FinalProject

The basis for this paper is a machine learning and data mining dataset called Online News Popularity found in UC Irvine’s dataset repository [1]. The dataset comes with a number of features already defined and a success metric. The success of each article is defined as the number of shares the news article receives. The goal of this advanced project is to discuss and evaluate the addition of several new predictive attributes to the Online News Popularity Dataset, related to the title of the articles. These include the popularity rating of words, word embeddings, and parts of speech. The research question is whether any of these additions increase the accuracy of predicting the success of an online news article and if so which features have the most benefit for the model.