A5061574958- Fluorine in Organic Chemistry
- Cyclopropane Reaction Mechanisms
- Recommender Systems and Techniques
- Topic Modeling
- Crystallization and Solubility Studies
- X-ray Diffraction in Crystallography
- Synthesis and Reactions of Organic Compounds
- Multimodal Machine Learning Applications
- Asymmetric Synthesis and Catalysis
- Peptidase Inhibition and Analysis
- Modular Robots and Swarm Intelligence
- 3D Shape Modeling and Analysis
- Smart Agriculture and AI
- Neuropeptides and Animal Physiology
- Image Retrieval and Classification Techniques
- Advanced Vision and Imaging
- Evaluation Methods in Various Fields
- Computer Graphics and Visualization Techniques
- Antioxidant Activity and Oxidative Stress
- Organic and Inorganic Chemical Reactions
- Environmental Engineering and Cultural Studies
- Lightning and Electromagnetic Phenomena
- Icing and De-icing Technologies
- Law, logistics, and international trade
- Diabetes Treatment and Management
City University of Hong Kong
2025
Westlake University
2023-2024
ShanghaiTech University
2024
University of Strathclyde
2020-2023
Gezhouba Group (China)
2023
University of Electronic Science and Technology of China
2020-2021
Energy Research Institute
2018
Tufts University
2011-2013
Yulin Normal University
2003-2009
Chinese Academy of Sciences
2004-2006
Recommendation models that utilize unique identities (IDs for short) to represent distinct users and items have been state-of-the-art (SOTA) dominated the recommender systems (RS) literature over a decade. Meanwhile, pre-trained modality encoders, such as BERT [9] Vision Transformer [11], become increasingly powerful in modeling raw features of an item, text images. Given this, natural question arises: can purely modality-based recommendation model (MoRec) outperforms or matches pure...
Accurate prediction of the compressive strength concrete is great significance to construction quality and progress. In order understand current research status in field, a bibliometric analysis relevant literature published this field last decade was conducted first. The 3135 journal articles from 2012 2021 Web Science core database were used as database, knowledge map drawn with help visualisation software CiteSpace 6.1R2 analyse at macro level terms spatial temporal distribution, hotspot...
Bortezomib, a dipeptidyl boronic acid and potent inhibitor of the 26S proteasome, is remarkably effective against multiple myeloma (MM) but not solid tumors. Dose-limiting adverse effects from "on target" inhibition proteasome in normal cells tissues appear to be key obstacle. Achieving efficacy tumors therefore likely require making more selective for tumor tissue over tissues. The simplest strategy that might provide such specificity would employ specific protease release an larger,...
Stereoselective approaches to α-trifluoromethylated α-amino acids (α-Tfm AAs) have been developed. The stereoconfigurations of the resulting α-Tfm AA precursors were well controlled by using different solvents. optically active (S)-2-amino-2-phenyl-1,1,1-trifluoropropanoic acid was synthesized this method.
Large foundational models, through upstream pre-training and downstream fine-tuning, have achieved immense success in the broad AI community due to improved model performance significant reductions repetitive engineering. By contrast, <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">trans</u> ferable one-for-all models xmlns:xlink="http://www.w3.org/1999/xlink">rec</u> ommender system field, referred as TransRec, made limited progress. The...
Bioactive peptides have evolved to optimally fulfill specific biological functions, a fact which has long attracted attention for their use as therapeutic agents. While there been some recent commercial successes fostered in part by advances large-scale peptide synthesis, development of agents significantly impeded inherent susceptibility protease degradation the bloodstream. Here we report that incorporation specially designed amino acid analogues at P1' position, directly C-terminal enzyme...
A highly stereoselective lithium−bromine exchange reaction of 2-trifluoromethyl-3,3-dibromoallylic alcohols is described. (E)- and (Z)-2-trifluoro-3-bromoallylic were obtained in THF hexane, respectively. The lithium carbenoid intermediate was stable even at −40 °C could be trapped by various electrophiles to afford functionalized 2-trifluoromethyl-3-bromoallylic alcohols.
Optically active 2-C-trifluoromethylerythritols, analogues of 2-C-methylerythritol, which is a key intermediate in the biosynthesis isoprenoid with mevalonate-independent route, were conveniently synthesized from 1,1,1-trifluoro-2,3-epoxypropane.
Micro-videos have recently gained immense popularity, sparking critical research in micro-video recommendation with significant implications for the entertainment, advertising, and e-commerce industries. However, lack of large-scale public datasets poses a major challenge developing effective recommender systems. To address this challenge, we introduce very large dataset, named "MicroLens", consisting one billion user-item interaction behaviors, 34 million users, micro-videos. This dataset...
High yields of γ-alkyl, -silyl and -aryl-substituted difluoropropargyl bromides have been obtained by monitoring concentration temperature in the reaction between CF2Br2 γ-substituted lithium acetylide. γ-Silyl substituents afforded corresponding α,α,α-difluorobromopropargyl alcohols good via cleavage tetrabutylammonium fluoride presence an aldehyde.
Whilst industrial robots have been widely used in many sectors, they are predominantly a structured factory environment. In recent years, off-site robotics investigated extensively and there some promising candidates emerging. One such category of is exoskeleton this paper provides an in-depth assessment their suitability assisting human operators undertaking manual operations typically found the construction industry. This work aims to objectively assess advantages disadvantages these two...
Learning a recommender system model from an item's raw modality features (such as image, text, audio, etc.), called MoRec, has attracted growing interest recently. One key advantage of MoRec is that it can easily benefit advances in other fields, such natural language processing (NLP) and computer vision (CV). Moreover, naturally supports transfer learning across different systems through features, known transferable systems, or TransRec. However, so far, TransRec made little progress,...
Abstract ChemInform is a weekly Abstracting Service, delivering concise information at glance that was extracted from about 200 leading journals. To access Abstract, please click on HTML or PDF.
Aiming at the problems of untimely fire warning and unreliable information in traditional systems, this paper designs a building electrical integrated protection system based on multi-sensor data fusion. This fusion fire-fighting hardware module, realizes processing through environmental brightness sensor, smoke temperature infrared flame sensor. At same time, algorithm is designed, number nodes input layer set to 4.At order reduce error, average value white noise variance solved, so as...
An automatic 3D semantic annotation algorithm is proposed based on the associated scene between models. This uses context of models to annotate non-annotated nodes in terms annotated nodes. Firstly, this abstracts extensible (X3D) into X3D conventional tree, transforming tree an in-line node tree. Additionally, ontology for reasoning and Lastly, annotates model files automatically. study analyses time space complexity algorithm. Experimental results confirm that improves not only validity...
Recommendation models that utilize unique identities (IDs) to represent distinct users and items have been state-of-the-art (SOTA) dominated the recommender systems (RS) literature for over a decade. Meanwhile, pre-trained modality encoders, such as BERT ViT, become increasingly powerful in modeling raw features of an item, text images. Given this, natural question arises: can purely modality-based recommendation model (MoRec) outperforms or matches pure ID-based (IDRec) by replacing itemID...