Learning Sentence Embeddings with Auxiliary Tasks for Cross-Domain Sentiment Classification

By Jianfei Yu, Jing Jiang. 2016

195 citations recorded by the source corpus.

In this paper, we study cross-domain sentiment classification with neural network archi-tectures. We borrow the idea from Structural Correspondence Learning and use two auxiliary tasks to help induce a sentence embedding that supposedly works well across domains for sentiment classification. We also propose to jointly learn this sentence embedding together with the sentiment classifier itself. Experiment results demonstrate that our proposed joint model outperforms several state-of-the-art methods on five benchmark datasets.

Subfields: Artificial Intelligence, Information Systems, Natural Language Processing

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. Bing Liu
2. B. Liu
3. Minlie Huang
4. Shuai Wang
5. Lei Zhang
6. Xiaoyan Zhu
7. Hu Xu
8. Lei Shu

In Information Systems

1. B. Liu
2. Bing Liu
3. Lei Zhang
4. Shuai Wang
5. Xiaoyan Zhu
6. Guangyi Lv
7. Sebastian Ruder
8. Minlie Huang

In Natural Language Processing

1. Bing Liu
2. B. Liu
3. Shuai Wang
4. Roi Reichart
5. Minlie Huang
6. Xiaoyan Zhu
7. Sebastian Ruder
8. Lei Zhang

Citations in the corpus

• An Overview of Multi-Task Learning in Deep Neural Networks (2017, background): “Recent examples of this strategy in the context of natural language processing are [Yu and Jiang, 2016] who predict whether an input sentence contains a positive or negative sentiment word as auxiliary tasks for sentiment analysis and [Cheng et al., 2015] who predict whether a name is present in a…”
• Deep learning for sentiment analysis: A survey (2018, methodology): “Yu and Jiang (2016) studied the problem of learning generalized sentence embeddings for cross-domain sentence sentiment classification and designed a neural network model containing two separated CNNs that jointly learn two hidden feature representations from both the labeled and unlabeled data. Zha”
• Sentiment Classification by Leveraging the Shared Knowledge from a Sequence of Domains (2019, methodology): “Thus, when we train matrix multiplication based networks like RNNs, we can keep some columns in the weight matrix unchanged, and only update those less important ones.”
• A Knowledge-Enhanced Pretraining Model for Commonsense Story Generation (2020, background): “Incorporating other auxiliary task objectives to complement the primary goal has been shown to improve the performance in many NLP tasks such as sentiment classification (Yu and Jiang, 2016) and conversation generation (Zhao et al.”
• Neural Structural Correspondence Learning for Domain Adaptation (2016, background): “Yu and Jiang (2016) presented a convolutional NN that learns sentence embeddings using two auxiliary tasks (whether the sentence contains a positive or a negative domain independent sentiment word), purposely avoiding prediction with respect to a large set of pivot features.”
• Projecting Embeddings for Domain Adaption: Joint Modeling of Sentiment Analysis in Diverse Domains (2018, methodology): “Yu and Jiang (2016) learn sentence embeddings that are useful across domains through multi-task learning.”
• Pivot Based Language Modeling for Improved Neural Domain Adaptation (2018, unknown): “(Pan et al., 2010; Gouws et al., 2012; Bollegala et al., 2015; Yu and Jiang, 2016; Ziser and Reichart, 2017)).”