Tnooz is running a panel about Big Data in Travel at the upcoming New York Travel Festival, where I am a participant.
Since the audience represents a wide variety of travel professionals, it got me thinking about what the “average” travel person would want to learn about Big Data.
NB: This is a viewpoint by Larry Smith, a partner at US-based Thematix.
I quickly concluded that discussion of Hadoop computer clusters, MapReduce data processing, and R Studio statistical analysis would not be appealing.
So what then? Where does everyone play?
The most popular people-generated Big Data files are photographs. We all love pictures, we love looking at them, and we love sharing them.
It’s the entire reason for being on Flickr and Pinterest, but also the actual essence of Facebook and Tumblr, and increasingly Twitter.
Every single web page on the internet has one or more images.
That’s big, that’s popular. And while no precision is possible, I’d bet half of all these billion pictures are within the travel realm.
Unfortunately computers do not understand what the photo is about, what it shows, or the meaning it conveys. A computer does not see, it does not infer meaning.
So how does one make an image fit the Big Data and travel value proposition?
Mostly to enable search and find, we use metatags and key words like Bohemian Hall, Czech Republic, or #NYTF.
The problem is these words lose value at scale - there is literally 183 billion results to a search of Czech Republic, and millions that feature Prague Castle.
A couple of key words do not represent the value, and are not useful in a search.
However, change is afoot and value is being rewarded at search services like Bing and Google.
What’s more, a new startup, Image Snippets, give you the ability to computationally describe an image, and conversely find an image that fits highly refined search criteria.
This is more than "people on street, NYC", this is "foreground happy family, looks Japanese, background jewelry storefront and goth girl, with panoramic perspective, on a dry sunny day, with leaves blowing on the street".
You can now add a story to every picture.
With this capability, every travel professional has a million, if not a billion, image opportunities.
These image opportunities are inspirations and motivations to travel. These images are persuasive tools for friends to follow in your footsteps.
At scale, these images are Big Data and they come from a lot of people in places over time.
You should be a producer of Big Data.
Later on, when you want to consume image data, there are fascinating things such as simple geographic locations and time data found in EXIF (EXchangeable Image File) that every modern digital camera and smartphone produces.
And if you use a system like Image Snippets, you can curate a highly refined library of images easily searched.
The Big Picture of Big Data is to be a producer first, and a consumer whenever you can afford it. You will see the value as soon as you get in the picture.
NB: This is a viewpoint by Larry Smith, a partner at US-based Thematix.
NB2: Digital picture image via Shutterstock.
NB3:New York Travel Festival details.