Recently in Crowdsourcing

Paul Marks, senior technology correspondent

(Image: James Whitaker/Digital Vision/Getty)

Looks like 30 is the age at which everyone grows up on Twitter. It's nigh on impossible for computers to guess the age of people on the network by analysing the type of language they use once they leave their 20s, researchers in the Netherlands have found.

Jim Giles, consultant

(Image: Westend61/Getty)

Luis von Ahn is an ambitious guy. He's best known for putting a twist on CAPTCHAs - the squiggly fragments of text that websites deploy to beat spammers' automated software - that he called reCAPTCHA. Von Ahn realised that when we solve a CAPTCHA we could be helping to decipher text that book-digitisation software has tried and failed to transcribe. Google liked the idea so much that it bought reCAPTCHA in 2009, and now uses the technique to help power its book-scanning project.

Von Ahn's current project may be his most ambitious yet. Duolingo is a website where people can learn a language for free and help to translate web content at the same time. Not everyone thinks it will work, but von Ahn now has evidence that one part of the challenge - the learning bit - is performing as hoped.

Hal Hodson, technology reporter

Next time you're stuck for someone to play Pictionary with, don't despair. Now you can always get a game against a computer and, reassuringly, you'll probably win.

Sketching is something nearly any human can do, even if we use inaccurate representations like outsized ears on rabbits, or stick figures to simplify complex objects. What's remarkable is that most of us can recognise these dodgy sketches, despite their divergence from reality. Pictionary is based on these very abilities.

Now machines can play too, after researchers from the Berlin Institute of Technology in Germany and Brown University in Providence, Rhode Island, trained a computer to recognise human sketches. Instead of trying to teach the computer a set of fundamental lines and shapes, they trained it using machine learning on drawings crowdsourced over the internet, with the computer making a best guess at what category a sketch should go in, based on the layout of its lines.

Lisa Grossman, reporter

(Image: Lisa Grossman)

Tae Kim steps back from the piano and shakes his hands limply. "If I ever do this again, I'm going to have to remember to take a break," he says.

Kim, a graduate of the New England Conservatory, had been playing the piano in the MIT Media Lab's "Opera of the Future" lab for three and a half hours at the lab's spring meeting earlier this week. But there was no sheet music on the music stand. Instead, Kim watched colourful bubbles on an iPad that displayed what people watching along online wanted to hear.

The piece was "an experiment in collaborative improvisation", says composer and lab director Tod Machover. People at home could listen to ten clips of music from Bach to the Beatles and rate their preferences. If listeners said, "This is nice, but I'd like a little more Radiohead and a little less Schubert," Kim had to respond by improvising in real time.

Jacob Aron, technology reporter

(Image: Geoffrey Robinson/Rex Features)

Social media such as Twitter and the BlackBerry Messenger (BBM) network were partially to blame for the riots which swept across the UK last summer, but now a new analysis of British tweets reveals that the public mood was turning long before the first riot took place.

Thomas Lansdall-Welfare and colleagues at the University of Bristol measured the emotional content of 484 million tweets from between July 2009 and January 2012, organising them into four lists: joy, fear, anger and sadness. It is not the first time researchers have analysed tweets in this way, but the Bristol team wanted to see if particular events were reflected in the emotions on Twitter, rather than just general sentiments.

Their analysis found that certain annual events always result in a spike of emotion each year. For example, there is always a large increase in joy on Christmas Day, along with smaller spikes on New Year's Day, Easter and Valentine's Day. Other emotions show similar spikes at regular points, such as an increase in sadness on Halloween.

Paul Marks, senior technology correspondent

(Image: Shoula/Getty Images)

A couple of fake reviews probably won't kill (or unfairly promote) somebody's product/hotel/restaurant - but sustained postings of fake reviews by large groups of colluding opinion spammers might succeed. To the rescue come software engineers from the University of Illinois in Chicago and Google - who this week reveal a software algorithm that can pinpoint review threads in which groups of fraudsters are trying to wrest control of online sentiment.

In a paper to be delivered at this week's World Wide Web 2012 conference in Lyon, France, 

Paul Marks, senior technology correspondent

(Image: KeystoneUSA-ZUMA/Rex Features)

A web-based filtering system that lets emergency services extract vital information from Twitter will be demonstrated at next week's World Wide Web 2012 conference in Lyon, France. Developed at Delft University of Technology in the Netherlands, the Twitcident system aims to help the services cope with the street chaos that might reign during urban terrorist attacks, extreme weather events, wildfires, earthquakes and industrial emergencies such as chemical explosions and spills.

The idea? When people are tweeting about, say, which roads are blocked by fleeing traffic, crowds, collapsed buildings or fallen trees, its useful to give paramedics, firefighters and police that information so they can plan other routes. Tweeted and retweeted warnings from members of the public are just perfect for that, but finding the relevant data is tricky. "Making the information accessible and findable in any given crisis context is a non-trivial scientific challenge," say the Delft team on the Twitcident home page.

Niall Firth, technology editor


(Image: Ricardo Garcia Vilanova/AFP/Getty Images)

Earlier this month, an unnamed woman in the village of Sahl Al-Rawj, Syria, left the safety of her hiding place to plead for the lives of her husband and son as government forces advanced. She was captured and five soldiers took turns raping her as she was forced to watch her husband die.

Her shocking story - officially unverified - is just one of many reports of sexual violence against women that has come out of Syria as fighting continues between government forces and rebels. Now a crowd-mapping website, launched this week, will attempt to detail every such rape and incident of sexual violence against women throughout the conflict.

The map is the creation of the Women under Siege initiative, and uses the same crowdsourcing technology developed by Washington DC-based Ushaidi, which is also being used to calculate the death toll in the recent fighting.

Paul Marks, senior technology correspondent

(Image: © Google -  © 2011 Cnes/Spot Images, DigitalGlobe, GeoEye)

If you like puzzling over the meaning of the Nazca lines in southern Peru, crop circles in central England or poring over maps of Area 51, then you'll like this. Digital mapping fans today noticed some wild and wacky forms of indeterminate function in China's barren wastes in the Gobi desert.

The bizarre shapes vary widely. Two of them are grids that look a little like street floor plans without buildings and are around 2 kilometres long by just over a kilometre wide. Conspiracy theorists pounced immediately. Were these some sort of alien markings or the remains of a lost civilisation?

Jacob Aron, technology reporter

(Image: Zoological Society of London)

A new iPhone app that lets you spot leopards, elephants and other animals in the wild could also help conservationists identify new species or determine whether populations are under threat.

The Instant WILD app, released today by the Zoological Society of London (ZSL), connects users to motion-sensitive cameras set up near animal habitats in Kenya, Sri Lanka and Mongolia. New images are sent to the app in real-time when a camera senses a nearby animal, which then tasks users to pick which species it belongs to. You can also follow a specific camera to get alerts whenever a new image is captured, or access a version of the app online.

It would normally takes days for a conservationist to sift though these animal images, but ZSL hopes that crowdsourcing the results will help speed up the work. "By asking people to help us identify species through the app, we are turning wildlife conservation into the massive team effort that it needs to be," says ZSL conservation director Jonathan Baillie.