Learning Adversarial Networks for Semi-Supervised Text Classification via Policy Gradient

Li, Yan; Ye, Jieping

doi:10.1145/3219819.3219956

Cited by 43 publications

(17 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Semi-supervised Methods. Many semi-supervised methods have been explored on sentence-level sentiment classification, such as pretraining with Restricted Boltzmann Machine or autoencoder [23,26], auxiliary task learning [24] and adversarial training [25,27]. However, there are only few studies [16,19] on semi-supervised target-level sentiment classification.…”

Section: Deep Methodsmentioning

confidence: 99%

Learning Discriminative Neural Sentiment Units for Semi-supervised Target-Level Sentiment Classification

Jin

Yang

Pang

et al. 2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Target-level sentiment classification aims at assigning sentiment polarities to opinion targets in a sentence, for which it is significantly more challenging to obtain large-scale labeled data than sentence/document-level sentiment classification due to the intricate contexts and relations of the target words. To address this challenge, we propose a novel semi-supervised approach to learn sentiment-aware representations from easily accessible unlabeled data specifically for the finegrained sentiment learning. This is very different from current popular semi-supervised solutions that use the unlabeled data via pretraining to generate generic representations for various types of downstream tasks. Particularly, we show for the first time that we can learn and detect some highly sentiment-discriminative neural units from the unsupervised pretrained model, termed neural sentiment units. Due to the discriminability, these sentiment units can be leveraged by downstream LSTM-based classifiers to generate sentiment-aware and context-dependent word representations to substantially improve their sentiment classification performance. Extensive empirical results on two benchmark datasets show that our approach (i) substantially outperforms state-of-the-art sentiment classifiers and (ii) achieves significantly better data efficiency.

show abstract

Section: Deep Methodsmentioning

confidence: 99%

Learning Discriminative Neural Sentiment Units for Semi-supervised Target-Level Sentiment Classification

Jin

Yang

Pang

et al. 2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…Recently, generative adversarial networks (GANs) [10] have demonstrated its superior performance in many tasks [21,37]. A few methods leverage the adversarial training to learn more robust network representation [8,12,36,48].…”

Section: Anomaly Detection and Robust Representation Learning On Graphsmentioning

confidence: 99%

REA: Robust Cross-lingual Entity Alignment Between Knowledge Graphs

Pei

et al. 2020

Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

Cross-lingual entity alignment aims at associating semantically similar entities in knowledge graphs with different languages. It has been an essential research problem for knowledge integration and knowledge graph connection, and been studied with supervised or semi-supervised machine learning methods with the assumption of clean labeled data. However, labels from human annotations often include errors, which can largely affect the alignment results. We thus aim to formulate and explore the robust entity alignment problem, which is non-trivial, due to the deficiency of noisy labels. Our proposed method named REA (Robust Entity Alignment) consists of two components: noise detection and noise-aware entity alignment. The noise detection is designed by following the adversarial training principle. The noise-aware entity alignment is devised by leveraging graph neural network based knowledge graph encoder as the core. In order to mutually boost the performance of the two components, we propose a unified reinforced training strategy to combine them. To evaluate our REA method, we conduct extensive experiments on several real-world datasets. The experimental results demonstrate the effectiveness of our proposed method and also show that our model consistently outperforms the state-of-theart methods with significant improvement on alignment accuracy in the noise-involved scenario.

show abstract

“…Hence, VAE initially achieves great success in generating the latent representation from high dimensional data. Recently, semi-supervised methods with generative models have attracted more attention, which can achieve a better result compared with simple supervised methods [20], [29]- [31]. Xu et al propose a semi-supervised sequential variational autoencoder framework by using conditional RNN for text classification.…”

Section: B Generative Modelmentioning

confidence: 99%

Adversarial Mobility Learning for Human Trajectory Classification

et al. 2020

View full text Add to dashboard Cite

Understanding human mobility is one of the important but challenging tasks in Location-based Social Networks (LBSN). Recently, a user mobility mining task called Trajectory User Linking (TUL) has become an essential and popular topic, aiming at identifying user identities through exploiting their mobility patterns. Existing methods mainly focus on learning sequential mobility patterns by capturing long-short term dependencies among historical check-ins. However, users have personalized moving preferences, which have not been considered in previous work. Besides, how to leverage the prior knowledge behind human mobility needs to be further investigated. In this work, we present a novel semi-supervised method, called AdattTUL, to make adversarial mobility learning for human trajectory classification, which is an end-to-end framework modeling human moving patterns. AdattTUL integrates multiple human preferences of check-in behaviors and involves an attention mechanism to dynamically capture the complex relationships of user check-ins from trajectory data. In addition, AdattTUL leverages an adversarial network to help in regularizing the data distribution of human trajectories. Extensive experiments conducted on real-world LBSN datasets show that AdattTUL significantly improves the TUL performance.INDEX TERMS Trajectory user linking, adversarial model, autoencoder, attention mechanism, human mobility.

show abstract

Learning Adversarial Networks for Semi-Supervised Text Classification via Policy Gradient

Cited by 43 publications

References 14 publications

Learning Discriminative Neural Sentiment Units for Semi-supervised Target-Level Sentiment Classification

Learning Discriminative Neural Sentiment Units for Semi-supervised Target-Level Sentiment Classification

REA: Robust Cross-lingual Entity Alignment Between Knowledge Graphs

Adversarial Mobility Learning for Human Trajectory Classification

Contact Info

Product

Resources

About