Decoding the buzz words as a freshman - #1 (SA)

     Welcome to the first post of "Decoding the buzzwords as a freshman" series. Starting as a freshman majoring in computer science can be very intimidating in the current scenario as the talk of the town (LinkedIn especially) are computer science jargons like Big Data, Cloud Computing, Data science, IoT, machine learning, bitcoin, computer vision, reinforcement learning, Sentiment Analysis etc. The worst part is we are expected to know what these terms mean when we actually shouldn't be. Nontheless, an insight into what these topics entail can make lives easier as undergrads learning physics, chemistry and mathematics when all that the world seems to be telling us is to learn coding. (which ofcourse is essential) An understanding of the prerequisites of the hot selling skills in the market will help us find meaning in what we study in the first two years of college. Hence, this series is to vaguely enlighten you and instigate an interest in you for these "cool" areas of technology so that you can flaunt it among your other department peers. 

If the world population has to be classified into two categories, one has to consumers and the other, sellers. It's as simple as can be. The world works by the seller selling his product to the consumer by charging an amount which the seller uses to make more products and it's a cycle. What if the seller could know what the consumer wants first hand and makes products accordingly? That should be amazing! We have surveys and reviews for this purpose. Given the world population and the amount of people buying and reviewing products online, manual scavenging of issues, opinions, likes, dislikes about the product from the reviews and surveys is a near to impossible task. We need a faster and non manual technique for this purpose. 

What if there could be a technology that reads the thousands of opinions online and figures out what the writer intended to mean about the product/service/person/event? That should be super amazing! What if we develop a computational procedure that identifies the underlying emotion behind the text/sentence/document (which is our comments, tweets, blogs, reviews), comprehends the consumer opinion as a whole, consolidates the varied opinions and displays them as data ( could be graphs, pie charts)? Wouldn't it make the lives of the sellers super easy? We have developed algorithms that do exactly this. Yes! this. It has been given the name Sentiment Analysis. Using this data, the sellers would be able to correct their mistakes, alter their marketing and branding strategies and products and improve their businesses. Consumers too wouldn't be put through the same fault identification process again. It's a win-win situation. The best part, it saves up hell lot of time for everybody involved. This process is also known as Opinion Mining. Doesn't it sound cool guys? Well, before that, I would like to know what "cool" means to you. I'll be performing Sentiment Analysis on it. Wink!

How does it work? It uses Natural Language Processing (NLP), Machine Learning (ML) and other rules driven algorithms. You don't want to read further, do you? Yes, these terms are daunting. These concepts in turn incorporate concepts from calculus, differential equations, matrices, probablity etc to frame their algorithms. Don't they sound familiar? Along with a sound knowledge of coding, these concepts can be easily picked up. 

Essentially, Sentiment analysis answers binary questions which has answer Yes/No apart from "neutral". "Did people like the movie?" " Did people like our new mobile?" "Are people talking about this event?" "Popular opinion on the Budget 2k19?" "What will be the poll results?"  and the likes of it. You should know by now that this has applications not only in businesses (Business intelligence) but also in policy making, Sociology, psychology, e-commerce (obviously) among others. 

Microsoft's Azure Text Analytics API, Google cloud's Natural Language API, IBM's Watson among others make it available to the users to work their hands on Sentiment Analysis through their cloud services. Try your hand at these and let me know what you learnt. Happy learning!