AN ENSEMBLE APPROACH FOR SENTIMENT CLASSIFICATION: VOTING FOR CLASSES AND AGAINST THEM

Sentiment denotes a person's opinion or feeling towards a subject that they are discussing about in that conversation. This has been one of the most researched and industrially promising fields in natural language processing. There are several methods employed for performing sentiment analytics. Since this classification problem involves natural language processing, every solution has its own advantages and disadvantages. Hence mostly, a combination of these methods provides better results. Various such ensemble approaches exist. The objective of this work is to design a better ensemble approach that uses a complex voting method, where classifiers are given rights not only to vote in favour of classes but also against them. This in turn will give chances to the algorithms that are weaker in classifying a sentence toward a particular class but better at rejecting it. The performance of the ensemble is compared to the individual classifiers used in the ensemble and also the other simple voting ensemble methods to verify whether the performance is better compared to them. The designed ensemble is currently implemented for sentiment analytics. This can also be used for other classification problems, where generalization is required for better results.