Balneus

Australian Lefty on Politics, Governance, Science and Info Management

Smileys and Frownies becoming obsolete

Posted by Dave Bath on 2007-07-19

Are smilies, frownies and other emoticons about to get a lot more complex?  The Final Report of the W3C Emotion Incubator Group gives hints as to what is coming with markup languages to describe emotions.

Apart from delving into the various use-cases and psychological theories, for non-techies in the blogosphere, it means that we might see smilies that are animated or have subtle degrees of facial expressions to accurately reflect intent.  This research also has application to current-generation mobile phones.

There are a few drivers for this, including the benefits for clinicians and researchers, but the gaming industry (including "Second Life") will be the first to exploit this (by having a compact means of describing changes of expression.

One of the extant markup languages examined by the committee was EARL (Emotion Annotation and Representation Language).  Here are a couple of snippets:

  • Ambiguity coded through probability:
    <complex-emotion xlink:href="face12.jpg">
      <emotion category="pleasure" probability="0.5">
      <emotion category="friendliness" probability="0.5">
    </complex-emotion>
  • Major/minor emotion coded through intensity:
    <complex-emotion xlink:href="face12.jpg">
      <emotion category="pleasure" intensity="0.7" />
      <emotion category="worry" intensity="0.5" />
    </complex-emotion>
  • A simulated emotion masking a suppressed one:
    <complex-emotion xlink:href="face12.jpg">
      <emotion category="pleasure" simulate="1.0" />
      <emotion category="worry" suppress="1.0" />
    </complex-emotion>
  • Different emotions in different modalities:
    <complex-emotion xlink:href="face12.jpg">
      <emotion category="pleasure" modality="face" />
      <emotion category="worry" modality="voice" />
    </complex-emotion>

"Simulated emotion masking a suppressed one" would be very handy for political reportage!

All very interesting stuff.  Of course, security agencies will be interested in this too, as they are already into determining states of mind of people at airports (and hassling first-time flyers because they seem as nervous as a theoretical bomber).

It’s actually coming soon to a standard phone near you (especially if you have one of the newer Sony-Ericsson models with a stylus).  Another group associated with the same site that gives you EARL is eMoto, from the Swedish Institute of Computer Science.  If you look at the eMoto page, you’ll see pictures of phones where the SMS text overlays a background with a specific color or pattern.

eMoto is a mobile messaging service for sending and receiving affective messages. The application extends on both input and output channels when sending text messages between mobile phones. The aim is to convey more of the emotional content through the very narrow channel that a text message otherwise provides. Emotional communication between people meeting physically in the “real world” make use of many different channels, such as facial expression, body posture, gestures, or tone of voice, little of this physicality of emotions is used in a similar digital context. In eMoto users therefore use affective gestures to convey the emotional content of their messages which are then translated and communicated in colors, shapes and animations.

You can see the interaction mechanisms here

Now, from browsing around, there’s still the problem of internationalization.  While EARL might describe affect, and a "renderer" might generate culturally appropriate movements of a face (e.g. an Indian smiling head waggle might correspond to a European smiling nod), the Swedes don’t seem to have dealt with the different cultural interpretations of color (e.g. in China, white is for mourning, red is happy rather than angry).

Still, mappings are possible, although guessing from the target country may not always work.

There’s one application I’d like to see:  As EARL provides an "intensity" attribute, as a user I might be able to "tone-down" the emotions of an avatar in Second Life (I’m not a Second-Lifer), but toning down the expressions (or even getting an explanation) of those silly Japanese cartoons some kids watch would be a Good Thing.

How do I express that I’m worried I’ll get depressed and annoyed by overly demonstrative animated emoticons that will be appearing soon?


See Also

  • You might like to check out the emotional text analysis for textual affect available from a German university. Start here, click Next, type in more than 200 words (in English), then hit the Classify button.  It was tested with movie reviews and turned them into "no stars" (negative review) to "four stars" (loved it).

Advertisement

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.