Proposed Preprocessing Methods for Manipulate Text of Tweet


  • Hayder Mahmood Salman


Social media provides plentiful information to study people's ideas and opinions about events in this world, such as political, economic issues, disasters and others. The tweets that published must study and classify. In a classification system for text mining, an important step is the preprocessing phase. It is often underestimated. This paper uses Twitter data as a case study. The aim of the paper is to propose methods for preprocessing for each text of tweet. In this paper the propose in preprocessing steps that include proposed manipulate Hashtag state method that used to extract another important word to assist in other processing processes and proposed enhancement stemming algorithm to return the entered word to its source without Affix, which are both Suffix and Prefix extensions and final results of the preprocessing process are presented based on number of words after preprocessing.