Sentiment Analysis for Hindi Cinema Using Boosting Algorithms

In today’s rapidly evolving world, with ubiquitous access to technology, there are massive amounts of data being generated. This data contains key insights that shape better decision-making. Hence, tools that help us extract such insights from this data are of the utmost importance. Sentiment analysis is one such tool. It helps us determine the emotions behind a piece of text. Although there are many resources for sentiment analysis in English, resources for Hindi are limited. We aim to remedy this issue with our work where we scrape, annotate, and pre-process our own Hindi review corpus from the field of cinema. We propose a novel methodology to perform Hindi sentiment analysis using various boosting algorithms and create a foundation to aid better model and framework selection for vernacular natural language processing tasks.