Text Mining Predictive Modeling Algorithm for classifying Attitudes of customers with Accuracy Estimation

Abstract

Social media is one where everyone can share their views on different aspects. These attitudes or views can be captured and analyzed to afford solution for predefined problems. The mounting need for entity insights and the industrial challenges currently facing the field motivated to come up with a system that helps in knowing the customers opinions on particular aspect with its estimation. The paper introduced a Text Mining predictive modeling algorithm to categorize or classify the test data. This paper studies a Bayesian modeling approach for multi-class sentiment classification. With the help of Dimensionality reduction technique and a classifier model, trained data is tested on different sample sets. We tested the classifier against a 70-30 split of training and test data sets with different threshold values. Cross validation methods are used to calculate the accuracy of the classifier. This value is also compared with accuracies obtained for the same data sets and different threshold values. The empirical results show the efficiency of the proposed approach. The paper helps organizations to assess the campaign's triumph or study how to adjust for greater deed that can have an undeviating bang on returns and competitor offerings.