i’ve been thinking lately that we could potentially use social media monitoring tools to prevent suicides and mass murders. the idea struck me as i was working with some clients on a couple of issues / crisis management projects lately.
in the midst of the hurley burley of crisis mode, a news story caught my eye. killer george sodini went berserk in a pennsylvania gym and killed three, wounding nine before turning the gun on himself. the thing about it is, he blogged about doing it. that’s when the idea sparked.
so i did some digging
it’s not the first time, that social media was used by a killer or a suicide victim to declare their intentions. with a quick search, i found paul zolezzi, a model who declared he was going to kill himself on facebook and did. more interestingly though, i found hsu yu-sheng.
hsu yu-sheng
hsu is a gay and lesbian rights activist in taiwan, who on august 6 wrote a farewell note on his blog in english. After seeing the note, readers of his blog, launched a full scale effort to save him. friends and strangers alike, thousands of people banded together, to try to track him down and others posted kind comments to his blog.
police arrived at hsu’s place just in time and saved his life.
the idea
we use social media monitoring tools such as radian6 to listen to conversations on the blogosphere and elsewhere to protect brands. it’s not a stretch to deploy these tools to protect people.
how it would work
- radian6 set up to listen for a list of keywords
- suspect posts are parsed through to a heuristic analysis engine to further determine the sentiment of the post. radian6 is has a automatic sentiment engine built in, but we need one that would be tuned to suicidal/homicidal sentiments
- results that come up positive there are alerted to the on duty psychiatrist for an assessment and to alert the relevant authorities
there would be many issues that would have to be dealt with to make the system viable, feasible and workable and even then it would never be a certainty. what it would be is another tool in kitbag to tackle an extremely complex and difficult problem.
this post was first published on the howorth blog
About Jonathan Nguyen
This idea has merit and has crossed my mind a few times too. I think most organisations that currently look at suicide prevention lack the capacity to even understand this technology, let alone apply it. Would be scope to partner though, if radian6 could donate the technology to an organisation that could become capable if they had access to it.
I volunteer for Reachout.com and we have used technology to help young people, initially through the website but also with forums… I would also like to see targeted ads to at risk people – though I'm not familiar enough with keyword capabilities for targeting, one day I hope this would be possible with Facebook. If you right “trigger” words, an ad comes up to direct you to some help in an online environment.
It's all so new and non-profits are often starved of information when it comes to technology – but if you are keen to make this more than an idea I'd love to share thoughts…
Hi Sarah, Radian6 saw my post and made an offer to help. Let's have a chat about this offline. I'll shoot you through an email.
Hi Sarah, Radian6 saw my post and made an offer to help. Let's have a chat about this offline. I'll shoot you through an email.
Hi Jonathon, I know it's an old post but I've just come across it and wanted to comment. I work with a range of high-risk online communities, and recently worked with a parenting community for two years. This presented community management challenges such as suicide, child at risk of harm, post-natal depression, relationship issues etc.
One of the difficulties in processing these issues is ascertaining who the person is IRL, and whether or not you are contravening privacy laws in doing so. Mandatory reporting legislation varies from state to state and in most cases online publishers don't fall under it (to the best of my knowledge).
The other issue is whether or not you break the trust of the community by doing so. I'm not suggesting members would be bothered by an attempt to help, but you may erode the sanctity of that space. For many people facing troubled times, a call for help on a forum might be the best action they can take. They will be listened to and responded to (by their peers and the site). If people stopped posting cries for help for fear of the police turning up, would the action of community custodians be counterproductive?
I don't have the answer but it's a topic I've been exploring lately in considering the value of anonymity on the web. I enjoyed your post, thank you!
Hi Alison
I suspect it's one of those areas where there will be no easy answer. I also think it's gong to be one of those areas where there will be no one rule. So for example, on a Forum it might not be appropriate to intervene, but what if it was a Facebook status update? Would one of your friends be totally justified in calling the police to your house? What about your blog?
It's a tough call, but it it's good that people like yourself are wanting to explore it. For it me it was an interesting concept to explore, but I'm not a mental health professional. It would be interesting to get that viewpoint as well.
Thanks for commenting!
You're right – it is a tough call. I think the best course of action is to contact the member privately, and offer compassion whilst referring them to a professional.
Your suggestion is a great one (I just got hung up on the logistics before!). I imagine there are a vast number of organisations that deal specifically with these concerns who'd benefit from social media monitoring.
It might only be a matter of time until software is developed to tackle these issues specifically in the same way Crisp's Net Moderator deals with cyber-grooming and cyber-safety.
Great chatting with you Jonathon