- Topic Modeling
- Advanced Graph Neural Networks
- Data Quality and Management
- Semantic Web and Ontologies
- Natural Language Processing Techniques
- Biomedical Text Mining and Ontologies
- Text and Document Classification Technologies
- Landslides and related hazards
- Domain Adaptation and Few-Shot Learning
- Multimodal Machine Learning Applications
- Cloud Computing and Resource Management
- Hydrological Forecasting Using AI
- Bayesian Modeling and Causal Inference
- Genetic Associations and Epidemiology
- Karst Systems and Hydrogeology
- Dam Engineering and Safety
- Machine Learning in Healthcare
- IoT and Edge/Fog Computing
- Bioinformatics and Genomic Networks
- Complex Network Analysis Techniques
- Advanced Text Analysis Techniques
- Gene expression and cancer classification
- Nuclear Physics and Applications
- Image and Video Quality Assessment
- Expert finding and Q&A systems
Hong Kong Polytechnic University
2025
Hohai University
2025
Anhui University
2024
Tencent (China)
2021-2023
Tsinghua University
2017-2023
University Town of Shenzhen
2023
PLA Army Engineering University
2022
Tianjin Normal University
2020-2022
Guangzhou Metro Group (China)
2021
Donghua University
2019-2020
With the advent of big data age, volume has been changed from trillionbyte to petabyte with incredible speed. Owing fact that cloud storage offers vision a virtually infinite pool resources, can be stored and accessed high scalability availability. But single cloud-based risks like vendor lock-in, privacy leakage, unavailability. Multi-cloud mitigate these geographically located providers. In this scheme, one important challenge is how place user's cost-effectively paper, an architecture for...
Existing knowledge graph (KG) embedding models have primarily focused on static KGs. However, real-world KGs do not remain static, but rather evolve and grow in tandem with the development of KG applications. Consequently, new facts previously unseen entities relations continually emerge, necessitating an model that can quickly learn transfer through growth. Motivated by this, we delve into expanding field this paper, i.e., lifelong embedding. We consider retention learning growing snapshots...
Abstract Cadmium (Cd) accumulation in rice is a serious threat to food safety and human health. Breeding varieties with low Cd one of the most effective approaches reducing health risks from Cd-polluted rice. However, genetic basis grains, especially indica varieties, has not been fully elucidated. The evaluation Cd-accumulation capacity was conducted among 338 diverse accessions grown Cd-contaminated soils different contents. Thirteen lines relatively accumulation, including six lines, were...
Identifying the key components in complex systems is crucial for ensuring functionality and integrity of system. For directed networks, both nodes edges are critical components, but little attention has been drawn to important edges. In this paper, a novel method, Cooperation–Competition Centrality (Co-Com Centrality), proposed quantify edge importance network. This method utilizes cooperation competition relationships integrate local global information about We compare performance Co-Com...
Abstract Cloud computing can provide users with basic hardware resources, and there are three instance types: reserved instances, on-demand instances spot instances. The price of is lower than others on average, but it fluctuates according to market demand supply. When a user requests instance, he/she needs give bid. Only if the bid not price, obtain right use this instance. Thus, very important challenging predict To end, we take most popular representative Amazon EC2 as testbed, history...
Over the years, reasoning over knowledge graphs (KGs), which aims to infer new conclusions from known facts, has mostly focused on static KGs. The unceasing growth of in real life raises necessity enable inductive ability expanding Existing work assumes that entities all emerge once a batch, oversimplifies scenario continually appear. This study dives into more realistic and challenging setting where multiple batches. We propose walk-based model tackle setting. Specifically, graph...
Visual Relation Detection (VRD) aims to detect relationships between objects for image understanding. Most existing VRD methods rely on thousands of training samples each relationship achieve satisfactory performance. Some recent papers tackle this problem by few-shot learning with elaborately designed pipelines and pre-trained word vectors. However, the performance models is severely hampered poor generalization capability, as they struggle handle vast semantic diversity visual...
With the development of cloud computing, market is becoming more and complicated. There are many providers different instance types, which brings users some confusion when they select types. In order to solve type selection problem in multi-cloud environment, a Cloud Instance Type Selection Algorithm based on Genetic (CITSA-GA) proposed. CITSA-GA mainly includes two-dimensional encoding with constraint between adjacent genes, operation adopting elite retention strategy roulette strategy,...
Zero-resource cross-lingual named entity recognition (ZRCL-NER) aims to leverage rich labeled source language data address the NER problem in zero-resource target language. Existing methods are built either based on transfer or representation transfer. However, former usually leads additional computation costs, and latter lacks explicit optimization specific task. To overcome above limitations, we propose a novel prototype-based alignment model (PRAM) for challenging ZRCL-NER PRAM models...
Genome-wide association study (GWAS), as one primary approach for genetic studies, has been successfully applied to a variety of complex diseases, leading the discovery substantial disease-associated loci. These discovered associations provide unprecedented opportunities deepening our understanding such risk variants, genes, and pathways. However, it is non-trivial extract biological knowledge from GWAS data due existence several non-negligible factors. For example, majority associated loci...
In view of the nonuniform landslide driving force in practical cases and common landslide–pipeline separation phenomenon, a new pipeline–soil model under action parabolic is proposed this paper, closed analytical solutions for are derived based on elastic foundation theory. Moreover, case long transportation pipeline crossing Luozhentian was studied using method, comparisons between results finite-element method (FEM) also made to validate exactitude solutions. addition, deflection stress...
In the community of Linked Data, anyone can publish their data as Data on web because openness Semantic Web. As such, RDF (Resource Description Framework) triples described same real-world entity be obtained from multiple sources; it inevitably results in conflicting objects for a certain predicate entity. The objective this study is to identify one truth An intuitive principle based common sense that an object reliable source trustworthy; thus, provide trustworthy reliable. Many discovery...