Big Data Sensemaking

This week for Thought Vectors we are to read Vannevar Bush’s 1945 Atlantic essay As We May Think and choose a nugget. This means to:

Take a passage from the reading that grabs you in some way and make that passage as meaningful as possible.

creative commons licensed ( BY-ND ) flickr photo shared by frau-Vogel

Choosing just one nugget was difficult. My list of things I want to explore and discuss in this article are vast and although I tried to sort them by discipline, they are intertwingled (Thanks, Tom for introducing me to that absurd word)

The scientist in me wants to dig into the physics, chemistry and mathematics but that associative trail goes down an infinite rabbit hole.

I was struck by Vannevar’s concern for information overload and how he identified the need for curation but then realized that I’d read a very good analysis of that already by Maria Popova over at Brain Pickings (Thanks, Gardner for retweeting that).

In the end, I chose this just this one line:

“For mature thought there is no mechanical substitute” (p.3)

The timing is interesting as a bot has just possibly passed the Turing test and disguised itself as a 13year old boy.

One of my favourite Radio Lab episodes Talking to Machines describes how a man unknowingly fell in love with a computer chat bot and how a therapy bot was proven to be alarmingly successful. What I found most interesting was not that the artificial intelligence was particularly good or complex but rather humans were drawn to these machines because all they did was listen. We love sounding boards and mirrors, places to learn about ourselves through the presence of others.

In Vannevar’s vision, he wants us to be constantly recording and data keeping but he imagines each individual being responsible for their associative trails. His dream is that we can share our trails with each other.

In our current reality, it is certainly all being tagged and recorded, but do we have full access to our trails? Even if and when we do have access, do we take the time to create meaning from our data? We are constantly inputing data into The System but how frequently do we examine them to make meaning for ourselves about our experiences?

Without our own contextual interpretation of how we came upon a concept and the paths our thought vectors travelled, all we have is raw meaningless data.  Algorithms can determine the data’s relevance and indeed, Amazon, Google, Facebook, Netflix, Delicious, Twitter are all busy determining who I am and what I like based on the data I willingly provide to them daily.

Spurious correlations was mentioned a few times just last week, which probably isn’t statistically significant and just a coincidence but it was enough to make me rethink big data in the absence of human interpretation. Fascinating and ridiculous correlations exist between divorce rates and margarine sales, drowning and Nicholage Cage films, Space launches and sociology doctorates. Data don’t lie!

George Siemens talked about these hilarious spurious correlations during his #apereo14 talk in Miami on Wednesday: Learning Analytics and Sensemaking. He contended that automated tasks that can be reproduced infinitely at scale (movie clips/ autograded quizzes etc) cannot be the future of higher education and the value of a university experience comes from the pedagogy from the instructor.

creative commons licensed ( BY-NC-SA ) flickr photo shared by giulia.forsythe
Similarly, analyzing learner data must be made sensible within a context by a human.

Shortly afterwards, I came across this great article by Maciej Cegłowski, The Internet Remembers Too Much which both addresses the Vannevar Bush request of remembering everything but serves as a warning signal that we’ve let algorithms and large centralized data centres not only control the remembering but also the meaning and interpretation of the data.

Both examples show that we’ve accomplished the masssive data collection and abilty to index, search, curate, share. What’s missing for many, myself included, is the recall, review, reflection.

What I respect a great deal about many of the Thought Vector participants is that they make reflection a part of their daily practice. Now if I could only do that with a little more efficiency of language and time.

---

 

Associative Trails – primed and ready for more details about a “concept experience”

“experience and reflect on a particular exercise related to a key concept in the essay. The idea here is to turn concepts into experiences–in other words, to take a key “dream” from the essay and make it something you do, something you make. For each essay, we’ll specify the experience and set up the parameters”

Physics – Chemistry

• ELECTRONS!
• thermionic switch (vacuum tubes / Diodes? / Edison effect)
• current / capacitors
• fibre optic
• lasers electrostatic printing (dry photography?)
• nanotubes
• radio waves
• frequency spectrum

Technology – Hardware

• Laser printer
• magnetic steel – hard disk drives
• Monitors – CRT, LCD
• 802.11 wifi
• Wearable devices: GoPro, Google Glass

Computer Science – Mathematics

• Machine Language (universal language)
• Input-Output
• Logic
• Array
• Decision tree
• Regular expressions
• Big Data analysis – data mining – Spurious correlations  – SQL query / Facebook graph
• Wikipedia
• meta tag / hashtag

Library Science – Information Literacy

• Record Keeping
• Indexing
• Open access
• Reproducing content without mention of copyright restrictions

Humanities – Education – Social Science – Psychology – Neuroscience

• recall
• memory augmentation
• personal agency
• privacy
• Critical thinking
• Selection
• Reflection
• Interpretation / Sense making / Connecting / Associating
• Analysis
• Collaboration
• Synthesis

Pop Culture

• SteamPunk – miniature mechanical levers accomplishing complex computations
• Matrix  – direct input without sensory experience

This list is not exhaustive!