Attentive convolutional gated recurrent network: a contextual model to sentiment analysis
- PDF / 1,538,921 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 38 Downloads / 184 Views
ORIGINAL ARTICLE
Attentive convolutional gated recurrent network: a contextual model to sentiment analysis Olivier Habimana1 · Yuhua Li1 · Ruixuan Li1 · Xiwu Gu1 · Wenjin Yan1 Received: 20 April 2019 / Accepted: 2 May 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Considering contextual features is a key issue in sentiment analysis. Existing approaches including convolutional neural networks (CNNs) and recurrent neural networks (RNNs) lack the ability to account and prioritize informative contextual features that are necessary for better sentiment interpretation. CNNs present limited capability since they are required to be very deep, which can lead to the gradient vanishing whereas, RNNs fail because they sequentially process input sequences. Furthermore, the two approaches treat all words equally. In this paper, we suggest a novel approach named attentive convolutional gated recurrent network (ACGRN) that alleviates the above issues for sentiment analysis. The motivation behind ACGRN is to avoid the vanishing gradient caused by deep CNN via applying a shallow-and-wide CNN that learns local contextual features. Afterwards, to solve the problem caused by the sequential structure of RNN and prioritizing informative contextual information, we use a novel prior knowledge attention based bidirectional gated recurrent unit (ATBiGRU). Prior knowledge ATBiGRU captures global contextual features with a strong focus on the previous hidden states that carry more valuable information to the current time step. The experimental results show that ACGRN significantly outperforms the baseline models over six small and large real-world datasets for the sentiment classification task. Keywords Sentiment analysis · Convolutional neural network · Recurrent neural network · Attention mechanism · Contextual features
1 Introduction Nowadays, with the notable increase of Web 2.0 tools like online social media and e-commerce platforms, users freely express their ideas and thoughts in the form of text [1, 5, 6, 39, 58]. Consequently, many organizations became increasingly interested in getting the hidden insights from these user-generated content (UGC) [13, 24, 36] to assist in decision making and monitoring public opinion. Therefore, sentiment analysis has received a substantial amount of attention from many researchers as one of the natural language processing tasks that focuses on finding the opinions articulated in the UGC. * Yuhua Li [email protected] * Ruixuan Li [email protected] 1
School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, China
To get good results in sentiment analysis requires modeling and prioritizing informative contextual features. Considering the following review text extracted from the Amazon dataset, which talks about the sandals: “I received this day and I’m not a fan of it but I thought it would be puffier as it looks in the pic but it is not what I wanted to do with the sandals she was gonna wear it now I’m going to fi
Data Loading...
