“Predicting and Evaluating the Popularity of Online News”
The Online News Popularity data set from the University of California – Irvine Machine Learning Data Repository is provided for your use as well as two related journal articles, “Predicting and Evaluating the Popularity of Online News” by He Ren and Quan Yang, and “A Proactive Intelligent Decision Support System for Predicting the Popularity of Online News” by Kelwin Fernandes, Pedro Vinagre and Paulo Cortez.The URL for the data set description is: http://archive.ics.uci.edu/ml/datasets/Online+News+Popularity#., the task is to use this one dataset and compare the results of these three different machine learning techniques to evaluate which of those provides the best results.Because you are assessing the “market share” based on different measures of popularity, in this case the best results will be the results that provide the best understanding of what drives market share. This is different than determining exactly what the market share is.