Opinion Mining with Deep Recurrent Neural Networks

By Ozan Irsoy, Claire Cardie. 2014

392 citations recorded by the source corpus.

Recurrent neural networks (RNNs) are connectionist models of sequential data that are naturally applicable to the analysis of natural language. Recently, “depth in space” — as an orthogonal notion to “depth in time” — in RNNs has been investigated by stacking multiple layers of RNNs and shown empirically to bring a temporal hierarchy to the architecture. In this work we apply these deep RNNs to the task of opinion expression extraction formulated as a token-level sequence-labeling task. Experimental results show that deep, narrow RNNs outperform traditional shallow, wide RNNs with the same number of parameters. Furthermore, our approach outperforms previous CRF-based baselines, including the state-of-the-art semi-Markov CRF model, and does so without access to the powerful opinion lexicons and syntactic features relied upon by the semi-CRF, as well as without the standard layer-by-layer pre-training typically required of RNN architectures.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Graph Convolutional Encoders for Syntax-aware Neural Machine Translation (2017, methodology): “To not only summarize the past of a word, but also its future, a bidirectional RNN (Schuster and Paliwal, 1997; Irsoy and Cardie, 2014) is often used.”
Deep multi-task learning with low level tasks supervised at lower layers (2016, methodology): “We experiment with a multi-task learning (MTL) architecture based on deep bi-directional recurrent neural networks (bi-RNNs) (Schuster and Paliwal, 1997; Irsoy and Cardie, 2014).”
Harvesting Paragraph-level Question-Answer Pairs from Wikipedia (2018, background): “Binary Overlap counts every predicted answer that overlaps with a gold answer span as correct, and Proportional Overlap give partial credit proportional to the amount of overlap (Johansson and Moschitti, 2010; Irsoy and Cardie, 2014).”
Simple and Accurate Dependency Parsing Using Bidirectional LSTM Feature Representations (2016, methodology): “Our proposal (Section 3) is centered around BiRNNs (Irsoy and Cardie, 2014; Schuster and Paliwal, 1997), and more specifically BiLSTMs (Graves, 2008), which are strong and trainable sequence models (see Section 2.3).”
A Primer on Neural Network Models for Natural Language Processing (2015, methodology): “The use of BI-RNNs for sequence tagging was introduced to the NLP community by Irsoy and Cardie (2014).”
Semi-supervised Multitask Learning for Sequence Labeling (2017, background): “However, recent work has shown that neural network architectures are able to achieve comparable or improved performance, while automatically discovering useful features for a specific task and only requiring a sequence of tokens as input (Collobert et al., 2011; Irsoy and Cardie, 2014; Lample et al.”
Easy-First Dependency Parsing with Hierarchical Tree LSTMs (2016, background): “ive regarding its meaning. One way of incorporating context is the Bidirectional RNN (Schuster and Paliwal, 1997). Bidirectional RNNs are shown to be an effective representation for sequence tagging (Irsoy and Cardie, 2014). Bidirectional RNNs the black fox who really likes apples did not jump over ”
How Polarized Have We Become? A Multimodal Classification of Trump Followers and Clinton Followers (2017, background): “Recursive neural networks (RNN), in particular its variant Long Short Memory Networks (LSTM) [12], have been very successful in modeling sequential data [43, 23], with wide applications in text classiﬁcation [16], especially in sentiment analysis [14, 19].”
Emotional States vs. Emotional Words in Social Media (2015, methodology): “Other promising sentiment analysis tools use machine learning techniques like neural networks [15], conditional random fields [37], or support vector machines (e.”
DAG-Recurrent Neural Networks for Scene Labeling (2015, methodology): “Our proposed DAG-RNN is a generalization of chain-RNNs [1][11], tree-RNNs [27][37] and 2D-RNNs [9][23], and it enables the network to model long-range semantic dependencies for graphical structured images.”
A Neural Approach to Automated Essay Scoring (2016, positive): “Apart from these examples, recurrent neural networks have also been used for opinion mining (Irsoy and Cardie, 2014), sequence labeling (Ma and Hovy, 2016), language modeling (Kim et al.”
Dimensional Sentiment Analysis Using a Regional CNN-LSTM Model (2016, background): “…networks (NN) such as convolutional neural networks (CNN) (Kim, 2014; Kalchbrenner et al., 2014), recurrent neural networks (RNN) (Graves, 2012; Irsoy and Cardie, 2014) and long short-term memory (LSTM) (Wang et al., 2015; Liu et al., 2015) have been successfully employed for categorical…”
Sentiment Analysis in Tourism: Capitalizing on Big Data (2019, methodology): “Neural network and deep learning models (Irsoy & Cardie, 2014; Socher et al., 2013) and k-nearest neighbour method (Schmunk et al., 2014) were also employed for semantic analysis in the literature.”
Sentiment Analysis of Comment Texts Based on BiLSTM (2019, methodology): “Irsoy and Cardie [20] used the Recurrent Neural Network based on time series information to obtain sentence representation, which further improved the accuracy of sentiment classification.”
Simple Recurrent Units for Highly Parallelizable Recurrence (2017, background): “, 2014; Gal and Ghahramani, 2016; Zoph and Le, 2016), opinion mining (Irsoy and Cardie, 2014), and situated language understanding (Mei et al.”