Author(s): P.Bharathi, Dr. B. Umadevi
The growth of internet gives more opportunities for the buyers in selecting their products. They have an open forum through which they can make their purchase by comparing the like products by various organizations. The e-commerce platform allows the user to make the business with pos tags by the various types of visitors. The Collaborative and SVM algorithm are applied for predicting user rating behaviour. The author collects the data from the websites of the different organizations. The product of the different organizations is being posted as the likings and disliking as values of stars. The significance of the product may be tagged as the comments posted by the different product users. So many people will contribute same results or feedback. The data mining techniques classifies the similarities and dissimilarities among the tags. It predicts the repeated items and transforms it into text and their density values are calculated for each product. Using the clustering algorithm the product features are validated. The existing Collaborative filtering algorithm is applied to filters the user rating behavior. But it does not give significant result. So we propose the SVM algorithm for predicting the behavior for making the better performance from the available data set.