PREDICTING ONLINE NEWS POPULARITY USING NEURAL NETWORK

Abstract

In this era with the rapid growth of internet, smartphone and other gadgets, everybody enjoys reading, sharing online news. We can direct relate online news popularity with the number of shares, number of comments and number of likes of that news that means popularity is directly proportional to number of share, comment and likes. In this project, Our Goal is to find out the best suitable model to predict the popularity of online news, using artificial neural network. Artificial neural network have several advantage over machine learning algorithm, artificial neural networks have some interesting properties that made these family of machine learning algorithms very appealing when confronting difficult patter-discovery tasks. Artificial neural network has two kind, one is feed forward unidirectional ANN and other is feedback cycle ANN (Back propagation). ANN with Back propagation (BP) learning algorithm is widely used in solving various classification and forecasting problems our data comes from Mashable, a well-known online news website. Our dataset consist of 40K rows and 15 input parameter and 1 output parameter. Artificial Neural Network turns out to be the best approach for prediction, goal of this thesis to predict the number of share of news article and it can achieve an accuracy with optimal parameters. Our work can help online news companies to predict news popularity before publication.