Shining Liang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Shining Liang has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Wanli Zuo
College of Computer Science and Technology, Jilin University, Changchun, China
3 shared publications - Daxin Jiang
StepFun, Shanghai, China
2 shared publications - Jian Pei
Duke University
2 shared publications - Linjun Shou
Microsoft Edge Consumer
2 shared publications - Xianglin Zuo
Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education, Changchun 130012, China and College of Computer Science and Technology, Jilin University, Changchun 130012, China
2 shared publications - Lin Yue
Northeast Normal University, Changchun 130024, China
1 shared publication
Latest Publications
- Open Access
Label-aware Multi-level Contrastive Learning for Cross-lingual Spoken Language Understanding
Despite the great success of spoken language understanding (SLU) in high-resource languages, it remains challenging in low-resource languages mainly due to the lack of labeled training data. The recent multilingual code-switching approach achieves better alignments of model representations across languages by constructing a mixed-language context in zero-shot cross-lingual SLU. However, current code-switching methods are limited to implicit alignment and disregard the inherent semantic structure in SLU, i.e., the hierarchical inclusion of utterances, slots, and words. In this paper, we propose to model the utterance-slot-word structure by a multi-level contrastive learning framework at the utterance, slot, and word levels to facilitate explicit alignment. Novel code-switching schemes are introduced to generate hard negative examples for our contrastive learning framework. Furthermore, we develop a label-aware joint model leveraging label semantics to enhance the implicit alignment and feed to contrastive learning. Our experimental results show that our proposed methods significantly improve the performance compared with the strong baselines on two zero-shot cross-lingual SLU benchmark datasets.
Open AccessLabel-aware Multi-level Contrastive Learning for Cross-lingual Spoken Language Understanding
Despite the great success of spoken language understanding (SLU) in high-resource languages, it remains challenging in low-resource languages mainly due to the lack of labeled training data. The recent multilingual code-switching approach achieves better alignments of model representations across languages by constructing a mixed-language context in zero-shot cross-lingual SLU. However, current code-switching methods are limited to implicit alignment and disregard the inherent semantic structure in SLU, i.e., the hierarchical inclusion of utterances, slots, and words. In this paper, we propose to model the utterance-slot-word structure by a multi-level contrastive learning framework at the utterance, slot, and word levels to facilitate explicit alignment. Novel code-switching schemes are introduced to generate hard negative examples for our contrastive learning framework. Furthermore, we develop a label-aware joint model leveraging label semantics to enhance the implicit alignment and feed to contrastive learning. Our experimental results show that our proposed methods significantly improve the performance compared with the strong baselines on two zero-shot cross-lingual SLU benchmark datasets.
- Open Access
CalibreNet: Calibration Networks for Multilingual Sequence Labeling
Lack of training data in low-resource languages presents huge challenges to sequence labeling tasks such as named entity recognition (NER) and machine reading comprehension (MRC). One major obstacle is the errors on the boundary of predicted answers. To tackle this problem, we propose CalibreNet, which predicts answers in two steps. In the first step, any existing sequence labeling method can be adopted as a base model to generate an initial answer. In the second step, CalibreNet refines the boundary of the initial answer. To tackle the challenge of lack of training data in low-resource languages, we dedicatedly develop a novel unsupervised phrase boundary recovery pre-training task to enhance the multilingual boundary detection capability of CalibreNet. Experiments on two cross-lingual benchmark datasets show that the proposed approach achieves SOTA results on zero-shot cross-lingual NER and MRC tasks.
Open AccessCalibreNet: Calibration Networks for Multilingual Sequence Labeling
Lack of training data in low-resource languages presents huge challenges to sequence labeling tasks such as named entity recognition (NER) and machine reading comprehension (MRC). One major obstacle is the errors on the boundary of predicted answers. To tackle this problem, we propose CalibreNet, which predicts answers in two steps. In the first step, any existing sequence labeling method can be adopted as a base model to generate an initial answer. In the second step, CalibreNet refines the boundary of the initial answer. To tackle the challenge of lack of training data in low-resource languages, we dedicatedly develop a novel unsupervised phrase boundary recovery pre-training task to enhance the multilingual boundary detection capability of CalibreNet. Experiments on two cross-lingual benchmark datasets show that the proposed approach achieves SOTA results on zero-shot cross-lingual NER and MRC tasks.
- Open Access
- 17 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 15423-15435
IDDSAM: An Integrated Disease Diagnosis and Severity Assessment Model for Intensive Care Units
People are admitted to intensive care units (ICUs) because they need complete support for failing organ systems, constant monitoring, routine nursing care, and treatment. A critical or intensive illness is different from conventional or chronic diseases that most people are likely to have previously encountered. Such an illness is often unexpected and without warnings and can suddenly strike the previously fit. High levels of treatment and support are generally required to prevent life-threatening complications for the patents. Two of the most noticeable actions during an ICU stay are disease diagnosis and severity assessment of the patients. Unlike the majority of previous approaches where diagnosis and severity assessment are studied separately, we treat these actions as two tasks in an integrated procedure that clinicians must be able to quickly and accurately conduct such that patients are given the best possible chance for therapeutic success. In this paper, we propose an integrated disease diagnosis and severity assessment model (IDDSAM) to diagnose and assess diseases. Moreover, accompanying the prediction, we also provide an evidence-based explanation. IDDSAM is a multisource multitask model that is based on an attention mechanism and utilizes shareable information from laboratory tests, bedside monitoring, and complications to support patients’ severity assessment and in-hospital disease diagnoses. We use 50,430 ICU cases consisting of 46,520 patients from 50 kinds of diseases over nine classifications to evaluate our proposed model. The experimental results demonstrated that our model outperforms the existing state-of-the-art mortality and diagnosis prediction framework by 3.79% on average in terms of accuracy for the mortality prediction tasks and by 14.51% on average for the diagnosis tasks.
- 17 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 15423-15435
Open AccessIDDSAM: An Integrated Disease Diagnosis and Severity Assessment Model for Intensive Care Units
People are admitted to intensive care units (ICUs) because they need complete support for failing organ systems, constant monitoring, routine nursing care, and treatment. A critical or intensive illness is different from conventional or chronic diseases that most people are likely to have previously encountered. Such an illness is often unexpected and without warnings and can suddenly strike the previously fit. High levels of treatment and support are generally required to prevent life-threatening complications for the patents. Two of the most noticeable actions during an ICU stay are disease diagnosis and severity assessment of the patients. Unlike the majority of previous approaches where diagnosis and severity assessment are studied separately, we treat these actions as two tasks in an integrated procedure that clinicians must be able to quickly and accurately conduct such that patients are given the best possible chance for therapeutic success. In this paper, we propose an integrated disease diagnosis and severity assessment model (IDDSAM) to diagnose and assess diseases. Moreover, accompanying the prediction, we also provide an evidence-based explanation. IDDSAM is a multisource multitask model that is based on an attention mechanism and utilizes shareable information from laboratory tests, bedside monitoring, and complications to support patients’ severity assessment and in-hospital disease diagnoses. We use 50,430 ICU cases consisting of 46,520 patients from 50 kinds of diseases over nine classifications to evaluate our proposed model. The experimental results demonstrated that our model outperforms the existing state-of-the-art mortality and diagnosis prediction framework by 3.79% on average in terms of accuracy for the mortality prediction tasks and by 14.51% on average for the diagnosis tasks.