Big Data, Data Analytics and Data Science | BI-BIGDATA.com

Recently sentiment analysis has gained much attention and interest to huge proportions which is also being fuelled by the many social media platforms mushrooming each day. And sometimes we actually do not understand how this is going to help? But before this we need to understand what sentiment analysis is and how it really works?

Sentiment Analysis is a research topic at the crossroads of information retrieval and computational linguistics concerned with enabling automatic systems to determine the human opinions from web content with the help of Natural Language Processing. Sentiment Analysis is process which gathers the latest inputs from human source or several human sources and uses it to determine the general opinion of person, place or the thing. All comment boxes available over the internet sites are technological sources of sentiment analysis which helps business owners, managers to find strengths and weaknesses of their business. It’s a process to determine the attitude of an audience with respect to some topic. It facilitates an ability to monitor interactions and to identify and categorize drivers of dissatisfaction between customers, buyers and partners. Also known as opinion mining, opinion extraction and emotion detecting. Sentiment analysis is innovative new approach to customer satisfaction surveys. Sentiment analysis is all about understanding what your customers, buyers and partners are talking about your product/services online.

Social sentiment analysis is the use of social media like Twitter, Facebook, blogs and comments to understand the wisdom of the crowd. To achieve this, what we do is we streamed the twitter live data stream by putting in certain keywords about a certain subject matter and stored in NoSQL database like MongoDB which supports semi-structured or unstructured JSON documents. Once we got all live tweets about subject matter then we tried to understand from Natural Language Processing what people are saying about that specific topic by using Python which comes really handy in text analysis. A lot of the efforts we put in to understand how to filter out sarcastic tweets text, what’s positive, what’s negative and what’s sarcastic and also understand the hundreds of opinion that people use on twitter. This is very useful for business owners to understand what the buzz is all about and it also helps you to understand where you should be putting your marketing dollars. What is reaction to your social media campaigns for this reaction to your advertising campaigns? This is just beginning to be really critical part of the decision makers to understand the role of social media over the web in the next ten-fifteen years for their predictive analysis.

The ability to look at million tweets/comment content over the web in a couple of hours about a given subject gives you insights about what people are really thinking. It is like having a million person focus group and that’s different…!!  The cool thing about these new tools is that you have a focus group of a million people and that’s I think much more representative and much more useful. I see it transforming into real business requirements.

Now let’s talk about Natural Language Processing, without this sentiment analysis is not effective at all. Natural Language Processing is analyzing and parsing of natural language which human speaks. Its computer aided text analysis of human language and the goal is to enable machines to understand human language and extract meaning from text. It is a field of study which falls under the category of machine learning and more specifically computational linguistics. There are lot of challenges in passing through and understand what’s going on in human’s mind with his language.

Some of the areas where sentiment analysis can be used by business in their day to day activities like fraud detection in insurance & banking, track promotions and campaigns in retail, fundamental and technical analysis in trading, travel domains to name a few.

Hope this helps..

Sandip