Moving Beyond Data Visualization | Frank Evans | TEDxOU

[Applause] what do you see more than a million pixels on screen most of them various shades of orange and brown and yet in a moment your brain can put together a cohesive picture of the information that it needs a million dollars worth of technology every computer vision machine learning trick ever invented cannot as reliably as quickly and as efficiently get to the same conclusion put together the same picture and yet your brain can take these millions of individual pieces of information build the individual patterns put them together into a cohesive model about the world around you and say to you in no uncertain terms danger run this is what our brains do they take collections of individual facts and they build around them a way of understanding the world I'm a data scientist and that's kind of a fancy buzzword term for someone who uses computer tricks to try to replicate some of the same things I take a collection of individual facts or data and I try to build from them a way of understanding the world a way of creating knowledge from those facts and systematically doing so is referred to as data analysis and today this is how we communicate with the world this the buzzword de vivre around this particular object is a dashboard now while the individual data visualizations haven't really changed in the past 75 years even though the artistic quality has markedly improved the way that we communicate is still this same approach this is seen not only as the pinnacle of the way to communicate the knowledge that we've learned from data from this collection of facts it's also seen as its terminus this is how we know we're done this is how we are trying to get to this is the thing that we're trying to accomplish once we're here we've done our job we've answered the question I don't want to just answer individual questions anymore I want data science and data analysis to be able to build explorable worlds that take advantage of our humanity the build to work within the capability of humanity to explore and to collaborate and to engage humanity in its best way and the way that this works is through data applications a date application is anything but static it allows a person to walk around in a world to see something maybe new to pick something up to understand something from a different perspective to see something new and a completely familiar light or to see something familiar in a completely different light to investigate something you didn't even know that you wanted to understand to be able to interact and to collaborate to contribute your wisdom your capability to contribute your humanity into data science into your ability to understand the world around you I think it's appropriate that we refer to these as dashboards because as dashboards they're like the dashboard in your car you glance at it from time to time to answer a specific question but how fast am I going do I have enough gas is the engine running at the appropriate speed or are the oil and the water both still there but they aren't the purpose of the car the purpose of the car is to go somewhere the purpose of the car you you don't drive a car with your eyes affixed on the dashboard you drive your car with your eyes out the windshield exploring the world around you I don't want to build a better dashboard I want to build a windshield and I want to begin with what is possibly the central grand challenge of humanity medical research so medical research has largely unchanged over the past several decades an experiment is designed data is collected it's processed it's analyzed it's written into an article it's published in a journal someone else can read that article and understand what's in that journal and yet if you start to tear those pieces apart what we are talking about is a given journal is a collection of individual facts that are right to be explored this particular gene is associated with this protein this other protein when it is in abundance causes a reduction in this particular syndrome and the associated metadata around it this particular experiment was done on a hundred subjects all of them were mice this other experiment was done on twelve subjects they were human it's been replicated three times that is an incredibly rich data set that needs to be capable of exploit and it would be easy very easy to look at some of these to look at this type of data to summarize it in something like a dashboard answer a few basic simple questions but to do so would be lie what could be possible if a researcher could truly explore the overarching world that was created we took an incredibly tiny sliver of medical research and built a data application around it specific to only a few hundred articles restricted within neurophysiological research around post-traumatic stress a stack of research maybe yay high if you were to print it out which we didn't but no researcher remotely has time to read all of that much less be able to understand it and explore it and yet we started with data visualizations and made an explorable world where one person may want to know what is directly connected and what is directly found someone else may want to know what is indirectly connected through pathways of unrelated or seemingly unrelated concepts while one person may be interested in understanding species at which an individual research topic was conducted someone else may actually be interested in establishing the replicability to understand causality the point was not that an application was capable of answering those questions it's that it didn't know those were going to be the questions asked it simply allowed exploration and allowed the person that was using it to go and pick up a piece of data and look at it through the lens of their domain familiarity it allows somebody to answer questions whether or not they even knew what they really wanted to do specific purpose or no Harebrained Schemes passing ideas they all pet theories even could have the same capability that only years ago would have required hours hundreds and thousands of hours of relentless focus and now these things can be asked easily and vividly only because we were able to trigger humanity's capability and want for exploration it eight applications are only about exploration they're also about another very powerful aspect of humanity which is collaboration and for inspiration around the capability of engaging humanity's collaboration I want to turn to probably the most successful the most famous collaborative data application in history Wikipedia now Wikipedia as a data visualization is incredibly easy to understand it's just text you read like a book but as a data application its capability was around facilitating effective collaboration the underlying technology just allows multiple people to create and edit documents that's all it does and yet perhaps it's fitting that the first thing we did with it was create one giant document to collect and organize the entirety of central human knowledge now Wikipedia on its own is impressive enough and there are these fantastic visualizations that have been put together how if Wikipedia were printed out like a standard encyclopedia and the volumes it would be the size of a small motel but I don't want to I want to put Wikipedia in its entirety to the side for just a moment and I want to concentrate entirely on one tiny corner the part of Wikipedia related to star trekkin now I'm going to deviate for just a moment I want to tell you about the first edition of the encyclopædia Britannica when the Encyclopedia Britannica was first put together it took years of editing and writing to bring together a number of sources from science history and literature together into what they referred to as a dictionary of all human knowledge the first edition when it was published was just shy of 2,400 pages the part of Wikipedia currently related only to Star Trek is approximately 12% larger III can feel you I can feel you I promise saying to yourself how depressing that is what a waste of our potential of our attention of our capability but I am NOT going to let this story off the hook that easily we built a collaboration capability we built an ability to cognitively collaborate and organize knowledge that was so efficient and effective that one of the things we did with it was we catalogued an entire base of human knowledge larger than the entirety of collected human knowledge just a few generations ago about a relatively trivial television show but we did it together in our spare time for free that is not depressing that is inspiring now whether or not you are willing to grant me the decisively generously way I would need to refer to the Star Trek Wikipedia portal as a Grand Challenge of humanity I will unapologetically label as such the ability to cognitively collaborate in such an efficient and effective manner that it could exist our brain centrally are capable of creating knowledge from fact creating an understanding of the world from data and as we systematically do this we will be able to take on challenges everything from the Grand Challenges of humanity all the way down to the daily mundane challenges of modernity when we can engage the best of humanity with technology we can solve anything so Marty McFly's father was right when you put your mind to it you can accomplish anything and humanity when it is engaged and when it is capable of exploring when it's capable of collaborating this is what humanity does best humanity by its nature is social we love to work together to solve problems big and small humanity is curious we love to learn about new things no matter whether it is cataloguing exactly how many red shirts were killed in the original series 43 or understanding the intricacies of the human mind humanity is ambitious in all things let's run with that thank you [Applause]

Channel: TEDx Talks

Views: 15,592

Rating: 4.7837839 out of 5

Keywords: TEDxTalks, English, Technology, Data, Innovation, Visualization

Id: WkhJqhKtdf8

Length: 12min 51sec (771 seconds)

Published: Mon Mar 12 2018