An Online Review Summarization on Abstractive Method Using LSTM

Abstract

In this modern era online services are increased faster than before. This is because online services reviews are very much effected and thousands of review posted every day. This study makes an online review summarization. Two approaches are broadly use in text summarization i). Extractive. ii) Abstractive. In this study we work on abstractive text summarization and used Deep learning method. our proposed method works on Amazon Food fair service customers’ reviews. In this method we execute sequence to sequence model using encoder and decoder architecture. The input and the output length remain same in sequence to sequence model. To solve this problem Encoder and decoder are used where the input and output sequence are of different length. Jointly differing the Recurrent Neural Network(RNNs), Gated Recurrent Neural Network (GRU) or Long Short Term Memory are preeminence as Encoder and Decoder. In this novel study we use LSTM (long short term memory) components. The main difficulties of this paper is training section. this is because most of data are used to train. Finally, this work help user to get a summarized review on Amazon food service.