News Analysis Using Word Cloud

In internet era, one can get news from huge number of sources. However, many news sources are biased in giving more coverage to specific content, persons or party. Eventually, the reader’s thoughts are also influenced by the news source’s biases. In this paper, a method has been proposed to instantly visualize the news topics discussed by various sources on internet. Word clouds make it very easier to decide the biases of a news source. Various algorithms, namely, Porter stemmer, Snowball, Lancaster, Rake, tf-idf, text-rank, and tag cloud algorithm have been tested to effectively extract the key words covered by a news source. Extraction time and count of correctly identified terms have been used as metrics to compare the algorithms. It is concluded that tf-idf is better than rake and text rank algorithm due to its right balance between speed and accuracy.