Effective LSTMs for Target-Dependent Sentiment Classification

By Duyu Tang, Bing Qin, Xiaocheng Feng, Ting-Wei Liu. International Conference on Computational Linguistics, 2015

1,000 citations recorded by the source corpus.

Target-dependent sentiment classification remains a challenge: modeling the semantic relatedness of a target with its context words in a sentence. Different context words have different influences on determining the sentiment polarity of a sentence towards the target. Therefore, it is desirable to integrate the connections between target word and context words when building a learning system. In this paper, we develop two target dependent long short-term memory (LSTM) models, where target information is automatically taken into account. We evaluate our methods on a benchmark dataset from Twitter. Empirical results show that modeling sentence representation with standard LSTM does not perform well. Incorporating target information into LSTM can significantly boost the classification accuracy. The target-dependent LSTM models achieve state-of-the-art performances without using syntactic parser or external sentiment lexicons.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/0b0dc14b8a8dcccbfc62a38355fff2f6a361e9d2

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

1. Erik Cambria
2. Bing Liu
3. Lidong Bing
4. Ting-Wei Liu
5. Bing Qin
6. Duyu Tang
7. Wai Lam
8. Soujanya Poria

Citations in the corpus

• Deep learning for sentiment analysis: A survey (2018, background): “Since LSTM can capture semantic relations between the target and its context words in a more flexible way, Tang et al. 70 proposed Target-dependent LSTM (TD-LSTM) and Target-connection LSTM (TC-LSTM) to extend LSTM by taking the target into consideration.”
• Text Classification Improved by Integrating Bidirectional LSTM with Two-dimensional Max Pooling (2016, background): “Tang et al. (2015) developed target dependent Long ShortTerm Memory Networks (LSTM (Hochreiter and Schmidhuber, 1997)), where target information is automatically taken into account.”
• INSIGHT-1 at SemEval-2016 Task 5: Deep Learning for Multilingual Aspect-based Sentiment Analysis (2016, background): “However, without relying on dependency and constituency trees (Nguyen and Shirai, 2015) or positional information (Tang et al., 2015), the model falls short of being able to reliably triangulate aspects, particularly in sentences with 2https://www.yelp.com/dataset_challenge Lg. Dom.”
• Description Based Text Classification with Reinforcement Learning (2020, methodology): “It has broad applications such as sentiment analysis (Pang et al., 2002; Maas et al., 2011; Socher et al., 2013; Tang et al., 2014, 2015b), aspect sentiment clas-sification ( Jo and Oh, 2011; Tang et al., 2015a; Wang et al., 2015; Nguyen and Shirai, 2015; Tang et al., 2016b; Pontiki et al., 2016;…”
• A Challenge Dataset and Effective Models for Aspect-Based Sentiment Analysis (2019, methodology): “Third, attention based models without effectively modeling word order (e.g., MemNet and AEN) perform worst on MAMS since they lose word order information and cannot identify which part of context describes the given aspect.”
• A survey on empathetic dialogue systems (2020, background): “A similar problem has been raised for emotion or sentiment classification [140, 141, 142], but has been missing in the context of dialogue systems.”
• Transformation Networks for Target-Oriented Sentiment Classification (2018, methodology): “To overcome this limitation, researchers bring in the attention mechanism to model target-context association (Tang et al., 2016a,b; Wang et al., 2016; Yang et al., 2017; Liu and Zhang, 2017; Ma et al., 2017; Chen et al., 2017; Zhang et al., 2017; Tay et al., 2017).”
• Aspect Level Sentiment Classification with Deep Memory Network (2016, methodology): “(3) We compare with three LSTM models (Tang et al., 2015a)).”
• Syntax-Aware Aspect Level Sentiment Classification with Graph Attention Networks (2019, methodology): “The last hidden states of these two LSTM networks are concate-2 nated for predicting the sentiment polarity (Tang et al., 2016a).”
• Aspect Level Sentiment Classification with Attention-over-Attention Neural Networks (2018, methodology): “Restaurant Laptop Majority 0.535 0.650 LSTM 0.743 0.665 TD-LSTM [23] 0.756 0.681 AT-LSTM [28] 0.762 0.689 ATAE-LSTM [28] 0.772 0.687 IAN [11] 0.786 0.721 AOA-LSTM 0.812 (0.797 ± 0.008) 0.745 (0.726 ± 0.008) Table 2.”
• Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence (2019, background): “The earliest work on (T)ABSA relied heavily on feature engineering (Wagner et al., 2014; Kiritchenko et al., 2014), and subsequent neural network-based methods (Nguyen and Shirai, 2015; Wang et al., 2016; Tang et al., 2015, 2016; Wang et al., 2017) achieved higher accuracy.”
• Aspect-based sentiment analysis via affective knowledge enhanced graph convolutional networks (2021, neutral): “In ABSA, existing neural network-based approaches generally attempt to learn the key sentiment information from the context and determine the sentiment polarity according to the specific aspect or target [4,8,9,13–19].”
• Multi-grained Attention Network for Aspect-Level Sentiment Classification (2018, methodology): “Previous works (Tang et al., 2016b; Chen et al., 2017) mainly employed the simple averaged aspect vector to learn the attention weights on the context words.”