Classification of Publication Papers by Using K-nearest Neighbor Algorithm

Abstract

Classification is a data mining or machine learning technique used to predict group membership for data instances. Several major kinds of classification method including decision tree induction method, Bayesian networks method, k-nearest neighbor classification method, case-based reasoning, genetic algorithm and fuzzy logic techniques. Classification is the task of deciding whether a paper belongs to a set of pre-specified classes of papers. Automatic classification schemes can greatly facilitate the process of categorization. Categorization of documents is challenging, as the number of discriminating words can be very large. In this paper, we presented categorization of publication papers by applying k-nearest neighbor classification using the Euclidean Distance measure.K-nearest neighbor method is the simplest and most straightforward method among all classification methods. Hence, knearest neighbor method is used to classify different number of nearest neighbors for different categories, rather than a fixed number across all categories in this system.This system is intended to classify different categories from different papers in data sets and to save time for searching papers.