How sentiment analysis is used in the real world
The memo is out on sentiment analysis — didn’t you get it? Many large entities with spare resources to invest in the technology have already taken the dive, and smaller entities want in. Intel, IBM and even Twitter itself are already using tools to scan social media, news websites and other sources to make sense out of chaos and understand what’s being said about their brands and products online.
And why wouldn’t they? It’s quickly becoming clear that sentiment analysis is the future. In the business arms race, those with sentiment analysis are quickly leaving behind those who haven’t been able to use it.
Take Starbucks as an example of a company that’s using sentiment analysis right. There is an average of 10 tweets about Starbucks every second. Yes, every second. In order to effectively connect with customers, address poor experiences with the brand and form future marketing and development strategies, Starbucks needs some way to interpret this deluge of feedback in a constructive way.
For Starbucks, sentiment analysis does the work that no human employee would be fast enough, or patient enough, to do. The tools at their disposal are highly adept at understanding the intent of all those tweets and compiling them into statistics that help the executives make important decisions. Filters also allow them to sift through the information, producing public opinion of cleanliness standards at a specific Starbucks location or this year’s Christmas cup in specific cities or states.
Similarly, a company like Uber could use sentiment analysis to scan Web 2.0 platforms and get public feedback on price, safety and service. For nearly 15 years people have felt the incessant need to vent their frustrations and give voice to their satisfactions on social media, like screaming into the void. Well, it turns out someone is listening and it isn’t the NSA (well, not only the NSA).
And businesses aren’t the only ones using sentiment analysis to their advantage. Even NGOs, like the American Cancer Society (ACS) are using the technology to figure out how to increase awareness and promote participation.
“It’s nice to have people telling us what they want to hear and what they want to know instead of having to go ask them,” ACS research analyst Liz Keck told the Business Insider.
Her experience working with sentiment analysis for the ACS highlights an important aspect of sentiment analysis and one that should give businesses an incentive to get on board sooner rather than later: with a mixture of machine learning and a hint of human guidance, a good sentiment analysis system can learn and improve on itself with time. As Keck told the Business Insider, this gradual improvement leads to greater nuance and understanding; for example, being able to determine that a social media post that reads “Kill cancer” is actually positive and not negative despite being made up of two very negative words.
And it isn’t just the technology itself that’s improving, it’s also the way it’s being delivered. A new ICO by the name of Senno is finally making sentiment analysis accessible to small entities and even individuals who lack the computing resources and deep pockets of the big business players that have been using the technology until now. This has been achieved by tying Senno’s sentiment analysis platform to blockchain and giving third-parties the tools they need to easily develop their own apps to run on the platform.
It’s the dawn of a new age — the age of sentiment analysis.
