A novel text mining approach based on TF-IDF and Support Vector Machine for news classification

With the development of weblogs and social networks, many news providers share their news headlines on different websites and weblogs. One of the main text mining topics is how to classify news into different groups. This study aims to classify news into various groups so that users can identify the most popular news group in the desired country at any given time. Based on Term Frequency-Inverse Document Frequency (TF-IDF) and Support Vector Machine (SVM), a news classification method was proposed. The proposed approach is comprised of three different steps: 1) text preprocessing, 2) feature extraction based on TF-IDF, and 3) classification based on SVM. The proposed approach was evaluated using two BBC datasets and five groups of 20Newsgroup datasets. The classification precisions were obtained as 97.84% and 94.93% for BBC and 20Newsgroup datasets respectively. These are very desirable results in comparison with other classification methods.