Analysis of users’ Sentiments from Kannada Web Documents

In today's world, there is an explosive growth of data from terabytes to petabytes in the internet. The major problem is not the availability of data but starving for knowledge from the data. Sentiment analysis is an important current research area in the field of web content mining. Sentiment analysis and opinion mining is the study that analyzes people's opinions, sentiments, evaluations, attitudes, and emotions from written language. In this paper, we extend our ideas pertaining to Sentiment Analysis to the regional language Kannada, spoken mainly in Karnataka, a state in southern part of India. We have explored the usefulness of semantic approaches and machine learning approaches, used predominately on English language data set, from Kannada web documents. We found the average accuracy of machine learning approaches to be better than the average accuracy of semantic learning approaches for Kannada data set.