Voting Classification Approach for Sentiment Analysis of Twitter Data

This research work focuses on performing SA (sentiment analysis) from the text data. The sentiments are analyzed in different stages such to pre-process the data, extract the attributes and classify the data. Initially, SVM (Support Vector Machine) is implemented to analyze the sentiments. Subsequently, the integration of SVM is done with KNN (K-Neural Network) to classify the data. 5 diverse datasets are executed to test this integrated model. Eventually, the voting classification technique is put forward in order to analyze the sentiments. In this technique, GNB (gaussian naive bayes), RF (random forest) and SVM algorithms are integrated. The introduced techniques are implemented on Python. The outcomes are evaluated on the basis of several metrics including accuracy and execution time. The accuracy acquired from the introduced approach is superior and the execution time is lower in comparison with SVM model to analyze the sentiments.