Data analytics could put a stop to fake news, says IBM

Jason Burns of IBM Ireland insists that automation is becoming ever more important in the data space

IBM Ireland analytics architect Jason Burns has said that data analytics could be used to combat fake news, election interference and fraud, while warning that they are growing rapidly.

Burns was speaking on the topic of big data at the monthly Cork Chamber business breakfast, as reported by the Irish Examiner.

He said that data analysis relies on automation; humans are not equipped to deal with the massive amounts of information being generated today.

"It's an impossible task for someone at a manual level - instead, you take the knowledge gained from such people and build it into the data-analytics model. There is no other way to do it."

He added, "You need automation just to be able to process it, to be able to work out where the information is coming from, and to segment the people disseminating that.

"That is just one aspect — who is spreading this information, what is the origination of it and how do we stop the proliferation of it."

There are ethical and even bias issues facing analysts who want to use data in this way, however. "We cannot stop people expressing themselves freely," said Burns. "While it may be fake to you, it's not to some. How do we agree what is fake?"

Burns was optimistic about the near future, however. He was especially positive about moves by Facebook and Twitter to fight the fake news phenomenon:

"Facebook, Twitter — the amount of posts and tweets per day is growing exponentially. Facebook has set up giant teams to combat election interference by rogue nations.

"But it is all about setting up the teams to build the models, and informing those models of what is going on. You see it with it with tax authorities like Revenue, trying to build analysis models of who is likely to defraud," he said.