Government offices in Europe and the Middle East are using an analytics tool which can recognize sarcastic comments posted online
Here’s something some social media-challenged Canadian government officials might want to look into.
French software maker Spotter said it has developed an analytics tool that has an 80 per cent accuracy rate in identifying sarcastic comments posted online.
The company, which is represented in Canada and the United States by New York-based Spotter Americas Inc., said the software uses a combination of linguistics, semantics and heuristics to create algorithms that produce a report on a user’s online reputation.
Among the organizations using the solution are United Kingdom’s Home Office, the European Union Commission, Dubai Courts and Air France, according to a report from the British Broadcasting Company.
Spotter said its reputation management solutions combine data from conversations on social media with coverage in online media or traditional media sources to “create a new generation of reputational measurements.”
The solution uses a software-as-a-service platform, dynamic dashboards, analytical models and alert to help users monitor their reputation by keeping tabs of the tone of posts made by clients and customers on company Web sites and social networks such as Twitter and Facebook.
The algorithm is so far able to “reflect various tones” in 29 languages, according to Spotter.
A Spotter spokesperson said one of the most common subjects for sarcasm is bad service. For instance, he said, Air France might have a customer that tweets the airline a “thanks” for a severely delayed flight.
However, Simon Collister, a social media and public relations lecturer at the London College of Communication, expressed doubts that software could accurately detect sarcasm, which tends to be “so dependent on context and human languages.”
The bot threat
Some of the most serious threats networks face today are "bots," remotely controlled robotic programs that strike in many different ways and deliver destructive payloads, self propagating to infect more and more systems and eventually forming a "botnet."