Wenxin Hu

Wenxin Hu


Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

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

Representative Publications

View All Publications
14 October 2020
Verified
Book Chapter
Closed access
PDF file
ACAihua Cai
Aihua Cai
WHWenxin Hu
Wenxin Hu
JZJun Zheng
Jun Zheng

17 May 2019 in Crystals
Verified
Journal Article
Open access
PDF file
HXHao Xuan
Hao Xuan
JLJinfeng Liu
Jinfeng Liu
HLHongyuan Luo
Hongyuan Luo
Yanqiang HanYanqiang Han
Yanqiang Han
WHWenxin Hu
Wenxin Hu
XHXiao He
Xiao He
Molecular crystal plays an important role in many fields of science and technology, but it often crystallizes in different polymorphs with different physical properties. To guide the experimental synthesis of candidate materials, the atomic-scale model is frequently used to predict the most stable polymorph and its structural properties. Here, we show how an ab initio method can be used to achieve a rapid and accurate prediction of sulfathiazole crystal polymorphs (an antibiotic drug), based on the Gibbs free energy calculation and Raman spectra analysis. At the atmospheric pressure and the temperature of 300 K, we demonstrate that form III (FIII) is the most stable structure of sulfathiazole. The agreement between the predicted and experimental crystal structures corresponds to the order of stability for five sulfathiazole polymorphs as FI < FV < FIV < FII < FIII, which is achieved by employing the density functional theory (DFT) calculations.

19 November 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JCJiayi Chen
Jiayi Chen
WWWen Wu
Wen Wu
LSLiye Shi
Liye Shi
Yu JiYu Ji
Yu Ji
WHWenxin Hu
Wenxin Hu
Liang HeLiang He
Liang He
Sequential recommendations have made great strides in accurately predicting the future behavior of users. However, seeking accuracy alone may bring side effects such as unfair and overspecialized recommendation results. In this work, we focus on the calibrated recommendations for sequential recommendation, which is connected to both fairness and diversity. On the one hand, it aims to provide fairer recommendations whose preference distributions are consistent with users’ historical behaviors. On the other hand, it can improve the diversity of recommendations to a certain degree. But existing methods for calibration have mainly relied on the post-processing on the candidate lists, which require more computation time in generating recommendations. In addition, they fail to establish the relationship between accuracy and calibration, leading to the limitation of accuracy. To handle these problems, we propose an end-to-end framework to provide both accurate and calibrated recommendations for sequential recommendation. We design an objective function to calibrate the interests between recommendation lists and historical behaviors. We also provide distribution modification approaches to improve the diversity and mitigate the effect of imbalanced interests. In addition, we design a decoupled-aggregated model to improve the recommendation. The framework assigns two objectives to two individual sequence encoders, and aggregates the outputs by extracting useful information. Experiments on benchmark datasets validate the effectiveness of our proposed model.

28 September 2018 in Molecules
Verified
Journal Article
Open access
PDF file
PLPengfei Li
Pengfei Li
FLFengjiao Liu
Fengjiao Liu
XJXiangyu Jia
Xiangyu Jia
Yihan ShaoYihan Shao
Yihan Shao
WHWenxin Hu
Wenxin Hu
YMYe Mei
Ye Mei
For Diels–Alder (DA) reactions in solution, an accurate and converged free energy (FE) surface at ab initio (ai) quantum mechanical/molecular mechanical (QM/MM) level is imperative for the understanding of reaction mechanism. However, this computation is still far too expensive. In a previous work, we proposed a new method termed MBAR+wTP, with which the computation of the ai FE profile can be accelerated by several orders of magnitude via a three-step procedure: (I) an umbrella sampling (US) using a semi-empirical (SE) QM/MM Hamiltonian is performed; (II) the FE profile is generated using the Multistate Bennett Acceptance Ratio (MBAR) analysis; and (III) a weighted Thermodynamic Perturbation (wTP) from the SE Hamiltonian to the ai Hamiltonian is performed to obtain the ai QM/MM FE profile using weight factors from the MBAR analysis. In this work, this method is extended to the calculations of two-dimensional FE surfaces of two Diels–Alder reactions of cyclopentadiene with either acrylonitrile or 1-4-naphthoquinone at ai QM/MM level. The accurate activation free energies at the ai QM/MM level, which are much closer to the experimental measurements than those calculated by other methods, indicate that this MBAR+wTP method can be applied in the studies of complex reactions in condensed phase with much-enhanced efficiency.

24 March 2024 in Proceedings of the AAAI Conference on Artificial Intelligence
Verified
Journal Article
Open access
PDF file
YWYe Wang
Ye Wang
HPHuazheng Pan
Huazheng Pan
TZTao Zhang
Tao Zhang
WWWen Wu
Wen Wu
WHWenxin Hu
Wenxin Hu
The goal of document-level relation extraction (RE) is to identify relations between entities that span multiple sentences. Recently, incomplete labeling in document-level RE has received increasing attention, and some studies have used methods such as positive-unlabeled learning to tackle this issue, but there is still a lot of room for improvement. Motivated by this, we propose a positive-augmentation and positive-mixup positive-unlabeled metric learning framework (P3M). Specifically, we formulate document-level RE as a metric learning problem. We aim to pull the distance closer between entity pair embedding and their corresponding relation embedding, while pushing it farther away from the none-class relation embedding. Additionally, we adapt the positive-unlabeled learning to this loss objective. In order to improve the generalizability of the model, we use dropout to augment positive samples and propose a positive-none-class mixup method. Extensive experiments show that P3M improves the F1 score by approximately 4-10 points in document-level RE with incomplete labeling, and achieves state-of-the-art results in fully labeled scenarios. Furthermore, P3M has also demonstrated robustness to prior estimation bias in incomplete labeled scenarios.

8 April 2022
Verified
Book Chapter
Closed access
PDF file
QBQimin Ban
Qimin Ban
WWWen Wu
Wen Wu
WHWenxin Hu
Wenxin Hu
HLHui Lin
Hui Lin
WZWei Zheng
Wei Zheng
Liang HeLiang He
Liang He