Abstract

This research presents a unique software based on the programming language Twitter API and R. Twitter keywords are searched to get relevant tweets. Twitter APIs and Rs programming may extract these rich-opinion data sets about the contents of tweets, tweet writers, and tweets. This program has been expanded to geographical location search and post-time search in order to gather more complete Twitter feelings about political and economic problems. A new text preprocessing technique is suggested and being explored for Twitter data. The tweets collected may include a range of information about interference in many languages. This research presented for the first time a hybrid model for the categorization of Twitter sentiment. The performance of the Twitter polarity classification will be improved by combining it with a new feature chosen method based on the NRC lexicon and the classic classification algorithms KNN and Nave Bayes. The findings are assessed and verified.