Social media plays an important role in capturing the thoughts of people in their own representation of sentences. Twitter, the popular microblog, traps abundant rich information in it, in terms of short texts posted by users. Processing of such colloquial and informal sentences require a specific preprocessing technique that can alleviate its understanding and analysis. Understanding these natural language sentences to produce a desired result is a challenging task. The paper aims at analyzing, identifying and explaining the advantages of twitter specific preprocessing technique and find out if it performs equally well with the established baseline preprocessing method. Experiment is carried out for classifying tweets to assess the performance of preprocessing techniques. The results indicate the effectiveness of twitter specific preprocessing technique and its competence with the baseline text preprocessing technique. © 2019 Procedia Computer Science. All rights reserved.