This week there are our six key takeaways from Ad:Tech, wristbands that tell about customer’s real feelings, big data delusions and disillusions, and the ‘New Deal on Data’ ownership.
This Week in Context
Your Weekly Update on All Things Context, October 24 2014
We had great & productive time at Ad:Tech London, showcasing how the Argus Labs real-time moment detection and audience profiling can help with audience segmentation. The conference also offered plenty of interesting talks and panels. Read our six key takeaways on advertising technology.
“If you can program a computer, you have direct access to the deepest and most fundamental ideas in statistics.”
John Rauser, Chief Data Scientist at Pinterest
The executive dashboard is on the way out
1. Wristbands let marketers know consumer’s real feelings
For Saatchi & Saatchi, London-based Studio XO used sensor wristbands to gauge real-time emotional reaction of 2,300 people through capturing their biometrics data such as heartbeat, breathing and body temperature. Watch the results here. A similar experiment at South By Southwest gathered individual movement, sound levels and temperature feedback using the Lightwave wristband. The DJ used this feedback to adjust lightning and track selection, but much more is possible with this ‘crowd engagement data’.
A human DJ is likely to also be great at (en)gaging their audience, and should not concern himself too much with analytics. Yet such wearables do make sense for crowd control and safety, audience-controlled experiences and when your DJ is an algorithm.
2. Machine-Learning Maestro Michael Jordan on Big Data Delusions and Big Data Winter
A reality check and some sobering opinions on big data false positives, brain-like chips and the failure of neural realism from one of the world’s most respected authorities on machine learning. Jordan predicts a ‘big-data winter’, when the hype is over. “The field will continue to go forward, because it’s real, and it’s needed. But the backlash will hurt a large number of important projects,” he warns.
In ‘Artificial Intelligence is as elusive as ever‘, Joab Jackson also has a look at the current state of AI, helped by big data, and the improbability of an AI-winter and the need for more and richer predictive models.
3. Watson gets involved in clinical trials
Yet, as Jordan correctly points out, there’s specific fields where data and self-learning machines can do wonders. For example, when IBM’s Watson looks at health data, he proofs to be a valuable additions to the medic team, sometimes even spotting decease evolution before the human specialists do. Watson also takes charge of clinical trials by analyzing the medical records of patients with breast, colorectal, and lung cancer and suggesting appropriate matches with clinical trials. Soon, Watson might even be speaking to patients directly, as efforts are being undertaken to make him communicate information in a way that is actionable for them.
4. What if companies don’t own all the data?
Big data and the internet of things evolutions both assume that all the data being generated by internet companies and devices scattered across the planet belongs to the organizations collecting it. But what if it doesn’t? Alex Pentland, professor at MIT talks to Scott Berinato about his proposal for a set of principles and practices to define the ownership of data and control its flow; the New Deal on Data.
5. Redressing the balance between Android and iOS in mobile ad targeting
With 85 percent of the global smartphone audience using some flavour of Android, and iOS having only a 12 percent penetration, Julian Smith, head of innovation and strategy at Fetch, questions why mobile advertisers still think iOS first. Whilst iOS users are still more receptive and responsive to mobile advertising, the Android platform now allows for more precise targeting and is experimenting with tracking users from in-app ads to the mobile web. With a broader adoption of mobile data plans, the ‘Android Engagement Paradox’ is disappearing, and Android now surpasses iOS in mobile ad traffic and the ad revenue gap is closing.
A Work of Art & Tech
How do you visualise the thrill of driving a Porsche around on the track? Using Hexoskin shirts to measure breathing, acceleration, heart rates and other biometric data in near real-time, and then visualising this data as a ‘visceral flow of pulsating, glowing shapes and colours’.
Read more about the project by Porsche, TRAQS and Sosolimited in the Atlantic.
(Similar projects we have featured previously where driving data is recorded and used to engage with the driver on a more artistique level are Lexus’ Art is Motion and Volkswagen’s Play the Road – respectively painting a chauffeur’s portrait based on his driving style and surroundings, and generating music to match your driving style.)
Papers, Talks and Research
- Little Data, Big Data and Design at LinkedIn (user experience, design, big data, paper)
- How to identify gender in large datasets, ethically and responsibly, a comprehensive review (gender metrics, ethics, blog)
- Measuring emotion: Exploring the feasibility of automatically classifying emotional text (emotion detection, linguistics, dissertation)
- Learning to Execute (neural networks, paper)
- Neural Turing Machines (neural networks, paper)
- Word Translation Prediction For Morphologically Rich Languages With Bilingual Neural Networks (neural networks, translation prediction, paper) and a dutch article about the research: ‘Nieuwe methode helpt Google Translate aan betere vertaling‘.
Interested in ‘smart sensor technologies’? The 5th edition of the Intelligent Sensor Networks (ISN) conference takes place November 4 in Eindhoven (NL). Experience cooperative driving, learn about sensor networks for context-aware smart buildings, and intelligent mobility and transport technologies. Full ISN programme.
Enjoy the reads, and have a great weekend!
And if you haven’t done so yet, kindly consider subscribing to the Week in Context here.