A Novel Approach for Classification of Microblog Tweets

Nowadays, if something happens, then many tweets are posted. Peoples are generally influences what everyone else are saying. A person observes a thing according to their natures, whether they are positive or negative kind of person. When these persons tweet other persons influenced by their thoughts. These tweets contain the situational and non-situational information. Situational tweets may have attributes such as helpline number etc. While non-situational tweets may have the information about the charity etc. Therefore, the classification of tweets into these two categories are much required to extract the meaningful information. In this paper, a new methodology is proposed and implemented to classify situational and non-situational tweets. The present work is based on POS tagger, weighted parameter normalization and normalization of tweets. A comparative study is also presented in this paper in terms of F-measure. The results are satisfactory.