Classification of Newspaper Article Classification by Employing Support Vector Machine in Comparison with Perceptron to Improve Accuracy

Aim: The study's goal is to carry out text classification analysis.Materials and Methods: Accuracy is analyzed for text classification. It helps to organize the data entry data. The perceptron is mainly classified into two parts of data. Inaccurate newspaper articles have a low text classification accuracy rate and the classification is performed in various sectors in text analysis. Analysis of news articles is used to identify the text analysis whether the text in the article is negative or positive impressions are neutral and to identify the accuracy comparison between SVM and perceptron. Classification of text is performed on SVM statistical distribution of (N=42), acquired by statistical power at a rate of 80%. Results: The SVM's accuracy is 82.71%, which is higher than the Perceptron's (PER) accuracy of 75.86%. The significant value for accuracy is 0.196 (p>0.05). Conclusion: SVM performs better inaccuracy with the comparison of the perceptron. We can predict that the support vector machine which was proposed had better accuracy than the existing system perceptron.