Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

of

When My Data Actually Becomes My Data Slide 1 When My Data Actually Becomes My Data Slide 2 When My Data Actually Becomes My Data Slide 3 When My Data Actually Becomes My Data Slide 4 When My Data Actually Becomes My Data Slide 5 When My Data Actually Becomes My Data Slide 6 When My Data Actually Becomes My Data Slide 7 When My Data Actually Becomes My Data Slide 8 When My Data Actually Becomes My Data Slide 9 When My Data Actually Becomes My Data Slide 10 When My Data Actually Becomes My Data Slide 11 When My Data Actually Becomes My Data Slide 12 When My Data Actually Becomes My Data Slide 13 When My Data Actually Becomes My Data Slide 14 When My Data Actually Becomes My Data Slide 15 When My Data Actually Becomes My Data Slide 16 When My Data Actually Becomes My Data Slide 17 When My Data Actually Becomes My Data Slide 18 When My Data Actually Becomes My Data Slide 19 When My Data Actually Becomes My Data Slide 20 When My Data Actually Becomes My Data Slide 21 When My Data Actually Becomes My Data Slide 22 When My Data Actually Becomes My Data Slide 23 When My Data Actually Becomes My Data Slide 24 When My Data Actually Becomes My Data Slide 25 When My Data Actually Becomes My Data Slide 26 When My Data Actually Becomes My Data Slide 27 When My Data Actually Becomes My Data Slide 28 When My Data Actually Becomes My Data Slide 29 When My Data Actually Becomes My Data Slide 30 When My Data Actually Becomes My Data Slide 31 When My Data Actually Becomes My Data Slide 32 When My Data Actually Becomes My Data Slide 33 When My Data Actually Becomes My Data Slide 34 When My Data Actually Becomes My Data Slide 35 When My Data Actually Becomes My Data Slide 36 When My Data Actually Becomes My Data Slide 37 When My Data Actually Becomes My Data Slide 38 When My Data Actually Becomes My Data Slide 39 When My Data Actually Becomes My Data Slide 40 When My Data Actually Becomes My Data Slide 41 When My Data Actually Becomes My Data Slide 42 When My Data Actually Becomes My Data Slide 43 When My Data Actually Becomes My Data Slide 44 When My Data Actually Becomes My Data Slide 45 When My Data Actually Becomes My Data Slide 46 When My Data Actually Becomes My Data Slide 47 When My Data Actually Becomes My Data Slide 48 When My Data Actually Becomes My Data Slide 49 When My Data Actually Becomes My Data Slide 50 When My Data Actually Becomes My Data Slide 51 When My Data Actually Becomes My Data Slide 52 When My Data Actually Becomes My Data Slide 53 When My Data Actually Becomes My Data Slide 54 When My Data Actually Becomes My Data Slide 55 When My Data Actually Becomes My Data Slide 56 When My Data Actually Becomes My Data Slide 57 When My Data Actually Becomes My Data Slide 58 When My Data Actually Becomes My Data Slide 59 When My Data Actually Becomes My Data Slide 60 When My Data Actually Becomes My Data Slide 61 When My Data Actually Becomes My Data Slide 62 When My Data Actually Becomes My Data Slide 63 When My Data Actually Becomes My Data Slide 64 When My Data Actually Becomes My Data Slide 65 When My Data Actually Becomes My Data Slide 66 When My Data Actually Becomes My Data Slide 67 When My Data Actually Becomes My Data Slide 68 When My Data Actually Becomes My Data Slide 69 When My Data Actually Becomes My Data Slide 70 When My Data Actually Becomes My Data Slide 71 When My Data Actually Becomes My Data Slide 72 When My Data Actually Becomes My Data Slide 73 When My Data Actually Becomes My Data Slide 74 When My Data Actually Becomes My Data Slide 75 When My Data Actually Becomes My Data Slide 76 When My Data Actually Becomes My Data Slide 77
Upcoming SlideShare
What to Upload to SlideShare
Next
Download to read offline and view in fullscreen.

0 Likes

Share

Download to read offline

When My Data Actually Becomes My Data

Download to read offline

Society is currently going through a phase of having an adversarial relationship with personal data. Our data is gathered by third parties ranging from companies like Facebook and Google to governments and their agencies and although in theory we ourselves own our data, we don’t manage, get value from it, or use it ourselves. The only times we encounter our own data is when we read about abuses of it, or we get confused when we try to understand what GDPR means. One day we will live in a world where we actually own our own data and it will be managed for us, with our interests at heart, by trusted third parties analogous to how banks manage our wealth. Those third parties may increase the value of our data by pooling it, equivalent to banks lending money, and by sharing it with organisations like social media companies, educational institutions, entertainment companies, etc. In such a world we would be delighted rather than afraid, to gather data and to have data gathered about ourselves and used for our benefit. In such a world, what are the data points that can be gathered, what is our digital footprint ? In this talk I will present an overview of what data can, and is gathered by people about themselves. I will cover off-the-self and popular sensors as well as the more unusual and uncommon and as a focus I will give an overview of sleep, how it can be measured and what use that can be. Gathering data about oneself is also known as lifelogging or the quantified self and I will draw inspiration and case studies from the work we have done in the area of lifelogging over the last 15 years. (thanks to Cathal Gurrin for some of the slides).

Related Books

Free with a 30 day trial from Scribd

See all

Related Audiobooks

Free with a 30 day trial from Scribd

See all
  • Be the first to like this

When My Data Actually Becomes My Data

  1. 1. When My Data Actually Becomes My Data Alan F. Smeaton Dublin City University @asmeaton
  2. 2. Who’s Alan Smeaton ? - DCU Professor for +20 years - Founding Director of Insight - Decorated, highly-cited, worked on 00’s research projects, loads of PhD graduates - Background in multimedia analysis, indexing and search - Known for his collaborations
  3. 3. • We have an adversarial relationship with our personal data • Usually gathered by third parties and in theory we own our data, but we don’t manage, get value from, or use it ourselves • We read about abuses of it • We can also gather data ourselves, about ourselves, for standalone niche applications (sleep, steps, energy, screen usage, travel, etc.), or we can do extreme forms of this called lifelogging • Lets have a look at that niche application area, and some of the extreme lifelogging 3
  4. 4. Digital Footprints • Digital Footprints used to be the electronic evidence of a computer user's activity (online, local, etc.), used for debugging • Early search engines logged queries/clicks • Clickthroughs were mined, turned into AdWords, pushed the boundaries of ML, created Google • Now a digital footprint is the electronic evidence of your existence, since everything is now digital and so much is logged by third parties anyway
  5. 5. Digital Footprints • We’re aware we consciously and unconsciously leave digital footprints – conscious ones from • Web searches • Website visits and cookies • Internet connections • Purchases • Social media check-ins, posts and photos • These are what we expect Google et al. to know about us … like all marketing, they use this to segment their market (us) so we get adverts (like me viewing YouTube videos with fingers crossed) • With rich data they segment their market into N=1 by turning footprints into personality profiles
  6. 6. Digital Footprints • So that’s our digital footprints • Some we know about, its obvious, and we accept, we even like it … the Faustian pact • Some we don’t realise, we’re surprised at, but we’re OK with • Some is a bit creepy, maybe crossing a line, intrusive • Our awareness varies hugely – vast majority don’t realise, those that do know, don’t actually know it all
  7. 7. • Most of the time, 3rd parties gather data about us but occasionally we gather data about ourselves, we gather it • That’s called lifelogging 7
  8. 8. Depending on what you want to measure, there is likely to be a device (or an app)
  9. 9. Digitise as much as you can of life experience… for many reasons (memory, health, etc.). Lifelogging Sense and analyse factors of interest through numbers to gain knowledge Using knowledge for self- improvement through experimentatio n Quantified Self Biohackin g
  10. 10. 1928 Bucky Fuller Archive 1980s Steve Mann 2004 Williams (Sensecam) 2010 Quantified- Self 2006+ Memory Studies 2004 G. Bell (MyLifeBits) 1946 Vannevar Bush
  11. 11. Richard Buckminster Fuller • Interested in “a very accurate record of a human being" … so he made himself his own case study .. put everything in and created a very rigorous record … the Dymaxion Chronofile… • He documented his life, philosophy and ideas scrupulously by a diary every 15 minutes, now on display at the Stanford Library. 11 Richard Buckminster "Bucky" Fuller was an American architect, systems theorist, author, designer, inventor and futurist. Fuller was the second World President of Mensa from 1974 to 1983
  12. 12. With more than 140,000 papers and 1,700 hours of audio and video, all stretching to more than 1,400 linear feet of material, Fuller’s life might be the most documented life of all time. From 1917 to his death in 1983 he collected all documentation including (mail, newspaper clippings, drawings, blueprints, models, and even bills. The Dymaxion Chronofile has been at the Stanford University Libraries Department of Special Collections since 1999. There you can pick any day of these years of his life and find out exactly what he was doing nearly to the hour by flipping through a scrapbook.
  13. 13. Vannevar Bush (External Memory) 13 Vannevar Bush was an American engineer, inventor and science administrator, who during World War II headed the U.S. wartime military R&D including initiation of the Manhattan Project. "As We May Think" has turned out to be a visionary and influential essay. https://www.theatlantic.com/magazine/archive/1945/07/as-we-may-think/303881/
  14. 14. https://mediartinnovation.com/2014/06/06/vannevar-bush-memex-1945/ Vannevar Bush (External Memory)
  15. 15. Steve Mann’s Wearcam • Steve Mann (University of Toronto) built a wearable camera called Wearcam (wearcam.org). Steve would wear the camera and it uploaded images to the WWW for others to see. • Mann has been referred to as the "father of wearable computing", having created the first general-purpose wearable computer. Mann has also been described as "the world's first cyborg”. 15 Steven Mann is a Canadian researcher and inventor best known for his work on augmented reality, computational photography, particularly wearable computing and high dynamic range imaging.
  16. 16. 2012 Steve Mann’s Wearcam
  17. 17. Gordon Bell An early employee of Digital Equipment Corporation (DEC), Bell designed several PDP machines and later became Vice President of Engineering, for VAX computers He is the experiment subject for the MyLifeBits project, an experiment in life- logging and an attempt to fulfill Vannevar Bush's Memex. 17 Gordon Bell is an American electrical engineer, pioneer and investor. He is the founder of the MyLifeBits project, an experiment in life- logging based on Vannevar Bush's Memex.
  18. 18. Lifelog Life Experience The individual will have a lifelog / human ledger for many aspects of life experience… activities, experiences, behaviours, information, biometrics… huge volumes of data captured passively. Gordon Bell – the lifelog philosophy
  19. 19. Cathal Gurrin In 2006 he put on a wearable camera, he’s still wears one, every day all day, 12 years and billions of data points later His interest is in multimodal multimedia analysis, indexing and retrieval
  20. 20. Unstructured Lifelog Data 20
  21. 21. Wearable cameras to capture our activities 21
  22. 22. A complete trace of the individual
  23. 23. Know their activities and interests
  24. 24. Know their habits
  25. 25. Know their interests and experiences
  26. 26. And know what they consume
  27. 27. Devices Apple Watch & HealthKit Wahoo Chest Strap Fitbit Activity Tracker / Scales Nokia Activity Tracker / Scales / BP … Aggregation Software HealthKit Microsoft Health Gyrosco.pe We’ll come back to this
  28. 28. Issues with Wearable Camera Images Huge levels of redundancy 28
  29. 29. Issues with Wearable Camera Images Vary in quality from unusable to photo-album 29
  30. 30. Issues with Wearable Camera Images Don’t typically have a salient object Capture the hands of the wearer 31
  31. 31. CATHAL’S LIFELOG Autographer Panasonic 4K Google Glass Moves App Withings BASIS Watch Reporter RescueTime LoggerMan Camlapse MyTracks OpenPaths CameraPhoneInstagram Media Lifelogging Biometrics/ Activity Lifelog Information Access SMS Backup Youtube Log VoiceRecorder WebServices Swarm Location Digital Paper PDF Anno. Web Pages Consumption e-book/mag Last.fm NarrativeClip 3 2
  32. 32. Volume Growth over a Decade
  33. 33. Which allows for the development of various types of interface. 3 4
  34. 34. Events with colour coded minutes… showing the dominant colours 3 5
  35. 35. Minute-by- minute segmentation and summarisation of life activities
  36. 36. Segmentation of raw data into units such as events or moments. These can be enriched automatically with metadata, increasing their value. Events are analogous to our episodic memory and can be segmented based on many forms of data.
  37. 37. Quantified Self Personal Insights Data-driven Health Augmented Wellness Behaviour Change Enhanced Security Population-wide Analytics Augmented Community Augmenting Human Memory Nomenclators Augmented Memory Enhanced Productivity & Education Enhanced Interactions Rich Sharing and Reminiscing Augmented Cognition Some (Individual) Use Cases for Lifelogging Health & Wellbeing Memory & Cognition
  38. 38. Quantified-Self Analysis Self-discovery Reflect Contextual Reminders Remind Sousveillance. Protection of me and bystanders Protect Find an item from the digital self Validate a memory Contextual support Answer Reminiscence Therapy Social applications Reminisce Digital Agents acting on our behalf, during life and after Represent The most interesting aspect is the potential for memory support, where the lifelog works in synergy with your own memory. Adapted from Abigail J. Sellen and Steve Whittaker. 2010. Beyond total capture: a constructive critique of lifelogging. Communications of the ACM. 53, 5 (May 2010), 70-77.
  39. 39. • Back to OTS or regular quantified self rather than extreme lifelogging • Lets have a deeper look at one form of lifelogging … sleep 40
  40. 40. • Sleep is “active”, our brains do not shut down and are almost as active, cataloging memories • Sleep 5 stages – wake, relaxed wakefulness, light sleep, deep sleep and REM sleep • Starts from N1, goes through N2 to N3 (deep) and then back up towards REM sleep, there is an ordering • If you sleep for 8 hours, you have 5 full cycles Sleep Explained
  41. 41. • Each phase has characteristics – REM – body paralysed, HR, RR increased, body temperature drops, vivid dreams, brain active, towards latter end of the night, memory consolidation – N1 – conscious of surroundings, hypnic jerks – N2 – brief arousals, decreased HR, RR – N3 or deep sleep – slowest HR, RR, difficult to wake and then groggy Sleep Explained
  42. 42. • Insufficient sleep makes you more stupid, fatter, unhappier, poorer, sicker, worse at sex, more grumpy, more likely to get cancer, Alzheimer’s and to die in a car crash ! • Recent years have focused on this, we’re more aware because we ourselves can now measure it • [ Orthosomnia — a preoccupation with perfecting one’s sleep data ] Why is sleep important ?
  43. 43. • Sleep labs record EEG, body and eye movement, HR, HRV, RR, Oxygen saturation, etc. • They pool all these into a polysomnograph for an holistic overview, which looks like … Measuring Sleep (Properly)
  44. 44. 1. Phone apps 2. Wrist-worn accelerometer devices 3. Movement radar or sonar 4. The Ōuru ring Each of these uses a subset of sensors Sensing Sleep – Our Options
  45. 45. Presented as a Hypnogram
  46. 46. • How ? They record .. – Movement (in the bed, under pillow) – Microphone (listening to your breathing) – Sonar – inaudible frequency emitted and listened to, just like bats ! • There are many available, some freebies, some paid … SleepScore, Sleep Cycle, Pillow, etc. 1. Smartphone Apps
  47. 47. • FitBit, Jawbone, Withings, LARK, etc. • These “just” do movement but directly, so more accurately than phone apps 2. Wrist-worn Accelerometers
  48. 48. • BiancaMed -> ResMed -> S+ was first to market • Very low levels of transmitted radio-frequency power acting as sonar • Contactless, measures motion, plus room temperature, brightness and noise level (from phone, yes, phone is listening as you sleep) 3. Sonar / Radar
  49. 49. 4. The Ōura Ring • Includes accelerometer, gyroscope, temperature, and heart rate • HR sampled using IR spectrum at 250 Hz so able to do much more than wrist-worn • Contactless battery charge lasts 7 days, data capacity 6 weeks • Low power Bluetooth download to phone
  50. 50. • Consumer grade wearables are not sleep labs, they use a proxy for an orchestra of sensors ! • Each individual sensor will have errors (movement, sweat, etc.) • The algorithms to compute “sleep efficiency” are opaque and proprietary • I compared S+ with Ōura for me over 8 weeks and … Accuracy of sleep tracking ?
  51. 51. • 56 days, but S+ not continuous (travel, S+ is not travel-friendly) – I missed 1, 1 and 7 days • Correlation is (only) 0.147 (rises to 0.16 when blanks eliminated) • And because ranges may not be normalized, the visualization reveals … S+ and Ōura
  52. 52. S+ and Ōura - 56 days continuous – 0.147
  53. 53. So What ? • Its all a bit … unsatisfactory • I’m feeling a bit … let down … it shouldn’t be like this, there’s something not right • I’m actually feeling … trapped .. its my data, about me, but I can’t fix this, for me, I can’t leverage benefit, • Not everyone is like me though, and would want to do this, but maybe you’d want somebody to do this for you, like a wealth manager or stock broker • There are some benefits at population level 56
  54. 54. Almost all vendors who allow us gather individual data, get value from pooling anonymised data • Fitbit – 150B hours of HR data, statistics from millions of sleep nights • 23a DNA testing – diseases, long-lost relatives • Strava – global heatmap of runs/cycles • Jaw – pinpointing earthquake epicentres ! Population-level Analytics from Pooled Individual Data
  55. 55. Resting Heart Rate is an excellent indicator of overall health. https://finance.yahoo.com/news/exclusive-fitbits-150-billion-hours-heart-data-reveals-secrets-human-health-133124215.html
  56. 56. Strava Global Heatmap
  57. 57. 2014 – Northern Calif. 6.0M Earthquake % people who woke up correlates with distance from epicentre
  58. 58. Back to my problem of … • Unsatisfactory … let down … trapped 63
  59. 59. • Open up about their algorithms and allow a “normalisation” of sleep metrics for interoperability and device/vendor independence • Include “errorbars” in their outputs • Allow exporting of our data, our original raw data, with those error margins Sleep Sensing Vendors ought to …
  60. 60. • Imagine a world where our data is managed for us by trusted third parties analogous to how banks manage wealth • They may pool it, like banks lend money • They may share mine with organisations like social media companies, educational institutions, entertainment companies, etc. but only if I say so, and I might get paid, or I might get better services • I would be delighted rather than afraid, to gather data and have data gathered about me for my benefit • Most people would not micromanage data, most people don’t micromanage wealth • We have seen the data points that can be gathered Utopia ? Naivete ?
  61. 61. • What’s the appetite for this ? • Is there even a demand ? • Are there green shoots against surveillance capitalism ?
  62. 62. • DuckDuckGo, StartPage, SearX.me, DisconnectSearch, MetaGer, Quant, all trace-free search engines Trace-free services
  63. 63. Devices Apple Watch & HealthKit Wahoo Chest Strap Fitbit Activity Tracker / Scales Nokia Activity Tracker / Scales / BP … Aggregation Software HealthKit Microsoft Health Gyrosco.pe Remember this ?
  64. 64. 6 9 • Consolidates health data from iPhone, Apple Watch, and third-party apps • Activity, Sleep, Mindfulness, and Nutrition • “You are in charge of your data” • “The Health app lets you keep all your health and fitness information under your control and in one place on your device. You decide which information is placed in Health and which apps can access your data through the Health app”
  65. 65. dacadoo
  66. 66. dacadoo
  67. 67. dacadoo
  68. 68. dacadoo
  69. 69. dacadoo
  70. 70. dacadoo • Revenue stream is corporate • Employer signs up as a service to employees, employees use it, employer gets healthier (more productive) staff
  71. 71. • I don’t see a spanner breaking up Surveillance Capitalism, too traumatic • I see a slow creep towards – Surveillance Socialism (don’t like that) – Surveillance Democracy (don’t like that either) – Sousveillance • Not happening overnight, generational

Society is currently going through a phase of having an adversarial relationship with personal data. Our data is gathered by third parties ranging from companies like Facebook and Google to governments and their agencies and although in theory we ourselves own our data, we don’t manage, get value from it, or use it ourselves. The only times we encounter our own data is when we read about abuses of it, or we get confused when we try to understand what GDPR means. One day we will live in a world where we actually own our own data and it will be managed for us, with our interests at heart, by trusted third parties analogous to how banks manage our wealth. Those third parties may increase the value of our data by pooling it, equivalent to banks lending money, and by sharing it with organisations like social media companies, educational institutions, entertainment companies, etc. In such a world we would be delighted rather than afraid, to gather data and to have data gathered about ourselves and used for our benefit. In such a world, what are the data points that can be gathered, what is our digital footprint ? In this talk I will present an overview of what data can, and is gathered by people about themselves. I will cover off-the-self and popular sensors as well as the more unusual and uncommon and as a focus I will give an overview of sleep, how it can be measured and what use that can be. Gathering data about oneself is also known as lifelogging or the quantified self and I will draw inspiration and case studies from the work we have done in the area of lifelogging over the last 15 years. (thanks to Cathal Gurrin for some of the slides).

Views

Total views

403

On Slideshare

0

From embeds

0

Number of embeds

12

Actions

Downloads

0

Shares

0

Comments

0

Likes

0

×