see homeless tweakers who aren’t skinny but their still looking for some $hit. Lol
 
My experience is that the more "advanced" a platform is, the less it works. Unfortunately I'm convinced this is mostly the case for usability and overall performance, not for extracting information about the users, or trying to sell them.something.
 
Top