Abstract—The quality of a product can be determined by
consumer reviews of the product because consumer opinion is
in general honest and sentimental to the product. The process of
opinion extraction is called opinion mining. This process is done
to determine the tendency of the reviewer to the reviewed object.
Deep Learning is a recently developed opinion extraction model.
This model is widely used for achieving performance in Natural
Language Processing. This work classifies the book review on
Amazon.com into positive reviews or negative reviews using a
combination of Convolutional Neural Network (CNN) and Long
Short Term Memory (LSTM) algorithms. Simulation results
show that the performance of this algorithm is approximately
65.03% for testing data and 99.55% for training data.
Index Terms—Book review, deep learning, long short term
memory, opinion mining.
I. Mukhlash, A. Z. Arham, F. Rozi and D. Adzkiya are with the
Department of Mathematics, Institut Teknologi Sepuluh Nopember, Kampus
ITS Sukolilo-Surabaya 60111, Surabaya, Indonesia (e-mail:
imamm@matematika.its.ac.id).
M. Kimura is with Department of Information Science and Engineering,
School of Engineering, Shibaura Institute of Technology, 3-7-5, Toyosu,
Koto City, Tokyo 135-8548, Japan (e-mail: masaomi@shibaura-it.ac.jp).
Cite: Imam Mukhlash, Anshar Zamrudillah Arham, Fakhrur Rozi, Masaomi Kimura, and Dieky Adzkiya, "Opinion Mining on Book Review Using Convolutional Neural Network Algorithm — Long Short Term Memory," International Journal of Machine Learning and Computing vol. 8, no. 5, pp. 437-441, 2018.