Data Science and Mobile Devices: Joined at the Hip
As number of wearable and IoT devices have soared, there is a new big data market emerging at horizon. Here, we have scratched the surface of this giant and got few connections between mobile devices and data science.
By Fred Sandsmark (OpenText Analytics), special for KDnuggets.
What qualifies as a mobile device? Believe it or not, we had a brief – and surprisingly illuminating – debate at my workplace the other day that started with this question.
Let me provide some context: Our team is gearing up to produce technology demonstration projects and other content that connects Big Data analytics to mobile devices. (Both literally and figuratively – more on that in a bit.) Some of this content is slated for a webinar scheduled for July 9; also, mobility is an umbrella theme for our corporate blogs for the next few weeks.
We needed to set some parameters for the content the team would create. After some discussion, one individual suggested that we focus the content toward wearable devices – things like smartwatches and fitness trackers. After all, we already have created some excellent proofs of concept related to wearables (one of our engineers hacked his Prius and sent its data to his watch), and IDC predicts that the wearable market will grow 173.3 percent in 2015. He argued that if we focused on wearables, we would be skating to where the puck will be (as hockey fans like to say).
But others worried that focusing too much attention on wearables might exclude the more traditional devices, such as smartphones and tablets, that are critical to our community today. One individual added that most smartwatch functionality still requires a tethered phone, and most fitness trackers must sync to a computer, meaning wearables are an adjunct to mobility rather than devices unto themselves.
After more discussion, we decided to keep the definition of mobile devices broad, while focusing some specific time and attention – particularly in the webinar – on the wearables market.
You may wonder how this debate relates to data science. There’s one overarching parallel that I see: In both worlds – mobility and data science – we as technologists and practitioners need to keep our eye on the future while simultaneously staying relevant to what real people actually do every day. We can and should debate which nascent concepts and technologies will become commonplace in the future, but we also need to understand and address the challenges that our users face today and the tools they use to address those challenges.
There’s another, more direct connection between data science and mobility: With increasing frequency, data science provides the critical link between Big Data – in all its volume, variety, and velocity – and data that is usable and actionable on a mobile screen. (This is what my colleague Allen Bonde likes to call “Small Data.”) Data science, in this context, is not entirely about algorithms and ideas; it also relates to the tools and techniques we use to make those connections happen.
Here are some important factors I have identified in the connection between data science and mobility:
- APIs: Increasing diversity of mobile devices – and their decreasing size – highlights the ongoing need for, and importance of, APIs. Ideally, APIs used by Big Data systems must have enough flexibility that they can address devices not yet invented.
- Alerts: Alerts are one key to a successful wearable experience, but it’s becoming harder and harder to get notifications right, as this article on the NeimanLab website explains.
- Two-Way: Mobile devices aren’t just receivers of data; they’re also generators of data. Any Big Data system related to mobile devices – and wearables in particular – ignores this truth at its peril.
- Security: A mobile device is a potential security weak link. (Actually, both the device itself and the network connection between the device and its supporting analytics systems are potential points of exploit.) Addressing this truth is vital when mobile devices are linked to Big Data systems containing valuable or secret information.
What do you see as the connections, both conceptual and practical, between data science and mobility? Countless interesting projects are out there, in areas ranging from medicine to logistics to marketing to meteorology. I’d love to know what everybody is working on.
Image courtesy of Flickr user General Physics Laboratory (GPL), https://www.flickr.com/photos/genphys/
AUTHOR BIO: Fred Sandsmark (@FredSandsmark) is a Marketing Content Writer at OpenText Analytics (formerly known as Actuate). Along with creating marketing collateral and social media content, he produces the Data Driven Digest, a weekly roundup of great data visualizations.
- Sisense 2015 predictions for BI, Big Data
- Hot or Not: Data Science Trends in 2015
- Interview: Mac Devine, CTO, IBM Cloud on the Conflux of Cloud, IoT & Big Data