Conghui Zhu


pdf bib
Demographics Should Not Be the Reason of Toxicity: Mitigating Discrimination in Text Classifications with Instance Weighting
Guanhua Zhang | Bing Bai | Junqi Zhang | Kun Bai | Conghui Zhu | Tiejun Zhao
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

With the recent proliferation of the use of text classifications, researchers have found that there are certain unintended biases in text classification datasets. For example, texts containing some demographic identity-terms (e.g., “gay”, “black”) are more likely to be abusive in existing abusive language detection datasets. As a result, models trained with these datasets may consider sentences like “She makes me happy to be gay” as abusive simply because of the word “gay.” In this paper, we formalize the unintended biases in text classification datasets as a kind of selection bias from the non-discrimination distribution to the discrimination distribution. Based on this formalization, we further propose a model-agnostic debiasing training framework by recovering the non-discrimination distribution using instance weighting, which does not require any extra resources or annotations apart from a pre-defined set of demographic identity-terms. Experiments demonstrate that our method can effectively alleviate the impacts of the unintended biases without significantly hurting models’ generalization ability.

pdf bib
Cross Copy Network for Dialogue Generation
Changzhen Ji | Xin Zhou | Yating Zhang | Xiaozhong Liu | Changlong Sun | Conghui Zhu | Tiejun Zhao
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

In the past few years, audiences from different fields witness the achievements of sequence-to-sequence models (e.g., LSTM+attention, Pointer Generator Networks and Transformer) to enhance dialogue content generation. While content fluency and accuracy often serve as the major indicators for model training, dialogue logics, carrying critical information for some particular domains, are often ignored. Take customer service and court debate dialogue as examples, compatible logics can be observed across different dialogue instances, and this information can provide vital evidence for utterance generation. In this paper, we propose a novel network architecture - Cross Copy Networks (CCN) to explore the current dialog context and similar dialogue instances’ logical structure simultaneously. Experiments with two tasks, court debate and customer service content generation, proved that the proposed algorithm is superior to existing state-of-art content generation models.


pdf bib
Understanding Data Augmentation in Neural Machine Translation: Two Perspectives towards Generalization
Guanlin Li | Lemao Liu | Guoping Huang | Conghui Zhu | Tiejun Zhao
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

Many Data Augmentation (DA) methods have been proposed for neural machine translation. Existing works measure the superiority of DA methods in terms of their performance on a specific test set, but we find that some DA methods do not exhibit consistent improvements across translation tasks. Based on the observation, this paper makes an initial attempt to answer a fundamental question: what benefits, which are consistent across different methods and tasks, does DA in general obtain? Inspired by recent theoretic advances in deep learning, the paper understands DA from two perspectives towards the generalization ability of a model: input sensitivity and prediction margin, which are defined independent of specific test set thereby may lead to findings with relatively low variance. Extensive experiments show that relatively consistent benefits across five DA methods and four translation tasks are achieved regarding both perspectives.

pdf bib
Selection Bias Explorations and Debias Methods for Natural Language Sentence Matching Datasets
Guanhua Zhang | Bing Bai | Jian Liang | Kun Bai | Shiyu Chang | Mo Yu | Conghui Zhu | Tiejun Zhao
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics

Natural Language Sentence Matching (NLSM) has gained substantial attention from both academics and the industry, and rich public datasets contribute a lot to this process. However, biased datasets can also hurt the generalization performance of trained models and give untrustworthy evaluation results. For many NLSM datasets, the providers select some pairs of sentences into the datasets, and this sampling procedure can easily bring unintended pattern, i.e., selection bias. One example is the QuoraQP dataset, where some content-independent naive features are unreasonably predictive. Such features are the reflection of the selection bias and termed as the “leakage features.” In this paper, we investigate the problem of selection bias on six NLSM datasets and find that four out of them are significantly biased. We further propose a training and evaluation framework to alleviate the bias. Experimental results on QuoraQP suggest that the proposed framework can improve the generalization ability of trained models, and give more trustworthy evaluation results for real-world adoptions.

pdf bib
Understanding and Improving Hidden Representations for Neural Machine Translation
Guanlin Li | Lemao Liu | Xintong Li | Conghui Zhu | Tiejun Zhao | Shuming Shi
Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)

Multilayer architectures are currently the gold standard for large-scale neural machine translation. Existing works have explored some methods for understanding the hidden representations, however, they have not sought to improve the translation quality rationally according to their understanding. Towards understanding for performance improvement, we first artificially construct a sequence of nested relative tasks and measure the feature generalization ability of the learned hidden representation over these tasks. Based on our understanding, we then propose to regularize the layer-wise representations with all tree-induced tasks. To overcome the computational bottleneck resulting from the large number of regularization terms, we design efficient approximation methods by selecting a few coarse-to-fine tasks for regularization. Extensive experiments on two widely-used datasets demonstrate the proposed methods only lead to small extra overheads in training but no additional overheads in testing, and achieve consistent improvements (up to +1.3 BLEU) compared to the state-of-the-art translation model.


pdf bib
Improving Pivot-Based Statistical Machine Translation by Pivoting the Co-occurrence Count of Phrase Pairs
Xiaoning Zhu | Zhongjun He | Hua Wu | Conghui Zhu | Haifeng Wang | Tiejun Zhao
Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP)

pdf bib
Detection on Inconsistency of Verb Phrase in TreeBank
Chaoqun Duan | Dequan Zheng | Conghui Zhu | Sheng Li | Hongye Tan
Proceedings of The Third CIPS-SIGHAN Joint Conference on Chinese Language Processing


pdf bib
Improving Pivot-Based Statistical Machine Translation Using Random Walk
Xiaoning Zhu | Zhongjun He | Hua Wu | Haifeng Wang | Conghui Zhu | Tiejun Zhao
Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing

pdf bib
Hierarchical Phrase Table Combination for Machine Translation
Conghui Zhu | Taro Watanabe | Eiichiro Sumita | Tiejun Zhao
Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)


pdf bib
The HIT-LTRC machine translation system for IWSLT 2012
Xiaoning Zhu | Yiming Cui | Conghui Zhu | Tiejun Zhao | Hailong Cao
Proceedings of the 9th International Workshop on Spoken Language Translation: Evaluation Campaign

In this paper, we describe HIT-LTRC's participation in the IWSLT 2012 evaluation campaign. In this year, we took part in the Olympics Task which required the participants to translate Chinese to English with limited data. Our system is based on Moses[1], which is an open source machine translation system. We mainly used the phrase-based models to carry out our experiments, and factored-based models were also performed in comparison. All the involved tools are freely available. In the evaluation campaign, we focus on data selection, phrase extraction method comparison and phrase table combination.

pdf bib
Locally Training the Log-Linear Model for SMT
Lemao Liu | Hailong Cao | Taro Watanabe | Tiejun Zhao | Mo Yu | Conghui Zhu
Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning

pdf bib
Expected Error Minimization with Ultraconservative Update for SMT
Lemao Liu | Tiejun Zhao | Taro Watanabe | Hailong Cao | Conghui Zhu
Proceedings of COLING 2012: Posters


pdf bib
A Unified Tagging Approach to Text Normalization
Conghui Zhu | Jie Tang | Hang Li | Hwee Tou Ng | Tiejun Zhao
Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics