The three V’s definition of big data is the one that most people are familiar with. This denotes data that has the defining properties of volume, variety and velocity, which respectively refer to the amount, the number of types and the speed of processing said data.
However, presenter and write Timandra Harkness, has reverse engineered an acronym to explain big data in another way. The acronym is DATA. See what she did there?
According to her, D stands for dimensions and bringing data together from different areas to see how it might correlate. Harkness said that she illustrated this in her book by speaking to a neurologist, whom she referred to as a brain scientist. The neurologist had taken brain scans images and cross-referenced them with the addresses of the pertaining patients and the weather in their locations. From these multi-dimensional profiles of the patient, he was able to pose the question of whether there was a link between the amount of sunshine the patient received and the progression of their multiple sclerosis.
The first A stands for automatic, as in the automatic collection of data through wearables and IoT devices. To clarify this, Harkness explained that there is a thing called Strava, a social app for athletes to connect with friends and record their workouts. Strava draws time, distance and other pertinent data from fitness trackers, sports watches and smartphones, and creates ranking tables and maps to help runners, cyclists and other fitness enthusiasts to benchmark themselves to friends and strangers.
The T is for time, which I inferred to mean that big data is collected over a period of time. To exemplify how time is an integral aspect of big data, Harkness referred to the work of a New York City researcher. He made a freedom of information request to access New York City taxi data, which he received on a giant hard drive. “He cleaned the data and anonymised it and you can see patterns through time. And this is possible again because data is collected through time, then you can project those patterns into the future and use those projections to make predictions,” said Harkness. She then made a pertinent point, a mathematical model can project but only a human being can use that to make a prediction.
The final A is for artificial intelligence, especially the way it was foreseen by Alan Turing, “the grandfather of AI.” Harkness gave her own definition of artificial intelligence by saying: “We are not talking about a sentient robot. We’re talking about a very sophisticated computer program that can help you draw meaning from data.”
She spoke fondly of the future model of artificial intelligence that Turing described in a paper written in 1950 titled, ‘Computer Machinery and Intelligence’. “He was projecting forward to the idea that you could have machines that would think like a person. He imagined it in quite an adorable level of detail,” said Harkness. Turing said that some downsides of a thinking machine in comparison to a human child were limited utility in the home and the possibility of bullying. He wrote that it couldn’t be ordered to fill the coal scuttle and that “you couldn’t send it to school without other children making excessive fun of it.”
Sticking with the topic of AI, Harkness asked the audience to associate the words artificial intelligence with Turing’s “little robot child, Little Al” instead of Terminator. She recounted Turing’s analogy of a computer program with a mother giving instructions to her child to check daily if her shoes at the cobblers are ready and collect them if they are. “As a description of how a computer programme works, it’s genius because the computer that he is describing there, basically an algorithm is a flow chart for computers,” she said.
Harkness stated that Turing was thinking beyond ‘if this, then that’ directions but instead was thinking of a computer that wouldn’t need detailed instructions of what to do at every stage but a computer that could adapt to new circumstances, learn from what it had done before and respond to very different conditions. “Without that, AI doesn’t work in the uncertain world that we live in.”
She mentioned a robot called iCub that researchers at the University of Plymouth are attempting to teach as if it were a child. Harkness also referred to Google Deep Mind, neural networks, IBM Watson winning the game show Jeopardy, image recognition, trial and error and machine learning.
As interesting as these anecdotes were, the links between the words in the acronym were somewhat tenuous. Joe Bloggs in the street might not find this explanation of big data very clear. Having said this, no doubt the book for which the acronym was created provides context that will enable a better understanding.
For me, the take-away message was the importance of the human in the mechanical loop as “a mathematical model can project but only a human being can use that to make a prediction.”
Timandra Harkness was speaking at an event hosted by Konica Minolta Marketing Services. She is the author of Big data: Does size matter?