<abstract xmlns="http://www.w3.org/1999/xhtml"><p>In recent years, topic modeling and deep neural network-based methods have attracted much attention in sentiment analysis of online reviews. This paper presents a hybrid topic model-based approach for aspect extraction and sentiment classification of textual reviews. Latent Dirichlet allocation applied for aspect extraction and two-layer bi-directional long short-term memory (LSTM) for sentiment classification. This work also proposes a hill climbing-based approach for tunning model hyperparameters. The proposed model evaluated on three different datasets. Compared to the single-layer Bi-LSTM model, the proposed model gives 95, 95, and 86% accuracy for the movie, mobile, and hotel domain, respectively.</p></abstract>

In recent years, topic modeling and deep neural network-based methods have attracted much attention in sentiment analysis of online reviews. This paper presents a hybrid topic model-based approach for aspect extraction and sentiment classification of textual reviews. Latent Dirichlet allocation applied for aspect extraction and two-layer bi-directional long short-term memory (LSTM) for sentiment classification. This work also proposes a hill climbing-based approach for tunning model hyperparameters. The proposed model evaluated on three different datasets. Compared to the single-layer Bi-LSTM model, the proposed model gives 95, 95, and 86% accuracy for the movie, mobile, and hotel domain, respectively.

An efficient sentiment analysis using topic model based optimized recurrent neural network

International Journal on Smart Sensing and Intelligent Systems

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

running-head

An efficient sentiment analysis using topic model based optimized recurrent neural network: Pathik and Shukla

In recent years, topic modeling and deep neural network-based methods have attracted much attention in sentiment analysis of online reviews. This paper...

Dataset domain	Total	+ve	‒ve
Restaurant from Yelp	1,000	500	500
Mobile from Amazon	1,000	500	500
Movies from IMDB	1,000	500	500

Model	Single-layer Bi-LSTM (Hameed and Garcia-Zapirain, 2020)		Two-layer Bi-LSTM		Two-layer HCT Bi-LSTM
Dataset	T	V	T	V	T	V
Amazon	0.83	0.51	0.91	0.70	0.95	0.76
Yelp	0.84	0.70	0.85	0.72	0.86	0.75
IMDB	0.71	0.81	0.90	0.81	0.95	0.82

Parameter	Value
Vocabulary size	10,000
Bi-LSTM	2 layer
Dense	1
Activation	Sigmoid
Optimizer	Adam function
Loss Function	Binary cross-entropy
Input Length	100
Learning rate	0.002
Epoch	10

An efficient sentiment analysis using topic model based optimized recurrent neural network

Pubblicato online: 22 giu 2021

Pagine: 1 - 12

Ricevuto: 21 feb 2021

DOI: https://doi.org/10.21307/ijssis-2021-011

Parole chiave
Bi-LSTM, LDA, Hill-climbing, Classification, Hyperparameters

© 2021 Nikhlesh Pathik et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6:

Figure 7:

Figure 8:

Figure 9:

Figure 10:

Figure 11:

Figure 12:

Dataset statistics.

Comparison of proposed HCL-Bi-LSTM model.

Various model parameters.

An efficient sentiment analysis using topic model based optimized recurrent neural network

Pubblicato online: 22 giu 2021

Pagine: 1 - 12

Ricevuto: 21 feb 2021

DOI: https://doi.org/10.21307/ijssis-2021-011

Parole chiaveBi-LSTM, LDA, Hill-climbing, Classification, Hyperparameters

© 2021 Nikhlesh Pathik et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 6:

Figure 7:

Figure 8:

Figure 9:

Figure 10:

Figure 11:

Figure 12:

Dataset statistics.

Comparison of proposed HCL-Bi-LSTM model.

Various model parameters.

Parole chiave
Bi-LSTM, LDA, Hill-climbing, Classification, Hyperparameters