Using data to go Back to the Future, to find the dream job and to, well, discover just about anything on Planet Earth. This week’s big data Roundup takes on some quite epic challenges.

Hunting for strange and interesting datasets but finding a dream job

Mapping the movements of ships in SF Harbour - the big data roundup

Here’s a novel way to land your dream data job. Unearth a treasure trove of strange and interesting datasets and then turn them into mesmerising maps and gifs. That’s how Sam Kronick landed himself a job with geodata giants Mapbox. He first used data from the United States Coast Guard, which shows the off-shore locations of most ships updated by the minute. He then combined this with information from the National Oceanic and Atmospheric Administration on the country’s underwater terrain, before mapping and animating the results. He now hunts for weird datasets full time!

Beating the bookies with big data

Predicting the results of sporting events sounds like every gambler’s dream. Remember what happened in Back to the Future? Now a team from Cardiff Uni are taking on the bookies by using big data collected from thousands of previous results. AI bots are programmed to make bets based on three years worth of data from 100 football leagues around the world, including results and in-play data. Prof Anatoly Zhigljavsky says that while it’s not about the money – but the challenge – one AI bot is generating a “small but steady profit”. While the team is still looking for AI that can win the big bucks, they have detected betting scams through unusual fluctuations in the odds. “Football betting is an interesting example,” says Zhigljavsky, “there is an infinite source of data, lots of uncertainty, and uncertainty is exactly what we like to deal with.”

Find the location of just about anything visible on Google Earth 

Find the location of just about anything visible on Google Earth

A new tool launched this week allows users to find matches for any feature found on Google Earth. The tool – GeoVisual Search – uses geospatial imagery and machine intelligence to allow users to find every snowy mountain peak, sports stadium, wind turbine, or any other object visible on satellite imagery. And to do so is as easy as clicking on one you know about and letting the machine intelligence take over. GeoVisual Search lets anyone run an automatic query on global satellite imagery to find the location of just about any identifiable feature. Try it yourself.

Big data is so powerful, nation states will fight over it

“I think big data is so powerful that nation states will fight over how much data matters. He who has the data can do the analytics and the algorithms . . .” These are the words of Eric Schmidt, chairman of Google’s parent company Alphabet. He was talking at the firm’s annual Google Cloud Next conference – where the firm also announced the acquisition of Kaggle, the fast growing community (read ‘social network’) for data scientists. Schmidt added: “Speaking at the scale that we talked about [data] will provide huge nation-state benefits, in terms of global companies and benefits for their citizens, and so on.”

Mapping the changing refugee crisis

Mapping the crisis of refugees with UN Data

In every corner of the earth ordinary people are forced to leave their homes, often without notice, often never to return. When they cross international borders, these people are called refugees. The Refugee Project is a narrative, temporal map of refugee migrations since 1975. Using UN data, the project shows the shifting and escalating problems, and tells the stories of the major refugee crisis’ of the past 40 years. Sobering yet fascinating.

Posted in Big data, IoT, Media roundup On March 10, 2017 By