Instagram photos reveal predictive markers of depression
epjdatascience.springeropen.com
epjdatascience.springeropen.com
They used the color of the photographs, number of people in the photos, level of engagement, and happiness rated by Mechanical Turk to determine depression.
How many comments vs shares does a content post have?
Comments - disagree with it, want to correct something, controversy
Shares - agree with it, want to spread it more widely, echo chamber
Also, the fact that you wrote n:n with the colon in between would make it by definition a ratio.
I'm curious how ML will do that, not how but accuracy would be cool to make (train) the model. Also with thumbnails that capture the viewer's attention.
The other jobs were brute searching sites in hopes to improve SEO.
Transcribing photos of receipts (funny those ads, some poor schmuck being paid 12 cents per 10 min job).
Oh I remember an awkward one it was some social site's user photo uploads, I had to tag images with babies. So weird to see so many 'public' yet seemingly intimate photos. Not nude, just random people's photos.
I did this for a whole week as much as I could do sanely, made $55 not worth it. Washing plates makes more money.
I'm curious on how this works. Do you basically search for specific keywords and click on specific URLs in order to signal Google that the URL is more relevant?
Yeah you had to search some term and find the link. It seemed to work as they would be beyond the first page result usually or at least not #1 but say #4 on page 1.
But it could be biased you know based on your own search history I often check externally with say Tor to see how sites rank from other viewpoints than my own.
Anyway hope that is helpful. You could also just sign up and look around. I think I still have a 99.99% satisfaction rate a lot of jobs you need something close to that to apply but it is shit work though. Some go up to $1.00+ woweee but those take like half an hour to do it's ridiculous.
"Data collection was crowdsourced using Amazon’s Mechanical Turk (MTurk) crowdwork platform. Separate surveys were created for depressed and healthy individuals. In the depressed survey, participants were invited to complete a survey that involved passing a series of inclusion criteria, responding to a standardized clinical depression survey, answering questions related to demographics and history of depression, and sharing social media history. We used the CES-D (Center for Epidemiologic Studies Depression Scale) questionnaire to screen participant depression levels [26]. CES-D assessment quality has been demonstrated as on-par with other depression inventories, including the Beck Depression Inventory and the Kellner Symptom Questionnaire [27, 28]. Healthy participants were screened to ensure no history of depression and active Instagram use. See Additional file 1 for actual survey text."
So it sounds like they found depressed and healthy subjects via MTurk and verified them using surveys. I guess that's slightly more rigorous than just "people on MTurk thought the person in this photo looks depressed"
I should burn this nickname and be more diligent about persona management. Or get off social media entirely.
That will also be part of your Citizen Score, or it will impact your life like having a bad Citizen Score. Kind of like how you can't get a loan because you haven't had a loan before and you don't have a credit rating.
Wouldn't that be a bigger red flag?
I'm sure a ghost is an indicator for some things, but the majority of it is just speculation right? I have trouble thinking of how a lack of information about someone could be harmful as a source of information.
Now, I could see it becoming harmful in that I have to have some type of Citizen Score to apply for jobs, be accepted for loans, get insurance, etc etc. That's quite reasonable, imo. Yet, it's still seems different than being a source of information, e.g. this person is unstable, or this person is depressed, etc.
I play the same game. :)
If your employer/government/etc asks you for your facebook/google+/instagram/snapchat/etc account and you say you don't have any, it might come off as being suspicious. As if you are trying to hide something. The same thing with your banks/financial institutes/etc.
Why can't it simply indicate that I have no interest in sharing my personal life with those companies?
"Using modern machines/VMs, C64 demoscene devs are able to create content and develop new efficient algos for the old hardware that probably wouldn't have been possible using only the old hardware alone. They can develop algos and art content using sophisticated tools and backport to the old hardware."
Is just as true today as is this:
"Using invasive data mining, insurance companies are able to make more accurate inferences about their high-risk groups that probably wouldn't have been possible using only the data they are legally allowed to use to set rates. They can use sophisticated tools offered from data brokers to make inferences and then backport them into policy changes that meet the current guidelines/regulations for the industry."
In either case I don't know enough to explain what "backporting" entails. But there's just no way insurance companies aren't using this kind of invasive data already.
I declined an offer for a role exactly like this. Disgusting stuff.
I have been seeing a woman's posts lately who is going through a bitter, ugly divorce, and it's astounding how she represents everything publicly via Facebook when I know the full story due to a family member being involved.
This research is for finding a clinical condition and make less expendable the cost of health programs, not to evaluate the life of others based just on opinions. Suicides are the second-leading cause of death among teenagers. But, some people are just "not a happy person". That doesn't mean that they are unhappy, sad or suffering from depression.
I really like this comics from Oatmeal that talk about that. http://theoatmeal.com/comics/unhappy the essay that this comic is based thttps://outline.com/r6q4Ga
The algorithm didn't focus much on emotions of the subjects in the photos.
The cynical dystopian insinuation around technology, on a technology site no less, is getting old.
A technology site is exactly the place for cynicism about technology. The idea of seeing an ad for a suicide hotilne on a social media site is even more distopian than the op's crack about mandatory happiness.
I disagree, because I think we need to be the idealists and drivers instead of those pushing back, but I also don't think yours is necessarily an unreasonable position to have.
The idea of seeing an ad for a suicide hotilne on a social media site is even more distopian than the op's crack about mandatory happiness
I'm not sure I follow why. Ideally an offline social network (like your friends and family) would correctly infer your mental state from interactions and be able to intervene appropriately. A digital social network, where your friends and family, as well as the digital homunculus are all observing your activity should be able to infer your mental state as well, if not better than the offline system and intervene where necessary as well.
When I see Zuckerberg's fake, forced smile, I see a person trapped under there somewhere crying for help. I don't even mean this in a snarky way, I don't hate the guy, and I'm interested in the "Bill Gates transformation of being less insecure". But still, anyone near that and oblivious to it should read and listen when it comes to mental health, not talk and certainly not write algorithms that "intervene". If FB was able to want to help with it on any meaningful level, it would either disband or change drastically, and not have "be middleman for everything" as goal number one like so many do.
> idealists and drivers instead of those pushing back
You say this like there is a train, and it either moves or doesn't. Instead it's a multi-dimensional space with a million things that can all change position and shape and color. Technology isn't the question, but who is using it, why and for what.
We already know that Facebook/Instagram can manipulate your real-world emotions and thoughts by what it chooses to show you.
We also know that susceptibility to ads for different products varies depending on emotional state and context (pre 'surveillance capitalism' it was well known that you could get much bettter responses to aspirational products as the weekend approached, and to products like make-up on Sunday evenings as the work-dread sets in).
The question is, if you have a load of high-value inventory to shift, but your audience isn't susceptible to it, why not treat this as an optimization problem and make them?
It's the right thing to do for your shareholders, completely legal, and near undetectable.
Susceptibility to action is binary. If the audience isn't susceptible then they won't ever convert, or take action.
I'll never buy a Yoni Egg [1], and no amount of targeting or putting advertisements in front of me will cause me to make a purchase of one.
However there might be a world in which I would buy a box of feminine products, even though I am a man. For example if my daughter is sick or incapable of getting them for herself. In which case, a social or other network knowing that
1. I have a daughter
2. That she is sick
3. What her time based need of these products is
4. That there is a preference for a certain product in this category
5. I can get the product cheaply or easily
Makes my life easier, by saying: "Hey [user] we see that you may be in a scenario that this particular product/service would be helpful right now"
Even though at that point I might not even know that I need to help get that when I go to the store because of social anxiety or embarrassment preventing from asking.
So yes, I believe in an all seeing all knowing AI god that will guide and optimize our behaviors. That is not a sarcastic comment.
And it is a world where non participation might equate to a negative profile, so the bay we can do is understand the enemy and curate perfection, while engaging in a real way only when anonymous (so far as that can even be achieved).
There's a real tendency on HN to get all Ayn Rand and forget we live in a society. In relative terms, this is a very new business model. Companies will push for more power, and they get it unless people push back. As people close to the problem, we need to see ourselves as responsible for influencing the discussion about type of society we end up with.
We already use much less powerful things, words and images, mostly to manipulate. Plenty don't hesitate to trying to make people feel bad or unpopular for not having a product they don't actually need, and your susceptibility doesn't matter, that they also don't blink doing it to a mentally handicapped or psychologically unstable person is. But sure, just give marketing more power and at some point it will wrap around to being all nice and fluffy... I'm also not being sarcastic when I say you can live in that world over my dead body. It's not enough for you to want it, you need to want it more than I don't want it, and I doubt anxiety or fear of embarrassment is going to be enough motivation, not by orders of magnitude.
There are plenty of profitable industries built on the backs of people's mental unrest that has nothing to do with making them well.
Why do you think this technology would be deployed only aligned with the best interests of the users?
My guess is that it would optimize for the most proftiable behavior for the owner of the capital.
A free or cheap solution would not have the ROI a palliative / ongoing treatment would.
Why do you think depressed people are almost never given the option of writing gratitude letters, or are given drugs shown to increase the symptoms they proport to treat in the long term?
God help the millennials. The only winning move is not to play.
It's not like a single marker determines everything.
It's not even like having ALL markers means one is 100% depressed.
It merely gives you a confidence score in the person being depressed.
People who don't hang out with other people are more at risk for depression AND they are less likely to have group photos.
This is called a latent variable.