Classification and ranking of trending topics in twitter using tweets text

Abstract

Every day millions of twitter user tweet their views on various topics using short messages of 140 to 280 characters length. Here we monitor the public opinion in twitter thereby identifying trending topics. To identify the current trending topics on Twitter, tweet classification is done. Tweet classification is a process of classifying the tweets based on the topics using the keywords of the tweets as a feature. Tweets are extracted from twitter using Twitter API. In the existing system, trending topics in twitter can be identified by hash tags. This work is going to categorize the tweets into seventeen classes and identify the trending topics by using words. Dataset used in this work is the list of tweets taken for two to three days from Twitter. By classifying tweets based on sensitive words for a certain time, top k most trending words from a topic will be obtained that convey the dynamic trend.