A5016532375- Topic Modeling
- Recommender Systems and Techniques
- Breast Cancer Treatment Studies
- Advanced Graph Neural Networks
- Breast Implant and Reconstruction
- Energy Load and Power Forecasting
- Breast Lesions and Carcinomas
- Information Retrieval and Search Behavior
- Advanced Bandit Algorithms Research
- Cancer Treatment and Pharmacology
- Graph Theory and Algorithms
- Domain Adaptation and Few-Shot Learning
- Photorefractive and Nonlinear Optics
- Photonic and Optical Devices
- Advanced Battery Materials and Technologies
- Natural product bioactivities and synthesis
- Aquaculture disease management and microbiota
- Cancer-related molecular mechanisms research
- RNA modifications and cancer
- RNA Research and Splicing
- Global Cancer Incidence and Screening
- Epigenetics and DNA Methylation
- Molecular Biology Techniques and Applications
- HER2/EGFR in Cancer Research
- Glycosylation and Glycoproteins Research
Beijing Normal University
2023-2025
Huazhong University of Science and Technology
2025
University of Chinese Academy of Sciences
2024
Shanghai Institute of Microsystem and Information Technology
2024
Jiangsu Province Hospital
2015-2024
Nanjing Medical University
2013-2024
Tsinghua University
2024
Shandong Normal University
2023-2024
Soochow University
2024
Anhui University of Traditional Chinese Medicine
2021-2023
We elucidate the dissociation mechanism of LiFSI induced by ferroelectric fillers BaTiO 3 and enhanced spontaneous polarization oxygen vacancy defects.
Knowledge Graph Embedding (KGE) aims to learn representations for entities and relations. Most KGE models have gained great success, especially on extrapolation scenarios. Specifically, given an unseen triple (h, r, t), a trained model can still correctly predict t from ?), or h (?, such ability is impressive. However, most existing works focus the design of delicate modeling function, which mainly tells us how measure plausibility observed triples, but offers limited explanation why methods...
Knowledge graph (KG) entity typing focuses on inferring possible type instances, which is a significant subtask of knowledge completion (KGC). Existing methods usually exploit the representation to model transmission between entities and their types, cannot fully explore fine-grained identifying semantic an entity. To address these issues, we propose Neighborhood-Attention Neural Fine-Grained Entity Typing (AttEt), considers neighborhood information from KGs bridge types together. In this...
Product search is crucial for users to find and purchase products they need. Personalized product search, which models users’ intent provides tailored results, has become a prominent research problem in industry academia. Recent studies often leverage knowledge graphs (KGs) improve performance generate explanations results. However, existing KG-based methods treat explanation tasks separately explore paths KGs as explanations, creating gap between results generated explanations. Also,...
User behavior records serve as the foundation for recommender systems. While data exhibits ease of acquisition, it often suffers from varying quality. Current methods employ valuation to discern high-quality low-quality data. However, they tend black-box design, lacking transparency and interpretability. Besides, are typically tailored specific evaluation metrics, leading limited generality across various tasks. To overcome these issues, we propose an explainable versatile framework DVR...
Composite solid electrolytes (CSEs) are promising candidates for solid‐state lithium metal batteries. However, the poor cross‐phase Li+ transport restricts rate performance and cycle life of Herein, we revealed percolation behavior in poly(vinylidene fluoride) (PVDF)‐based CSEs with Li6.4La3Zr1.4Ta0.6O12 filler. The de‐coordination barrier from clusters determines interfacial capability. We then employed a designed N‐methyl‐2,2,2‐trifluoroacetamide (NMTFA) ligand to lower energy activate...
Composite solid electrolytes (CSEs) are promising candidates for solid‐state lithium metal batteries. However, the poor cross‐phase Li+ transport restricts rate performance and cycle life of Herein, we revealed percolation behavior in poly(vinylidene fluoride) (PVDF)‐based CSEs with Li6.4La3Zr1.4Ta0.6O12 filler. The de‐coordination barrier from clusters determines interfacial capability. We then employed a designed N‐methyl‐2,2,2‐trifluoroacetamide (NMTFA) ligand to lower energy activate...
Product search has been an important way for people to find products on online shopping platforms. Existing approaches in personalized product mainly embed user preferences into one single vector. However, this simple strategy easily results sub-optimal representations, failing model and disentangle user's multiple preferences. To overcome problem, we proposed a category-aware multi-interest encode users as preference embeddings represent user-specific interests. Specifically, also capture...
Selecting reliable negative training instances is the challenging task in implicit feedback-based recommendation, which optimized by pairwise learning on user feedback data. The existing methods usually exploit various samplers (i.e., heuristic-based or GAN-based sampling) data to improve quality of samples. However, these focused maintaining hard samples with a high gradient for training, causing false be selected preferentially. limitation noise amplification may lead overfitting and...
Totally implantable venous access ports are widely used in chemotherapy for malignant tumors. This retrospective study investigated the safety, technical feasibility, and device-related complications of totally implanted upper arm.Between May 2016 June 2018, 570 women with early breast cancer received were successfully a port arm. Device-related collected influencing factors analyzed major complications.Only one case underwent premature removal before end chemotherapy. observed 32 cases,...
Convolutional neural networks (CNN), recurrent (RNN), attention, and their variants are extensively applied in the sentiment analysis, effect of fusion model is expected to be better. However, confronted with some problems such as complicated structure, excessive trainable parameters, long training time. The classification traditional cross entropy loss function undesirable since sample category imbalance well ease difficulty not taken into account. In order solve these problems,...
Abstract Fish live in an aquatic environment rich various microorganisms and pathogens. mucosal-associated lymphoid tissue (MALT) plays a very important role immune defence. This study was conducted to characterize the response mediated by Cc IgZ3 common carp ( Cyprinus carpio. ) investigate proliferating + B lymphocytes gill. We determined expression of many different tissues following stimulation intraperitoneal injection TNP-LPS (2,4,6-Trinitrophenyl hapten conjugated lipopolysaccharide)...
The rate of breast-conserving surgery (BCS) is low in China. Many patients choose mastectomy even when informed that there no difference the overall survival compared with BCS plus radiotherapy. This study aimed to investigate factors influenced surgical choice eligible for BCS.Female breast carcinoma were enrolled a single center from March 2016 January 2017. They made their own decision regarding approach. Univariate analysis was employed determine associated different approaches....
Abstract Background: The tumor abnormal protein (TAP) test is used to screen for many cancers, but its use breast cancer has not been studied. Methods: Tests carcinoembryonic antigen (CEA), 125 (CA125), 15-3 (CA15-3), and TAP were administered 261 women with operable benign disease 348 cancer. cutoff value was the mean + 3 standard deviations patients (275.64 μm 2 ). Sensitivities specificities of single biomarker tests combined compared. defined as positive if any positive, negative...
Aim: Compare the two-weekly regimens of nanoparticle albumin-bound paclitaxel (nab-P) with solvent-based (sb-P) as neoadjuvant chemotherapy for breast cancer. Materials & methods: Patients (n = 162) operable early cancer received four cycles dose-dense epirubicin and cyclophosphamide followed by nab-P 83) or sb-P 79), trastuzumab when needed. Results: Across all patients, ypT0 ypN0 ypT0/is pathological complete response rates in group were not superior to those group. However,...