Dawn Nafus – Add GPS and Stir?: Some Context Needed for Context Awareness

If you have a mobile phone with GPS in it, you can now sense you location and summon appropriate directions. If you have mobile friend finding software, you are already able to know whether your husband is picking up milk on the way home or your friend is waiting for you at the bar. In the future it will be possible to use an expanded array of naturally occurring datastreams to make lots of inferences about human activity. But is this really the same thing as ‘context awareness’?

In an upcoming industry gathering (ecommmedia.com), I’ve been asked to reflect on opportunities in the increasing use of context aware technologies. Although there are lots of definitions, I see these more or less as technologies that leverage passively collected data rather than relying on direct input by users, often involving sophisticated algorithms to make some guesses about what it means. The result is that the device has some kind of ‘awareness’ of things outside it. Computer scientists define ‘context’ here as the myriad sensor readings possible—lighting, location, presence of other devices, etc. etc.. It is this narrow definition of context that makes life harder for ourselves, because adding more and more sensor readings is not really the same thing as adding more and more context.

From an anthropologist’s point of view, sensors can be pretty useful things when they are interpretable on a human level. But this means that we can’t, as an industry, think solely in terms of adding more and more data in hopes of eventual machine-based cleverness. We will start getting better value out of sensors once we do a better job of understanding how it can actually help people. For example, it turns out that my location is always relative to those I’m communicating it to. My being ‘at home’ means different things to my husband than to my boss than to my distant relatives. Sometimes being ‘at home’ is physically on the road or just not in a certain part of the house, if being ‘at home’ means communicating the idea that ‘I’m not working.’ Or, to those distant relatives, it really means something quite comforting–‘Dawn is doing the same daily rituals we used to do when we were all together’—for them, no need for an action to take place, or opportunity for interruption to be inferred and presented. For the future, this in fact may mean a person uses several apps, each designed relative to a different kind of relationship, and delivering the appropriate kinds of messages. Currently, it seems, the marketplace is skewed towards the generic, the ‘out of context,’ perhaps in hopes that in the name of user empowerment, the customers do all the work of making technologies appropriate for themselves.

It was a huge accomplishment to simply be able to sense location electronically. Now that we’ve done it, it is important to figure out the real context of ‘context awareness’—who needs to communicate what to whom. This is not a matter of simply adding more datapoints but creating occasions for human interpretation and interaction. Otherwise we are back to technology for technology’s sake. EComm participants will debate IMS, mobile web and all things disruption—all of which are important to debate. However, what we actually deliver with all of this is the kind of rethink the industry really needs to have.

Dawn Nafus is an anthropologist Intel, where she conducts social science research to inspire new products and strategies. She holds a PhD in Anthropology from Cambridge University, and was previously a research fellow at University of Essex. She has published widely on communication technology and society in academic journals, and worked with both public policy makers and industry leaders on issues such as widening participation in open source communities and knowledge transfer policies for local economic development. Her recent work has focused on how patterns of mobility and their social meanings affect technology usage. She is also working on a interdisciplinary project combining statistical modeling and social theory to explain cases of accelerated adoption rates at the national level. Her areas of regional expertise are Russia and the UK.

4 Responses to Dawn Nafus – Add GPS and Stir?: Some Context Needed for Context Awareness