- IoT and Edge/Fog Computing
- Privacy-Preserving Technologies in Data
- Topic Modeling
- Caching and Content Delivery
- Cloud Computing and Resource Management
- Peer-to-Peer Network Technologies
- Advanced Image and Video Retrieval Techniques
- Advanced Data Storage Technologies
- Network Security and Intrusion Detection
- Blockchain Technology Applications and Security
- Security in Wireless Sensor Networks
- Web Data Mining and Analysis
- Service-Oriented Architecture and Web Services
- Energy Efficient Wireless Sensor Networks
- Software Engineering Research
- Natural Language Processing Techniques
- Mobile Crowdsensing and Crowdsourcing
- Mobile Ad Hoc Networks
- Semantic Web and Ontologies
- Multimodal Machine Learning Applications
- Network Traffic and Congestion Control
- Mobile Agent-Based Network Management
- Cloud Data Security Solutions
- Cryptography and Data Security
- Advanced Computational Techniques and Applications
First Affiliated Hospital of Anhui Medical University
2023-2025
Anhui Medical University
2023-2025
Beijing Center for Disease Prevention and Control
2012-2025
Hunan University
2008-2024
Beihang University
2023-2024
National University of Defense Technology
2012-2024
Chinese Academy of Sciences
2009-2024
Wenzhou Medical University
2024
Zhejiang Lab
2024
Taizhou University
2024
Driven by the visions of Internet Things and 5G communications, edge computing systems integrate computing, storage, network resources at to provide infrastructure, enabling developers quickly develop deploy applications. At present, have received widespread attention in both industry academia. To explore new research opportunities assist users selecting suitable for specific applications, this survey paper provides a comprehensive overview existing introduces representative projects. A...
A data breach is the intentional or inadvertent exposure of confidential information to unauthorized parties. In digital era, has become one most critical components an enterprise. Data leakage poses serious threats organizations, including significant reputational damage and financial losses. As volume growing exponentially breaches are happening more frequently than ever before, detecting preventing loss pressing security concerns for enterprises. Despite a plethora research efforts on...
Federated learning (FL) has received considerable attention with the development of mobile internet technology, which is an emerging framework to train a deep model from decentralized data. Modern devices often have access rich but privacy-sensitive data, and computational abilities are limited because hardware restriction. In previous works based on federated averaging (FedAvg) algorithm, need perform lots calculations, it time-consuming in process global communication. Inspired by edge...
The pervasive view of the mobile crowd bridges various real-world scenes and people's perceptions with gathering distributed crowdsensing photos. To elaborate informative visuals for viewers, existing techniques introduce photo selection as an essential step in crowdsensing. Yet, aesthetic preference at very heart their experiences under contexts (e.g., travel planning), is seldom considered hardly guaranteed. We propose CrowdPicker, a novel framework adaptive awareness With observations on...
Noun phrases in queries are identified and classified into four types: proper names, dictionary phrases, simple complex phrases. A document has a phrase if all content words the within window of certain size. The sizes for different types determined using decision tree. Phrases more important than individual terms. Consequently, documents response to query ranked with matching given higher priority. We utilize WordNet disambiguate word senses Whenever sense term is determined, its synonyms,...
Current web search engines are built to serve all users, independent of the needs any individual user. Personalization is carry out retrieval for each user incorporating his/her interests. We propose a novel technique map query set categories, which represent user's intention. This categories can as context disambiguate words in query. A profile and general learned from history category hierarchy respectively. These two profiles combined into categories. Several learning combining algorithms...
Though destructive to network functions, insider attackers are not detectable with only the classic cryptography-based techniques. Many mission-critic sensor applications demand an effective, light, flexible algorithm for internal adversary identification localized information available. The attacker detection scheme proposed in this paper meets all requirements by exploring spatial correlation existent among networking behaviors of sensors close proximity. Our work is exploratory that...
This paper presents the first version of NIST Cloud Computing Reference Architecture (RA). is a vendor neutral conceptual model that concentrates on role and interactions identified actors in cloud computing sphere. Five primary were - Service Consumer, Provider, Broker, Auditor Carrier. Their roles activities are discussed this report. A goal for generating was to give United States Government (USG) method understanding communicating components system Federal IT executives, Program Managers...
A central problem in sensor network security is that sensors are susceptible to physical capture attacks. Once a compromised, the adversary can easily launch clone attacks by replicating compromised node, distributing clones throughout network, and starting variety of insider Previous works against suffer from either high communication/storage overhead or poor detection accuracy. In this paper, we propose novel scheme for detecting networks, which computes each social fingerprint extracting...
Inter-Component Communication (ICC) provides a message passing mechanism for data exchange between Android applications. It has been long believed that inter-app ICCs can be abused by malware writers to launch collusion attacks using two or more apps. However, because of the complexity performing pairwise program analysis on apps, scale existing analyses is too small (e.g., up several hundred) produce concrete security evidence. In this paper, we report our findings in first large-scale...
Model compression is significant for the wide adoption of Recurrent Neural Networks (RNNs) in both user devices possessing limited resources and business clusters requiring quick responses to large-scale service requests. This work aims learn structurally-sparse Long Short-Term Memory (LSTM) by reducing sizes basic structures within LSTM units, including input updates, gates, hidden states, cell states outputs. Independently can result inconsistent dimensions among them, consequently, end up...
The market for cloud backup services in the personal computing environment is growing due to large volumes of valuable and corporate data being stored on desktops, laptops smart phones. Source deduplication has become a mainstay that saves network bandwidth reduces storage space. However, there are two challenges facing service clients: (1) low efficiency combination resource-intensive nature limited system resources PC-based client site, (2) transfer since post-deduplication transfers from...
Clinical named entity recognition (CNER), which intends to automatically detect clinical entities in electronic health record (EHR), is a committed step for further text mining. Recently, more and deep learning models are used Chinese CNER. However, these do not make full use of the information EHR, either word-based or character-based. In addition, neural tend be locally unstable even tiny perturbation may mislead them. this paper, we firstly propose novel adversarial training based lattice...
Chip Multi-Processor (CMP) architectures have recently become a mainstream computing platform. Recent CMPs allow cores to share expensive resources, such as the last level cache and off-chip pin bandwidth. To improve system performance reduce volatility of individual threads, bandwidth partitioning schemes been proposed. While how affects is well understood, little understood regarding performance, interact with one another. In this paper, we propose simple yet powerful analytical model that...
The Internet of Things (IoT) is an important part the new generation information technology. It utilizes many sensor nodes to collect data and monitor environment can be applied in various fields. However, because energy limited batteries cannot replaced, lifetimes are by their batteries. Only a reasonable method for saving reduce loss communication process. Many studies have found that property battery-friendliness; is, if they continue working, drops rapidly. rest after working period...
Joint extraction of entities and their relations benefits from the close interaction between named relation information. Therefore, how to effectively model such cross-modal interactions is critical for final performance. Previous works have used simple methods as label-feature concatenation perform coarse-grained semantic fusion among instances, but fail capture fine-grained correlations over token label spaces, resulting in insufficient interactions. In this paper, we propose a deep...
IT companies need to monitor various Key Performance Indicators (KPIs) and detect anomalies in real time ensure the quality reliability of Internet-based services. However, due diversity KPIs, ambiguity scarcity lack labels, anomaly detection for KPIs has been a great challenge. Existing KPI methods have not explored properties detail our best knowledge. Therefore, we explore recognize common important form named <italic xmlns:mml="http://www.w3.org/1998/Math/MathML"...
With the widespread application of infotainment services in intelligent connected vehicles (ICVs), network traffic has grown exponentially, bringing huge burden and energy consumption to ICV network. Edge caching, which enables edges [e.g., or roadside units (RSUs)] with cache storages, is a promising technology alleviate this problem. In article, terms hybrid communication mode vehicle (V2V) RSU (V2R), an energy-aware caching scheme for proposed. Considering geographical distribution RSUs...
Joint extraction of entities and their relations benefits from the close interaction between named relation information. Therefore, how to effectively model such cross-modal interactions is critical for final performance. Previous works have used simple methods, as label-feature concatenation, perform coarse-grained semantic fusion among instances but fail capture fine-grained correlations over token label spaces, resulting in insufficient interactions. In this article, we propose a dynamic...
Pre-trained large language models (LLMs) have powerful capabilities for generating creative natural text. Evolutionary algorithms (EAs) can discover diverse solutions to complex real-world problems. Motivated by the common collective and directionality of text sequence generation evolution, this paper illustrates strong consistency LLMs EAs, which includes multiple one-to-one key characteristics: token embedding genotype-phenotype mapping, position encoding fitness shaping, selection,...
Background Vaccination is an effective measure to prevent herpes zoster and its related complications. The coverage of vaccination extremely low in China, there a notable lack research investigating the barriers promoting vaccine China. Objectives This study aims survey status (HZ) associated factors among individuals aged 50 years older, it also seeks identify dissemination, thereby providing scientific foundation for enhancing rate zoster. Methods From March August 2023, questionnaire was...