A5101639735- Service-Oriented Architecture and Web Services
- Human Mobility and Location-Based Analysis
- Topic Modeling
- Natural Language Processing Techniques
- Advanced Software Engineering Methodologies
- Semantic Web and Ontologies
- Mobile Crowdsensing and Crowdsourcing
- Complex Network Analysis Techniques
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Advanced Text Analysis Techniques
- Context-Aware Activity Recognition Systems
- Indoor and Outdoor Localization Technologies
- Software Engineering Techniques and Practices
- Software System Performance and Reliability
- Caching and Content Delivery
- Software Engineering Research
- Web Data Mining and Analysis
- Recommender Systems and Techniques
- Advanced SAR Imaging Techniques
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Non-Invasive Vital Sign Monitoring
- Traditional Chinese Medicine Studies
- Opportunistic and Delay-Tolerant Networks
- Text and Document Classification Technologies
Tsinghua University
2015-2024
Shanghai Institute of Technology
2020-2024
Beijing University of Technology
2022
Nanjing University of Science and Technology
2008-2022
Wuhan University of Technology
2020-2021
Beijing Academy of Artificial Intelligence
2021
China Electric Power Research Institute
2020
North China Electric Power University
2020
Beijing University of Posts and Telecommunications
2018
Cornell University
2012-2017
Delay tolerant networks (DTNs) may lack continuous network connectivity. Routing in DTNs is thus challenging since it must handle partitioning, long delays, and dynamic topology such networks. In recent years, social-based approaches, which attempt to exploit social behaviors of DTN nodes make better routing decision, have drawn tremendous interests design. this article, we summarize the properties DTNs, provide a survey approaches. To improve performance, these methods either take...
Event Detection (ED) is a fundamental task in automatically structuring texts. Due to the small scale of training data, previous methods perform poorly on unseen/sparsely labeled trigger words and are prone overfitting densely words. To address issue, we propose novel Enrichment Knowledge Distillation (EKD) model leverage external open-domain knowledge reduce in-built biases frequent annotations. Experiments benchmark ACE2005 show that our outperforms nine strong baselines, especially...
Localization is of great importance in mobile and wireless network applications. Time Difference Arrival (TDOA) one the widely used localization schemes, which target (source) emits a signal number anchors (receivers) record arriving time source signal. By calculating difference different receivers, location estimated. In such scheme, receivers must be precisely synchronized. But synchronization adds computational cost, brings errors may lower accuracy. Previous studies have shown that...
With the increasing stress and unhealthy lifestyles in people's daily life, mental health problems are becoming a global concern. In particular, mood related problems, such as disorders, depressions, elation, seriously impacting quality of life. However, due to complexity unstableness personal mood, assessing analyzing is both difficult inconvenient, which major challenge care. this paper, we propose novel framework called Mood Miner for uses mobile phone data - sensor communication...
In this paper, we propose a synthetic aperture radar (SAR) image despeckling method based on patch ordering and transform-domain filtering. Logarithmic transformation with bias correction is applied to the original SAR transform multiplicative noise model into additive model. Then, adopt two-stage filtering strategy. The first stage coarse which can suppress speckle effectively. stage, extract sliding patches from logarithmic image, order them in smooth way by simplified algorithm specially...
Graph neural network-based recommendation systems are blossoming recently, and its core component is aggregation methods that determine neighbor embedding learning. Prior arts usually focus on how to aggregate information from the perspective of spatial structure information, but temporal about neighbors left insufficiently explored.
Service processes in SOA are composed dynamically by services from different service providers. At run-time, some may become faulty and cause a process to violate its end-to-end quality of (QoS) constraints. We propose an effective approach for replacing only their neighboring maintain the original QoS use iterative algorithm search reconfiguration region that has replaceable meet constraint region. Services regions be replaced using one-to-one, one-to-many, or many-to-one mappings. By...
Event detection is a crucial and challenging sub-task of event extraction, which suffers from severe ambiguity issue trigger words. Existing works mainly focus on using textual context information, while there naturally exist many images accompanied by news articles that are yet to be explored. We believe not only reflect the core events text, but also helpful for disambiguation In this paper, we first contribute an image dataset supplement ED benchmarks (i.e., ACE2005) training evaluation....
Localization is of great importance in mobile and wireless network applications. TDOA one the widely used localization schemes, which a to-be-located object emits signal number receivers record arriving time signal. By calculating difference different receivers, location estimated. In such scheme, must be precisely synchronized, even slight noises are completely unacceptable for centimeter-level localization. Previous studies have shown that existing synchronization approaches low-cost...
Understanding the relationship between sleep and daily life can provide insights into a healthy style since quality is one of most important indicators people's health status. This paper studies extent to which person's be predicted by his/her context information. A combination machine learning technology medical knowledge used study relation quality, so that in real time according relation.We propose novel predicting framework from user data, without requiring users wear special devices. We...
In this paper, we propose an improved attributed scattering model to mathematically unify the models of several canonical primitives. These primitives include not only point- and line-segment-scatterers, such as trihedral, cylinder, dihedral, rectangular plane, but also arc scatterers, sphere top-hat. The estimation parameters can be posed ill-posed linear inverse problem. To overcome ill-posedness, employ incremental sparse Bayesian learning method realize sparsity-driven continuous...
In this letter, we propose an iterative synthetic aperture radar (SAR) image filtering method using the nonlocal sparse model. The original SAR is first transformed to logarithmic domain. Then, use model and regularization technique denoise log-intensity image. each iteration, update noisy then estimate noise variance. For patch in image, find several similar patches stack them together a group. This group filtered by simultaneous coding. all of groups are aggregated form denoised...
The prosperity of massive open online courses provides fodder for plentiful research efforts on adaptive learning. However, current open-access educational datasets are still far from sufficient to meet the need various topics Existing released often cover only small-scale data, lack fine-grained knowledge concepts. They even difficult curate and supplement due platform limitations. In this work, we construct MOOCCubeX, a large, knowledge-centered repository consisting 4,216 courses, 230,263...
As SOC and Web service technology become more widely used, large amounts of services need to be efficiently effectively composed meet complex businesses. In this paper, we proposed an approach resolve the composition problem over large-scale services. We used inverted table as index for a quick discovery, applied dependency graph (SDG) AND/OR algorithm basis parallel compostion. Considering semantic information described in service, our also recognizes transmits relationships Ontology Language (OWL).
Service dependency graph (SDG) is an AND/OR showing input output dependencies among service operations. As in SDG are indirectly expressed by reasoning on data models used interface definitions, their re-usability and expressiveness limited. In this paper, we propose enhanced version of graph, namely SDG+. SDG+ enhances with explicit declaration, which expresses directly static declarations. Based SDG+, developed our automatic composition algorithm for WS-Challenge 2007, wins the...
Objective . To compare the signals of pulse diagnosis fatty liver disease (FLD) patients and cirrhosis patients. Methods After collecting waves with disease, patients, healthy volunteers, we do pretreatment parameters extracting based on harmonic fitting, modeling, identification by unsupervised learning Principal Component Analysis (PCA) supervised Least squares Regression (LS) Absolute Shrinkage Selection Operator (LASSO) cross‐validation step for analysis. Results There is significant...
In this paper, we propose a novel ship detection approach in polarimetric synthetic aperture radar (SAR) images via variational Bayesian inference. First, express the SAR image as tensor, and decompose sum of sparse component associated with ships sea clutter component. These components are denoted by some latent variables. Then, introduce hierarchical priors variables to establish probabilistic model detection. By using inference, estimate posterior distributions Finally, result is obtained...
Addressed in This work is the issue of semantic annotation on Web services. As popularity services increases, automated discovery and composition relevant are more desired. However, current service standards, such as WSDL, not rich enough to fulfill these tasks, because they cannot specify semantics process composition. Thus it necessary describe semantically. Now, OWL standard ontology language, provides powerful features for expressing We an approach annotate with ontology. formalize a...
Semantic information, which is well-regulated and easy to be retrieved, has greatly enriched the expressive ability of Web. These advantages can applied in Web Services meet increasing complexity applications. In this paper, we propose a service composition approach. It combines large-scaled semantic information described WSCpsila09. Besides, QoS become critical issue evaluate performance Being different from improving single services, our approach focuses on overall composition. The...
In delay tolerant networks (DTNs), the lack of continuous connectivity, network partitioning, and long delays make design protocols very challenging. Previous DTN research mainly focuses on routing information propagation. However, with large number wireless devices' participation, how to maintain efficient dynamic topology becomes crucial. this paper, we study control problem in a predictable where time-evolving is known priori or can be predicted. We first model such as directed space-time...
In an academic conference, it is difficult to find people that share similar research interests with you, and also a chore add them into your personal online social network for later communication. Aiming at helping the conference attendees better organize their schedule expand network, we designed developed Find & Connect where used location encounters, together basic services, all through web user interface, in order get homophily physical interactions between users, then base these...
Air pollution has become a striking problem in recent years. When estimating the degree of human exposure to particular air pollutant, time–activity pattern is one most important factors, which able quantify time people spend different micro–environments, such as indoor and outdoor. Traditional surveys use method questionnaires telephone calls explore pattern. In this paper, we propose novel analyse by utilising mobile web usage log. We test on two datasets covering about four million users....
Normally, polarimetric SAR classification is a high-dimensional nonlinear mapping problem. In the realm of pattern recognition, sparse representation very efficacious and powerful approach. As classical descriptors SAR, covariance coherency matrices are Hermitian semidefinite form Riemannian manifold. Conventional Euclidean metrics not suitable for manifold, hence, normal cannot be applied to directly. This paper proposes new land cover approach SAR. There two principal novelties in this...