Piji Li

Piji Li

Auto-generated profile, based on publicly available information

Piji Li

Nanjing University of Aeronautics and Astronautics, Nanjing, China

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2020
Verified
Conference Paper
Open access
PDF file
WCWang Chen
Wang Chen
HCHou Pong Chan
Hou Pong Chan
PLPiji Li
Piji Li
IKIrwin King
Irwin King
Wang Chen, Hou Pong Chan, Piji Li, Irwin King. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020.

26 June 2023 in Proceedings of the AAAI Conference on Artificial Intelligence
Verified
Journal Article
Open access
PDF file
YLYougang Lyu
Yougang Lyu
PLPiji Li
Piji Li
YYYechang Yang
Yechang Yang
MRMaarten De Rijke
Maarten De Rijke
Pengjie RenPengjie Ren
Pengjie Ren
ZRZhaochun Ren
Zhaochun Ren
Natural language understanding (NLU) models often rely on dataset biases rather than intended task-relevant features to achieve high performance on specific datasets. As a result, these models perform poorly on datasets outside the training distribution. Some recent studies address this issue by reducing the weights of biased samples during the training process. However, these methods still encode biased latent features in representations and neglect the dynamic nature of bias, which hinders model prediction. We propose an NLU debiasing method, named debiasing contrastive learning (DCT), to simultaneously alleviate the above problems based on contrastive learning. We devise a debiasing, positive sampling strategy to mitigate biased latent features by selecting the least similar biased positive samples. We also propose a dynamic negative sampling strategy to capture the dynamic influence of biases by employing a bias-only model to dynamically select the most similar biased negative samples. We conduct experiments on three NLU benchmark datasets. Experimental results show that DCT outperforms state-of-the-art baselines on out-of-distribution datasets while maintaining in-distribution performance. We also verify that DCT can reduce biased latent features from the model's representation.

24 September 2022
Verified
Book Chapter
Closed access
PDF file
YSYundi Shi
Yundi Shi
PLPiji Li
Piji Li
CYChangchun Yin
Changchun Yin
ZHZhaoyang Han
Zhaoyang Han
LZlu zhou
lu zhou
ZLZhe Liu
Zhe Liu

1 January 2024
Verified
Conference Paper
Open access
PDF file
RWRenzhi Wang
Renzhi Wang
PLPiji Li
Piji Li
Renzhi Wang, Piji Li. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.

1 January 2024
Verified
Conference Paper
Open access
PDF file
SCSirry Chen
Sirry Chen
SFShuo Feng
Shuo Feng
LSLiang Songsong
Liang Songsong
CZChen-Chen Zong
Chen-Chen Zong
JLJing Li
Jing Li
PLPiji Li
Piji Li
Sirry Chen, Shuo Feng, Liang Songsong, Chen-Chen Zong, Jing Li, Piji Li. Findings of the Association for Computational Linguistics ACL 2024. 2024.

30 April 2023
Verified
Conference Paper
Open access
CYCongchi Yin
Congchi Yin
PLPiji Li
Piji Li
ZRZhaochun Ren
Zhaochun Ren
Dialogue structure discovery is essential in dialogue generation. Well-structured topic flow can leverage background information and predict future topics to help generate controllable and explainable responses. However, most previous work focused on dialogue structure learning in task-oriented dialogue other than open-domain dialogue which is more complicated and challenging. In this paper, we present a new framework CTRLStruct for dialogue structure learning to effectively explore topic-level dialogue clusters as well as their transitions with unlabelled information. Precisely, dialogue utterances encoded by bi-directional Transformer are further trained through a special designed contrastive learning task to improve representation. Then we perform clustering to utterance-level representations and form topic-level clusters that can be considered as vertices in dialogue structure graph. The edges in the graph indicating transition probability between vertices are calculated by mimicking expert behavior in datasets. Finally, dialogue structure graph is integrated into dialogue model to perform controlled response generation. Experiments on two popular open-domain dialogue datasets show our model can generate more coherent responses compared to some excellent dialogue models, as well as outperform some typical sentence embedding methods in dialogue utterance representation. Code is available in GitHub1.