When we read this afternoon on the BBC that the people behind French social media analytics tool Spotter had claimed they had developed a sarcasm detection tool, it caught our attention somewhat.
When it comes to using social media monitoring tools for sentiment analysis, results are a mixed bag to say the least. On their own, these tools are next to useless, says Simon Collister in the BBC’s article. You always need an analyst versed in communications strategy to interpret the data.
In my experience, the most effective uses of tools on social media monitoring are largely around tracking for absolutes. For example mentions of words, time of day, type of content or reach. The value of analysing this kind of data can be immense. It can fuel marketing, reputation management operations, R&D and numerous other business functions.
However the use of software to detect the tone of conversations online at scale, ie sentiment analysis, is rarely accurate enough to be if any use. And for high profile brands or government organisations, detecting a sudden shift in sentiment could be a valuable trigger in maintaining a competitive advantage.
So no sarcasm intended, we’re reaaaally interested in testing out Spotter, as this is a missing link in even the better quality tools in the social media monitoring toolkit right now.