Wednesday, April 22, 2015

Sentiment Analytics Results – Why They Are Inaccurate?

Now that you got your own brand new sentiment analytics application! With its help, you can parse through innumerable social data feeds to identify what is happening around your brand, who is saying what about your product, etc.

If you look at the chronicle data, sentiment analytics applications come a long way; from identifying just words to handling ambiguous phrases. The depth of digging has increased along with the breadth of the scope and reach, so far. As a result, the sentiment analytics applications nowadays can handle the following parameters: words, phrases, double negatives, etc.

Currently, the most sophisticated applications are coming up with geography-based slang customization features also. With this feature, now, you can first set geography, upload keywords related to that geography, and then run the sentiment analytics application. This way, the application analyzes the parsed words or phrases based on that respective geography’s slang. 

However, even after adding all the above features, still something is missing because of which the application is not displaying the accurate results. So, what is it that is missing? Answer: Context. 

People, while communicating with each other, say and understand words or phrases or sentences based on context. A word or sentence or sentence can be understood in two different ways in two different contexts based on the way the person utters it.  

Now, the billion dollar question is how to add ‘context’ to the application so that the operator opts for it before running any sentiment analysis to get accurate results!  

What do say?

No comments:

Post a Comment