SentiWordNet for Indian Languages

The discipline where sentiment/ opinion/ emotion has been identified and classified in human written text is well known as sentiment analysis. A typical computational approach to sentiment analysis starts with prior polarity lexicons where entries are tagged with their prior out of context polarity as human beings perceive using their cognitive knowledge. Till date, all research efforts found in sentiment lexicon literature deal mostly with English texts. In this article, we propose multiple computational techniques like, WordNet based, dictionary based, corpus based or generative approaches for generating SentiWordNet(s) for Indian languages. Currently, SentiWordNet(s) are being developed for three Indian languages: Bengali, Hindi and Telugu. An online intuitive game has been developed to create and validate the developed SentiWordNet(s) by involving Internet population. A number of automatic, semi-automatic and manual validations and evaluation methodologies have been adopted to measure the coverage and credibility of the developed SentiWordNet(s).