Emotion-Awareness: Do we need a ‘Mood Graph’?
As the internet is entering the ‘Age of Context’, companies are getting more and more aware of the importance (and value) of how we feel about the content we share, the places we visit and life in general. Of course, being more aware of your emotions can be a benefit to users too, but how do you – and the corporations – keep track of your moods?
Mood Data Sources
Manual User Input and Tagging
Social & Sharing Platforms
- LiveJournal was, as far as I know, the first platform to have predefined ‘mood’ status updates, starting more than a decade ago.
- Google recently introduced mood postings into their Google+ social platform in the form of animated emojis.
- Facebook added ‘I’m feeling… with a selection of emojis’ to their structured status updates in April.
- The newly launched Bitly for feelings allows you to add an emotion indicator to the content you share visible to all just by the shortner’s url: there’s yaythis.me, loathe.es, upvot.es, luvit.me, .. .
(ProTip: Share this article using yaythis.me/17rVuws or scaryto.me/1dyH7xs depending on how you feel about it.)
… and with their ‘Inspiration Index’, even LinkedIn is starting to inquire how you feel.
Dedicated apps and websites
There’s quite a few quantified self mood logging tools out there. My current favourite is the sensor-infused smartphone app EmotionSense, but others such as MoodPanda and MoodStats allow you to track your degree of happiness too.
Automated ‘mood-recording’ devices
More and more devices which sole purpose it is to – quite literally – read your mind using EEG, and translate that to actionable knowledge about what you’re thinking and feeling are now available or have successfully completed their crowdfunding campaigns. Major ones are:
- The ubercute Necomimi brain-computer interface with fluffy cat ears that move in response to the wearer’s emotional state (and that has been on my wishlist for ages).
- Muse, the brain-sensing headband that lets you control things with your mind by Interaxon.
- Emotiv’s older EPOC, and the next generation Emotiv Insight which measures and tracks your Attention, Focus, Engagement, Interest, Excitement, Affinity, Relaxation and Stress level, as well as detecting facial expressions such as surprise and smiling.
Mood derived from personal data
And of course, many services can make an ‘estimated guess’ as to how you are feeling, based on the, often public, data you produce.
- More and more research is happening into the correlations between our smartphone usage and our emotions.
- What music you listen to is a good indicator as to your mood, take for example EchoNest’s API, as explained by Daniel Portwood.
- Facial recognition expression software for video as well as images is getting better and better.
- Speech recognition software is constantly improved upon to accurately assess a customer’s mood.
- Textual sentiment analysis, however still rudimentary especially with more obscure languages such as Dutch, on anything you write. Although post-based analysis is for now mostly classifying posts as ‘negative’ or ‘positive’, given access to your full corpus of Tweets, researchers can already conclude what type of person you are.
- Google was recently granted a patent for Glass to track your emotions through your eyes (and to charge advertisers accordingly).
Do we need – and want – a Mood Graph?
Which such a huge range of different ‘sources’ (and data formats) for emotion tracking, could the internet benefit from – as there’s already Facebook’s Social Graph, Google’s Knowledge Graph and Runkeeper’s Health Graph – a Mood Graph? That’s the subject of this great Wired article ‘The Mood Graph: How Our Emotions Are Taking Over the Web’ by Evan Selinger, in which he splendidly describes the pro’s and con’s of of mapping our emotions against a structured set of emoji. A big benefit is that emojis and other key inputs of a mood graph might help us express ourselves better, as they make up for the lack of body language when communicating digitally.
Amongst the possible drawbacks Selinger points out is the fact that using a pre-fab structured, and thus confined way to express how we feel, might also limit the ideas we can express: “Just as we shape our tools, they shape us too. It’s a two-way street.” Then there’s also privacy (even if you don’t explicitly tell a platform about your current mood, it isn’t that hard to do an estimated guess) and the lack of value we get in return:
“And we’ll continue providing these companies and platforms all this emotional labor and data – letting them exploit and monetize it without paying us for it — because these “siren servers” (as Jaron Lanier calls them) have us all captive: we can’t resist the call of going where our friends are.”
The Benefit of Emotion-Aware Apps & Brands
Yes, that’s probably what is going to happen at first. But as we are becoming more aware of the value of our personal data – or that is what I like to believe – we’ll ask companies not only to be more transparent about the data they gather and how they use it, but also to offer us more in return for it. And it would benefit them too! Not only could we correct them when they are wrong; showing me the same leather jacket I can’t afford over and over hardly makes me like your brand; by making me happier – and not annoying me that often – they would win my loyalty.
For example, when I’m feeling down and a bit lonely, I don’t need to get a standard promotional email about winning a meet and greet with Robbie Williams, I want your app to change its music to accommodate my mood, with possibly, a playlist that slowly evolves into my ‘happy’ music. Who knows, you’ll actually make me feel better, I’ll be grateful, and I’ll renew my subscription. (Or at least, not cancel it.)
And that’s just a minor example. Let’s consider a more emotional-aware Siri or Google Now. Respectively, Siri could be a bit more considerate and less tongue-in-cheek regarding love when she gets word you’ve changed your Facebook relationship to ‘single’ once again and are listening to Nine Inch Nails. Google Now could try to cheer you up – or at least distract you – by notifying you about your favourite football team’s recent win. But for a variety of apps to be able to take our emotions into account, we do need a semantically-well-structured way to pass data from all the different sources listed above to them. So the answer is, yes, we do need a Mood Graph of sorts. To me, a structured dataset of my moods – which I can review, correct and revoke access to at any time – is not creepy, but essential for technology to become a ‘true’ companion, who listens and adjusts itself to my moods. (Admittedly, as transparency is a requirement, I might not want Facebook to host my Mood Graph.)
And yes, I would also very much appreciate my freezer to automatically order a replenishing of my Ice Twix stock when it’s made aware that I’ve had a crappy day, and my SmartTV should know that due to my state of mind, I’d probably prefer watching The Sopranos over Mamma Mia!
What about you? What ways do you see to ‘digitize’ your emotions into data applications can do something useful with? Whom would you trust with that data? And what would you like apps to do (differently) once they are aware of your emotions? I’d love to know, and there’s a huge comment box below this post! 😉