Text mining based sentiment analysis using a novel deep learning approach

Leveraging text mining for sentiment analysis, and integrating text mining and deep learning are the main purposes of this paper. The presented study includes three main steps. At the first step, pre-processing such as tokenization, text cleaning, stop word, stemming, and text normalization has been utilized. Secondly, feature from review and tweets using Bag of Words (BOW) method and Term Frequency Inverse Document Frequency is extracted. Finally, deep learning by dense neural networks is used for classification. This research throws light on understanding the basic concepts of sentiment analysis and then showcases a model which performs deep learning for classification for a movie review and airline sentiment data set. The performance measure in terms of precision, recall, F1-measure and accuracy were calculated. Based on the results, the proposed method achieved an accuracy of 95.38% and 93.84% for a movie review and Airline sentiment, respectively.