Earlier this week on Monday, I had the privilege to sit in on a lecture by Professor Laeeq Khan in social media analytics course being taught in the Scripps College of Communication. He spent time informing us about sentiment analysis. So what exactly is sentiment analysis? This concept is defined as a way of measuring what people think; positively or negatively. You might also hear individuals referring to sentiment analysis as opinion mining, opinion extraction, and sentiment mining. Sentiment analysis is important in analyzing social media content today because it can be a great source of information along with helping public relations and marketing departments. It may not always be accurate (roughly 60 to 70% accuracy rate on average) but it does provide vital insights.

A hybrid approach in which the human element is combined with the machine-level processing, seems to be an effective way to deal with the accuracy issue. A few monitoring services have already started combining automated sentiment analysis with an evaluation of human analysts. Human analysts can see what criteria creates a positive or negative mention. Humans are also better than machines when it comes to deciphering deeper meanings.  In other cases, companies can control costs of human analysis by reviewing a sampling of posts (Comcowich, 2016).

Overall, sentiment analysis programs can be beneficial or useless depending on the type of questions that are being answered. It is not only vital to consider the goals around which an analytics strategy is built but a careful consideration of the pros and cons of establishing a media monitoring and measurement plan is needed. A hybrid approach seems to be an effective option for companies at this point.