Posted byLeila Haddou and Will Franklin Sunday 16 March 2014 18.00 GMT The Guardian
In this digital age, even journalism is being automated. Now over to GUARBOT for the news ...
Is robot journalism the future? It's not quite titanium machines in trenchcoats door-stepping celebrities or buzzing about in newsrooms, but media organisations are increasingly looking to developers to come up with smart ways to incorporate computer algorithms into the daily grind of the news industry. We wanted to test this out, and challenged Will Franklin, a Guardian developer, to create our very own GUARBOT – in a day. Which, we'll admit, was a bit optimistic.
There is a logic at work here, though. The mundane task of trawling through wire copy to spot a newsworthy item could be seen as a waste of resources, especially if all that's required is straight reportage of facts and figures. Surely we should just let computers do the work, while humans get on with more investigative and analytical pieces?
Forbes.com already uses an artificial intelligence platform provided by the technology company Narrative Science to generate automated news from live data sets and content harvested from previous articles. What makes it possible is that business news content tends to be formulaic and data-heavy, listing places, stocks and company names. The LA Times, meanwhile, uses robots to report on earthquakes: the organisation relies on an algorithm that pulls in data on magnitude, place and time from a US Geological Survey site.
But will robots ever be able to write news stories with a distinct angle or provide analysis? In a study earlier this year, Swedish academic Christer Clerwall found that readers couldn't tell much difference between computer-generated and human-written sports articles, with the robot version said to score highly on description.
It has to be said that our rudimentary experiment in using algorithms to structure a simple article fared less well. We thought quinoa would be a suitable topic, and the program was tasked with trawling through a Guardian database for relevant terms from articles on the same subject. Then it was a matter of feeding a set of rules into the program to build a structure for the article. Here's what we got back:
The crime-ridden family of quinoa has taken US by storm this month. According to Peru, New York has confirmed that quinoa is more story than anything else they've ever seen. Quotes from top Yotam Ottolenghi eaters suggest that "crop" is currently clear top, possibly more than ground black pepper. Experts say both Salt and University need to traditionally grow to strengthen a common solution. Finally, it is worth slightly rattling that this article was peeled until it made sense.
Perhaps we won't be seeing a Polly Toynbot or George Monbot any time soon.
source; http://www.theguardian.com/media/shortcuts/2014/mar/16/could-robots-be-journalist-of-future/print
Is robot journalism the future? It's not quite titanium machines in trenchcoats door-stepping celebrities or buzzing about in newsrooms, but media organisations are increasingly looking to developers to come up with smart ways to incorporate computer algorithms into the daily grind of the news industry. We wanted to test this out, and challenged Will Franklin, a Guardian developer, to create our very own GUARBOT – in a day. Which, we'll admit, was a bit optimistic.
There is a logic at work here, though. The mundane task of trawling through wire copy to spot a newsworthy item could be seen as a waste of resources, especially if all that's required is straight reportage of facts and figures. Surely we should just let computers do the work, while humans get on with more investigative and analytical pieces?
Forbes.com already uses an artificial intelligence platform provided by the technology company Narrative Science to generate automated news from live data sets and content harvested from previous articles. What makes it possible is that business news content tends to be formulaic and data-heavy, listing places, stocks and company names. The LA Times, meanwhile, uses robots to report on earthquakes: the organisation relies on an algorithm that pulls in data on magnitude, place and time from a US Geological Survey site.
But will robots ever be able to write news stories with a distinct angle or provide analysis? In a study earlier this year, Swedish academic Christer Clerwall found that readers couldn't tell much difference between computer-generated and human-written sports articles, with the robot version said to score highly on description.
It has to be said that our rudimentary experiment in using algorithms to structure a simple article fared less well. We thought quinoa would be a suitable topic, and the program was tasked with trawling through a Guardian database for relevant terms from articles on the same subject. Then it was a matter of feeding a set of rules into the program to build a structure for the article. Here's what we got back:
The crime-ridden family of quinoa has taken US by storm this month. According to Peru, New York has confirmed that quinoa is more story than anything else they've ever seen. Quotes from top Yotam Ottolenghi eaters suggest that "crop" is currently clear top, possibly more than ground black pepper. Experts say both Salt and University need to traditionally grow to strengthen a common solution. Finally, it is worth slightly rattling that this article was peeled until it made sense.
Perhaps we won't be seeing a Polly Toynbot or George Monbot any time soon.
source; http://www.theguardian.com/media/shortcuts/2014/mar/16/could-robots-be-journalist-of-future/print
No comments:
Post a Comment