Abstract: Social networks have become a rich and large repository of information about us as individuals. Due to the growth of social network usage, grouping of like-minded users for further processing is a major research issue. Some social networks even allow users to identify others based on their user interests and tags. Interest of the users can be found with few options in web known as tagging and rating. Rating is the method that finds the opinion of users about the items. Geographical location is becoming a major factor that influences the interest of users. This study describes the performance of employing unsupervised learning techniques such as Support Vector Machine (SVM) and naive bayes model for grouping of the unanimous users with and without location information. Location-based unanimous user identification provides better results in grouping the users. Naive bayes and SVM classifier results are compared through error rate and confusion matrix. The analysis proves that SVM provides better performance in terms of accuracy when compared with NaiveBayes classifier. The classified location based user data is clustered using self-organizing maps for better recommendations.
J.S. Kanchana and D. Sujatha, 2016. Support Vector Machine based Classification and Clustering for Identifying Unanimous Users. Asian Journal of Information Technology, 15: 4150-4159.