The K-means with mini batch algorithm for topics detection on online news

Online media is the most important media for accessing a wide range of information, such as news. Nowadays, there are many news agencies publish digital news via online media. The popularity of the online news makes the increasing volume of available news. This leads to the necessity of automated methods for news analysis, i.e. topics detection. One of The topic detection methods is the K-means algorithm. However, this algorithm is slow for big datasets. Therefore, the mini batch approach is used to reduce the computational time. Our experiments show that the computational times of the mini batch approach is much faster for the comparable accuracies.