A5014896649- Service-Oriented Architecture and Web Services
- Semantic Web and Ontologies
- Business Process Modeling and Analysis
- Personal Information Management and User Behavior
- Data Management and Algorithms
- Multimodal Machine Learning Applications
- Complex Network Analysis Techniques
- Context-Aware Activity Recognition Systems
- Industrial Technology and Control Systems
- Recommender Systems and Techniques
- Human Pose and Action Recognition
- Topic Modeling
- Caching and Content Delivery
- Scientific Computing and Data Management
- Distributed and Parallel Computing Systems
- Expert finding and Q&A systems
- Advanced Computational Techniques and Applications
- Energy Efficient Wireless Sensor Networks
- Video Analysis and Summarization
- Natural Language Processing Techniques
- Big Data and Business Intelligence
- Graph Theory and Algorithms
- Hydrological Forecasting Using AI
- Mobile Crowdsensing and Crowdsourcing
- Advanced Fluorescence Microscopy Techniques
Army Medical University
2025
Xinqiao Hospital
2025
Kingdee (China)
2012-2024
Zhejiang University of Science and Technology
2021
Zhejiang University
2010-2020
Software (Spain)
2014-2016
China University of Geosciences (Beijing)
2014
Ollscoil na Gaillimhe – University of Galway
2005-2010
Enterprise Ireland
2008
National University of Ireland
2007
This article proposes an approach for identifying and recommending scientific workflows reuse repurposing. Specifically, a workflow is represented as layer hierarchy, which specifies hierarchical relations between this workflow, its sub-workflows, activities. Semantic similarity calculated hierarchies of workflows. A graph-skeleton based clustering technique adopted grouping into clusters. Barycenters in each cluster are identified, refer to core cluster, facilitating identification ranking...
Natural language moment localization aims at localizing video clips according to a natural description. The key this challenging task lies in modeling the relationship between verbal descriptions and visual contents. Existing approaches often sample number of from video, individually determine how each them is related query sentence. However, strategy can fail dramatically, particular when sentence refers some elements that appear outside of, or even are distant from, target clip. In paper,...
In this paper, we tackle a challenging task named video-language segmentation. Given video and sentence in natural language, the goal is to segment object or actor described by frames. To accurately denote target object, given usually refers multiple attributes, such as nearby objects with spatial relations, etc. propose novel Polar Relative Positional Encoding (PRPE) mechanism that represents relations ``linguistic'' way, i.e., terms of direction range. Sentence feature can interact...
Cellular processes such as cell cycle progression, mitosis, apoptosis, and migration are characterized by well-defined events that modulated a function of time. Measuring these in the context time its perturbation small molecule compounds RNAi can provide mechanistic information about cellular pathways being affected. We have used impedance-based time-dependent response profiling (TCRP) to measure characterize responses antimitotic or siRNAs. Our findings indicate mitosis leads unique TCRP....
With a huge volume of geospatial information being collected and number domain-specific functions developed for processing these information, an increasing Open Geospatial Consortium Web services (OWSs) are built available on the accessibility information. Given specific requirement specified by certain user, normally, composition (or chain) OWSs, rather than single OWS, can fulfill this requirement. Consequently, retrieving recommending sub-chains possible service invocations is important...
We propose a novel attentive sequence to translator (ASST) for localizing video clips by natural language descriptions. make two contributions. First, we an mechanism that aligns descriptions and content. A bi-directional Recurrent Neural Network (RNN) parses in directions. Given video-description pair, ASST generates vector representation. Each represents frame, conditioned the description. The representation not only preserves temporal dependencies between frames, but also provides...
In this paper, we propose a novel model for finding experts on given topic using social influence analysis in enterprise network. networks, employees usually talk about some topics relevant to their tasks. With the integration of technology BPM (Business Process Management), more expertise characteristics are reflected by actions networks. Social networks became an important place sharing expertise. We explore potential such as Yammer and IBM Connections, source evidence. work, utilized...
Innovation provides the means for enterprises to survive and grow. Due distributed heterogeneous characteristics, it's difficult manage innovation across extended enterprise. A novel vision Semantic Management (SIM) is introduced. The system architecture combines ontology, inference mediation technologies used create a semantic web of knowledge. functional framework SIM System (SIMS) based on metadata harvesting RDF access presented. An applied case study explained in detail. This, turn,...
The friction pair coatings are highly susceptible to failure in marine corrosive environments. To enhance their service performance, this study focuses on the pairs main shaft transmission of offshore wind turbines and investigates wear corrosion resistance different surface treatment processes. A theory model is developed based dynamic changes contact area predict under dry conditions. results indicate that Interzinc B coating exhibits best overall performance terms friction, wear,...
Knowledge workers rely on collaboration with a large number of knowledge. Even highly experienced knowledge find it difficult to keep up date resources available help them complete particular goal. This paper proposed context-aware resource recommendation system (CARRS) tackle this problem. When worker is working goal, CARRS continuously perceives his/her context, anticipates potential requirements and recommends the most relevant that are helpful for completing
Matrix multiplication is used in a variety of applications. It requires lot computation time especially for large-scale matrices. Parallel processing good choice matrix operation. To overcome the efficiencies existing algorithms parallel multiplication, scheme based on vector linear combination (VLC) was presented. The VLC splits procedure into two steps. first step obtains weighted vectors by scalar multiplication. second gets final result through with identical row numbers. We present...
Enterprise Social Network (ESN) service is getting more popular recently. It can help employees to communicate and collaborate efficiently with colleagues, customers suppliers. One significant phenomenon happening on ESN question & answer: people posting questions the network get answers from friends or friends-of-friends. However, existing platforms do not have good support this process. In paper, we propose a method better answer ESN, purely by using graph analysis approach. Based...
Nowadays, an increasing number of geospatial Web services (GWSs) are built and being available on the for accessibility processing information. Given requirement specified by certain users, normally a composition GWSs, rather than single GWS, can fulfill this requirement. Consequently, retrieving recommending sub-chains possible service invocations is important research challenge. Leveraging semantic similarity between name text description parameters, degree that represents invocation...
With the rapid development of cloud computing technology, cloud-based team collaboration applications are becoming popular on Web. Among all required features for a typical application, shared storage referred documents or produced artifacts by is must-have one. However, existing solutions far from satisfaction. Some them rely self-built infrastructure, which could be big burden, especially those small medium vendors. prevalence personal services, such as Dropbox and Google Drive, more allow...
To fully understand and to better reuse knowledge, it's necessary correlate knowledge with the context under which is generated managed.However, it remains a challenge support awareness in management system.This paper tries overcome difficulties by introducing key enabling technologies for modeling measuring: The ontology correlates its contexts through high level activity based model defined; algorithm Context similarity measurement designed exploiting hierarchical domain structure defined model.
This article proposes to identify and recommend scientific workflows promote their reuse repurposing. Specifically, a workflow is converted into layer hierarchy, which specifies hierarchical relations between this workflow, its sub-workflows, activities. Semantic similarity calculated hierarchies of in order construct network model. A graph-skeleton based clustering method adopted for grouping clusters. Barycenters clusters are identified facilitating cluster identification ranking...
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
Semantic Web services aim at enabling automatic service operations in order to keep human intervention a minimum, enable seamless interaction between the services, reduce risk of operation errors, and maximize efficiency interactions. Current semantic are functionally well-described, but behaviorally poorly-described especially for complex services. This defect may cause behavioral mismatches although provider requester compatible. paper presents description model with public private...