An Efficient Data Preparation Strategy for Sentiment Analysis with Associative Database

Abstract

Sentiment analysis is a process of categorizing and determining the expressed sentiments. It provides an explicit overview of extensive mass sentiments about particular subjects. Sentiment analysis involves various challenges because expressed opinions and sentiments contain an immense number of anomalies. Data preparation is a prerequisite assignment that can deal with those anomalies for sentiment analysis. This paper represents an efficient data preparation strategy for sentiment analysis using the associative database model. The efficient data preparation strategy involves three subtasks, such as eliminating non-sentimental sentences, eradicating unnecessary tokens, in addition to extracting vocabulary, and arranging that vocabulary uniquely through the associative database model. The experimental results show that the performance of the proposed data preparation approach is comparatively efficient. A comparison with some existing sentiment analysis approaches demonstrates that the accuracy of sentiment analysis has been comparatively improved and enhanced by integrating the proposed efficient data preparation strategy into it.