A5101967248- Topic Modeling
- Natural Language Processing Techniques
- Text and Document Classification Technologies
- Sentiment Analysis and Opinion Mining
- Multimodal Machine Learning Applications
- Advanced Image and Video Retrieval Techniques
- Advanced Text Analysis Techniques
- Text Readability and Simplification
- Speech Recognition and Synthesis
- Human Pose and Action Recognition
- Speech and dialogue systems
Tencent (China)
2022
Xiamen University
2019-2021
In aspect-level sentiment classification (ASC), it is prevalent to equip dominant neural models with attention mechanisms, for the sake of acquiring importance each context word on given aspect. However, such a mechanism tends excessively focus few frequent words polarities, while ignoring infrequent ones. this paper, we propose progressive self-supervised learning approach ASC models, which automatically mines useful supervision information from training corpus refine mechanisms....
We present ClidSum, a benchmark dataset towards building cross-lingual summarization systems on dialogue documents. It consists of 67k+ documents and 112k+ annotated summaries in different target languages. Based the proposed we introduce two settings for supervised semi-supervised scenarios, respectively. then build various baseline paradigms (pipeline end-to-end) conduct extensive experiments ClidSum to provide deeper analyses. Furthermore, propose mDialBART which extends mBART via further...
k-Nearest-Neighbor Machine Translation (kNN-MT) becomes an important research direction of NMT in recent years. Its main idea is to retrieve useful key-value pairs from additional datastore modify translations without updating the model. However, underlying retrieved noisy will dramatically deteriorate model performance. In this paper, we conduct a preliminary study and find that problem results not fully exploiting prediction To alleviate impact noise, propose confidence-enhanced kNN-MT...
Yubin Ge, Ly Dinh, Xiaofeng Liu, Jinsong Su, Ziyao Lu, Ante Wang, Jana Diesner. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
In aspect-level sentiment classification (ASC), it is prevalent to equip dominant neural models with attention mechanisms, for the sake of acquiring importance each context word on given aspect. However, such a mechanism tends excessively focus few frequent words polarities, while ignoring infrequent ones. this paper, we propose progressive self-supervised learning approach ASC models, which automatically mines useful supervision information from training corpus refine mechanisms....
Existing studies for multi-source neural machine translation (NMT) either separately model different source sentences or resort to the conventional single-source NMT by simply concatenating all sentences. However, there exist two drawbacks in these approaches. First, they ignore explicit word-level semantic interactions between sentences, which have been shown effective embeddings of multilingual texts. Second, multiple are simultaneously encoded an model, is unable fully exploit information...
We present ClidSum, a benchmark dataset for building cross-lingual summarization systems on dialogue documents. It consists of 67k+ documents from two subsets (i.e., SAMSum and MediaSum) 112k+ annotated summaries in different target languages. Based the proposed we introduce settings supervised semi-supervised scenarios, respectively. then build various baseline paradigms (pipeline end-to-end) conduct extensive experiments ClidSum to provide deeper analyses. Furthermore, propose mDialBART...
k-Nearest-Neighbor Machine Translation (kNN-MT) becomes an important research direction of NMT in recent years. Its main idea is to retrieve useful key-value pairs from additional datastore modify translations without updating the model. However, underlying retrieved noisy will dramatically deteriorate model performance. In this paper, we conduct a preliminary study and find that problem results not fully exploiting prediction To alleviate impact noise, propose confidence-enhanced kNN-MT...
In aspect-based sentiment analysis (ABSA), many neural models are equipped with an attention mechanism to quantify the contribution of each context word prediction. However, such a suffers from one drawback: only few frequent words polarities tended be taken into consideration for final decision while abundant infrequent ignored by models. To deal this issue, we propose progressive self-supervised learning approach attentional ABSA approach, iteratively perform prediction on all training...