A5100414975- Cryptography and Data Security
- Recommender Systems and Techniques
- Blockchain Technology Applications and Security
- Privacy-Preserving Technologies in Data
- Chaos-based Image/Signal Encryption
- Wheat and Barley Genetics and Pathology
- Human Mobility and Location-Based Analysis
- Fault Detection and Control Systems
- Sentiment Analysis and Opinion Mining
- Digital Marketing and Social Media
- Lattice Boltzmann Simulation Studies
- Fuel Cells and Related Materials
- Topic Modeling
- Advanced Graph Neural Networks
- Anomaly Detection Techniques and Applications
- Electrocatalysts for Energy Conversion
- Genetic Mapping and Diversity in Plants and Animals
- Internet Traffic Analysis and Secure E-voting
- Cryptographic Implementations and Security
- Advanced Bandit Algorithms Research
- Chromosomal and Genetic Variations
- Sparse and Compressive Sensing Techniques
- Opinion Dynamics and Social Influence
- Peer-to-Peer Network Technologies
- Genetics and Plant Breeding
Henan University
2015-2025
Dalian Maritime University
2022-2023
Beijing University of Technology
2021-2023
Shanghai University of Finance and Economics
2008
Outlier detection refers to the identification of data points that deviate from a general distribution. Existing unsupervised approaches often suffer high computational cost, complex hyperparameter tuning, and limited interpretability, especially when working with large, high-dimensional datasets. To address these issues, we present simple yet effective algorithm called ECOD (Empirical-Cumulative-distribution-based Detection), which is inspired by fact outliers are "rare events" appear in...
As the public cloud becomes one of leading ways in data-sharing nowadays, data confidentiality and user privacy are increasingly critical. Partially policy-hidden ciphertext policy attribute-based encryption (CP-ABE) can effectively protect while reducing leakage by hiding part access structure. However, it cannot satisfy need sharing with complex users large amounts data, both terms less expressive structures limited granularity hiding. Moreover, verification right to shared correctness...
Abstract Timed-release encryption (TRE) is a cryptographic primitive that can control the decryption time and has significant application value in time-sensitive scenarios. To solve reliability issue of nodes existing TRE anonymous interaction schemes, we propose blockchain-based protocol for query trapdoors. In our protocol, recipient divides encrypted trapdoor request information into n ciphertext fragments using secret sharing technology near time, employs idea onion routing to perform...
To break the narrow diversity bottleneck of wheat D genome, a set Aegilops tauschii-wheat introgression (A-WI) lines was developed by crossing Ae. tauschii accession T015 with common elite cultivar Zhoumai 18 (Zhou18). A high-density genetic map constructed based on Single Nucleotide Polymorphism (SNP) markers and 15 yield-related traits were evaluated in 11 environments for detecting quantitative trait loci (QTL). total 27 environmentally stable QTL identified at least five environments, 20...
Remote sensing image scene classification takes blocks as units and predicts their semantic descriptors. Because it is difficult to obtain enough labeled samples for all classes of remote scenes, zero-shot methods which can recognize scenes that are not seen in the training stage great significance. By projecting visual features class into latent space ensuring alignment, variational autoencoder (VAE) generative model has been applied address remote-sensing under a setting. However, VAE...
Next Point-of-Interest (POI) recommendation has shown great value for both users and providers in location-based services. Existing methods mainly rely on partial information users’ check-in sequences, are brittle to with few interactions. Moreover, they ignore the impact of multi-dimensional auxiliary such as user frequency, POI category preferences modeling dynamic changes over different time periods performance. To address above limitations, we propose a novel method next by long short...
User influence in online social networks has been measured by different metrics and algorithms, these methods roughly fall into two main genres: attribute-based approaches graph-based approaches. However, most only consider single metric, such as total view counts or retweet counts. And cannot apply to the platforms where it is difficult obtain graphs of some metrics. In this paper, we propose a triangular fuzzy number-based method measure user influence, which covers multiple graph free. By...
Cryptographic primitive of timed-release encryption (TRE) enables the sender to encrypt a message which only allows designated receiver decrypt after time. Combined with other technologies, TRE technology is applied variety scenarios, including regularly posting on social network and online sealed bidding. Nowadays, in order control decryption time while maintaining anonymity user identities, most solutions adopt noninteractive server mode periodically broadcast trapdoors, but because these...
To address privacy concerns, data in the blockchain should be encrypted advance to avoid access from all users blockchain. However, cannot directly retrieved, which hinders sharing Several works have been proposed deal with this problem. retrieval these schemes requires participation of owners and lacks finer-grained control. In paper, we propose an attribute-based keyword search scheme over blockchain, allows files based on their attributes. addition, build a file chain structure improve...
Proof of stake consensus algorithm (PoS) has the advantage not requiring arithmetic power, however, higher stake, more likely node will get right to account. As a result, accounting nodes are highly deterministic and rich richer, once with highest equity fails account for block properly. The rest have compete again rights, probability system stagnation increase sharply. To address these two shortcomings, based on dynamic delegation proof importance (DPoI) is proposed.The introduces an...
Summary With lots of public software descriptions emerging in the application market, it is significant to extract common features from these and recommend them new designers. However, existing approaches often according their frequencies which reflect designers' preferences. In order identify those users' favorite help design more popular software, this paper proposes make use data ratings products' downloads preferences extracted features. The proposed approach distinguishes perspective...
Next point-of-interest (POI) recommendation is to recommend the next POI that users may want visit according their check-in trajectories. When there are few check-ins in trajectories, it will cause cold-start problem traditional methods can't capture users' preferences accurately and no longer has ability forecast POI. Based on few-shot meta-learning method of MAML, this paper proposes a new named STMeta for users. Following learning, main idea proposed train model based long trajectories...
Abstract Cryptographic algorithm identification is the process of distinguishing or identifying cryptographic by analysing potential information various features in ciphertext under condition known ciphertext, which basis cryptanalysis work. To solve worse accuracy and stability problem single-layer scheme work as complexity enhances, interference between data number encryption algorithms increases, we propose a block cipher using deep learning Multi-Layer Perception (MLP) this paper. In...
In recommender systems, user reviews on items contain rich semantic information, which can express users' preferences and item features. However, existing review-based recommendation methods either use the static word vector model or cannot effectively extract long sequence features in reviews, resulting limited ability of feature expression. Furthermore, impact different useless interactions between users performance is ignored. Therefore, we propose an attentional factorization machine...
As the core of location-based social networks (LBSNs), main task next point-of-interest (POI) recommendation is to predict possible POI through context information from users’ historical check-in trajectories. It well known that spatial–temporal contextual plays an important role in analyzing users behaviors. Moreover, between POIs provides a non-trivial correlation for modeling visiting preferences. Unfortunately, impact such and spatio–temporal unequal interval on user selection POI,...