Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks570 from public data
- Graph Representation Learning via Graphical Mutual Information Maximization706
An unsupervised learning model trained by maximizing GMI between the input and output of a graph neural encoder is developed, which outperforms state-of-the-art unsuper supervised counterparts, and even sometimes exceeds the performance of supervised ones.
- 462
- 329
- ANOMALOUS: A Joint Modeling Approach for Anomaly Detection on Attributed Networks264
This paper proposes a novel joint framework to conduct attribute selection and anomaly detection as a whole based on CUR decomposition and residual analysis and results on both synthetic and real-world datasets corroborate the effectiveness of the proposed framework.
- 233
- 220
- 211
- Cross-person activity recognition using reduced kernel extreme learning machine173
Experimental results show that, the proposed TransRKELM model can adapt the classifier to new device users quickly and obtain good recognition performance.
- 172
- Virtual machine consolidated placement based on multi-objective biogeography-based optimization154
This work is the first approach that applies biogeography-based optimization (BBO) to virtual machine placement and it is shown that VMPMBBO has better convergence characteristics and is more computationally efficient as well as robust.
- An optimized approach for storing and accessing small files on cloud storage151
Experimental results demonstrate that the proposed schemes effectively improve the storage and access efficiencies of small files, compared with native HDFS and a Hadoop file archiving facility.
- 147
- 143
- 131
- BlueSky Cloud Framework: An E-Learning Framework Embracing Cloud Computing129
In this paper, an E-Learning framework based on cloud computing is presented, namely BlueSky cloud framework, which solves the challenges faced by E-learning, and improves the performance, availability and scalability of E- learning systems.
- A multi-constraint learning path recommendation algorithm based on knowledge map127
A new multi-constraint learning path recommendation algorithm based on knowledge map is proposed, in which the variables and their weighted coefficients considers different learning path preferences of the learners in different learning scenarios as well as learning resource organization and fragmented time.
- Self-weighted Robust LDA for Multiclass Classification with Edge Classes123
A novel self-weighted robust LDA with ℓ2,1-norm based pairwise between-class distance criterion, called SWRLDA, is proposed, which can automatically avoid the optimal mean calculation and simultaneously learn adaptive weights for each class pair without setting any additional parameter.
- Semantics-Guided Contrastive Network for Zero-Shot Object Detection114
A novel Semantics-Guided Contrastive Network for ZSD is developed, named ContrastZSD, a detection framework that first brings contrastive learning mechanism into the realm of zero-shot detection and outperforms the previous state-of-the-art on both ZSD and generalized ZSD tasks.
- 114
- Automatic Generation of Medical Imaging Diagnostic Report with Hierarchical Recurrent Neural Network114
- Cross-media analysis and reasoning: advances and directions104
By presenting approaches, advances, and future directions in cross-media analysis and reasoning, the goal is to draw more attention to the state-of-the-art advances in the field, and to provide technical insights by discussing the challenges and research directions in these areas.
- Scalable attack on graph data by injecting vicious nodes103
This paper develops a more scalable framework named Approximate Fast Gradient Sign Method which considers a more practical attack scenario where adversaries can only inject new vicious nodes to the graph while having no control over the original graph.
- Research on Extreme Learning of Neural Networks93
A novel algorithm called Regularized Extreme Learning Machine (RELM) based on structural risk minimization and weighted least square is proposed in this paper, which has better generalization, robustness and controllability than the original ELM.
- 91
- Ordinal extreme learning machine90
The proposed encoding-based framework for ordinal regression which includes three encoding schemes: single multi-output classifier, multiple binary-classifications with one-against-all (OAA) decomposition method and one- against-one (OAO) method, and the SLFN was redesigned for ordinals regression problems based on the proposed framework.
- 88
- 86
- Graph Few-shot Learning with Attribute Matching84
The proposed AMM-GNN leverages an attribute-level attention mechanism to capture the distinct information of each task and thus learns more effective transferable knowledge for meta-learning.
- A Fast Reduced Kernel Extreme Learning Machine80
Experimental results on a wide variety of real world small instance size and large instance size applications in the context of binary classification, multi-class problem and regression are reported to show that RKELM can perform at competitive level of generalized performance as the SVM/LS-SVM at only a fraction of the computational effort incurred.
- A survey of faceted search80
This paper first analyzes the representative facet search models, then presents a general faceted search framework, and surveys the related methods and techniques, including facet term extraction, hierarchy construction, compound term generation and facet ranking.
- BotMoE: Twitter Bot Detection with Community-Aware Mixtures of Modal-Specific Experts79
BotMoE is proposed, a Twitter bot detection framework that jointly utilizes multiple user information modalities (metadata, textual content, network structure) to improve the detection of deceptive bots and incorporates a community-aware Mixture-of-Experts (MoE) layer to improve domain generalization and adapt to different Twitter communities.
- SeqTrans: Automatic Vulnerability Fix Via Sequence to Sequence Learning79
This paper provides a novel approach called SeqTrans to exploit historical vulnerability fixes to provide suggestions and automatically fix the source code, and proposes to leverage data-flow dependencies to construct code sequences and feed them into the state-of-the-art transformer model.
- Simple to complex cross-modal learning to rank79
This paper involves the self-paced learning theory with diversity into the cross-modal learning to rank and learns an optimal multi- modal embedding space based on non-linear mapping functions that enhances the model’s robustness to outliers and achieves better generalization via training the model gradually from easy rankings by diverse queries to more complex ones.
- Knowledge Engineering with Big Data79
BigKE is presented, a knowledge engineering framework that handles fragmented knowledge modeling and online learning from multiple information sources, nonlinear fusion on fragmented knowledge, and automated demand-driven knowledge navigation.
- 78
- Mining query subtopics from search log data78
A clustering algorithm is proposed that can effectively leverage the two phenomena of user behavior to automatically mine the major subtopics of queries, where each subtopic is represented by a cluster containing a number of URLs and keywords.
- 74
- 71
- A new self-supervised task on graphs: Geodesic distance prediction70
This paper introduces a novel self-supervised strategy for graph representation learning by exploiting natural supervision provided by the data itself by randomly selecting pairs of nodes in a graph and training a well-designed neural net to predict the contextual position of one node relative to the other.
- 70
- Noisy Correspondence Learning with Meta Similarity Correction69
A Meta Similarity Correction Network (MSCN) is proposed to provide reliable similarity scores and an effective data purification strategy using meta-data as prior knowledge to remove the noisy samples.
- 69
- 68
- 67
- Semantics-Preserving Graph Propagation for Zero-Shot Object Detection66
This paper proposes a novel Semantics-Preserving Graph Propagation model for ZSD based on Graph Convolutional Networks (GCN), and develops a graph construction module to flexibly build category graphs by leveraging diverse correlations between category nodes, followed by two semantics-preserving graph propagation modules that enhance both category and region representations.
- 66
- 65
- 63
- 60
- KRACL: Contrastive Learning with Graph Context Modeling for Sparse Knowledge Graph Completion59
This work proposes a novel framework KRACL1 to alleviate the widespread sparsity in KGs with graph context and contrastive learning, and proposes the Knowledge Relational Attention Network (KRAT) to leverage the graph context by simultaneously projecting neighboring triples to different latent spaces and jointly aggregating messages with the attention mechanism.
- 58
- 57
- Clustering based Point Cloud Representation Learning for 3D Analysis57
This work proposes a clustering based supervised learning scheme for point cloud analysis that conducts within-class clustering on the point embedding space for automatically discovering subclass patterns which are latent yet representative across scenes.
- Multi-objective optimization for rebalancing virtual machine placement57
A novel solution to the VMrB problem is proposed, namely a Pareto-based Multi-Objective VM reBalance solution (MOVMrB), which aims to simultaneously minimize the disequilibrium of both inter-HM and intra-HM loads.
- Semi-supervised clue fusion for spammer detection in Sina Weibo57
A novel approach called Semi-Supervised Clue Fusion (SSCF) is proposed to conduct effective spammer detection in Sina Weibo and shows that this approach significantly outperforms state-of-the-art baselines.
- Identifying suspicious groups of affiliated-transaction-based tax evasion in big data56
Experimental tests that are based on two years of real-world tax data from a province in China demonstrate that 3TI can identify ATTE suspicious groups with higher accuracy and better generality than existing works.
- Exploring community structure of software Call Graph and its applications in class cohesion measurement55
It is shown that networks formed by software methods and their calls exhibit relatively significant community structures and two new class cohesion metrics are proposed to measure the cohesiveness of object-oriented programs, able to predict software faults more effectively than existing metrics.
- Dynamic author name disambiguation for growing digital libraries52
A novel IncAD model is proposed which aggregates metadata from a cluster of records to estimate the author’s profile such as her coauthor distributions and keyword distributions, in order to predict how likely it is that a new record is “produced” by the author.
- 50
- 50
- A new approach for parallel simulation of VLSI circuits on a transistor level50
This paper presents a new approach for parallel simulation of very large scale integration (VLSI) circuits on a transistor level with an efficient partitioning technique, which yields well-balanced partitions with few interconnects and speedup results for parallel analog simulation of large industrial designs.
- 48
- Relation-based test case prioritization for regression testing47
Results show that AGC outperforms existing techniques on large programs in terms of bug detection capability, and also achieves the highest mean APFD value, and the performance demonstrates a growth trend as the size of the program increases.
- Efficient Deep Web Crawling Using Reinforcement Learning47
A novel deep web crawling framework based on reinforcement learning, in which the crawler is regarded as an agent and deep web database as the environment is proposed, which outperforms the state of art methods in terms of crawling capability and breaks through the assumption of full-text search implied by existing methods.
- A Genetic Algorithm-Based Energy-Efficient Container Placement Strategy in CaaS46
It is demonstrated that an energy-saving gain can be achieved by optimizing the placement of containers under a nonlinear energy consumption model and an improved genetic algorithm called IGA is proposed for efficiently searching the optimal CP solution.
- A Fast SVD-Hidden-nodes based Extreme Learning Machine for Large-Scale Data Analytics46
This paper proposes an Extreme Learning Machine (ELM) approach for large-scale data analytic and embeds hidden nodes that are designed using singular value decomposition (SVD) into the classical ELM, demonstrating the superior generalization performance and efficiency of the FSVD-H-ELM.
- 45
- 45
- 43
- A financial ticket image intelligent recognition system based on deep learning42
An iterative self-learning framework of Financial Ticket Intelligent Recognition System (FTIRS) is proposed, which supports iteratively updating and extensibility of the algorithm model, which are the fundamental requirements for a practical financial accounting system.
- Learning to crawl deep web42
Experimental results show that the novel deep web crawling framework based on reinforcement learning outperforms the state of art methods in terms of crawling capability and relaxes the assumption of full-text search implied by existing methods.
- 41
- Learning to Rematch Mismatched Pairs for Robust Cross-Modal Retrieval40
The potential semantic similarity among unpaired samples makes it possible to excavate useful knowledge from mismatched pairs and the proposed L2RM, a general framework based on Optimal Transport that learns to rematch mismatched pairs is proposed.
- 40
- A Boundary Regression Model for Nested Named Entity Recognition40
This model is a multiobjective learning framework that simultaneously predicts the classification score of an NE candidate and refines its spatial location in a sentence and enables more potent nonlinear function approximators to enhance model discriminability.
- 40
- LMBot: Distilling Graph Knowledge into Language Model for Graph-less Deployment in Twitter Bot Detection39
A novel bot detection framework LMBot is proposed that distills the graph knowledge into language models (LMs) for graph-less deployment in Twitter bot detection to combat data dependency challenge and is compatible with graph-based and graph-less datasets.
- CLDG: Contrastive Learning on Dynamic Graphs39
A sampling layer to extract the temporally-persistent signals is proposed and the number of model parameters and training time is reduced by an average of 2,001.86 times and 130.31 times on seven datasets, respectively.
- Diverse fuzzy c-means for image clustering39
This work proposes a diverse fuzzy c-means for image clustering via introducing a novel diversity regularization into the traditional fuzzy c -means objective, which guarantees the learned image cluster centers to be different from each other and to fill the image data space as much as possible.
- Combining machine learning and human judgment in author disambiguation39
Comparing to conventional author disambiguation algorithms, the LOAD yields much more accurate results to assist human labeling and can save labeling time dramatically.
- 38
- 38
- A neuralized feature engineering method for entity relation extraction37
Experiments show that this approach considerably improves the performance compared to that of neural networks or feature-based models alone, exceeding state-of-the-art performance by more than 8% and 16.5% in terms of F1-score on the ACE corpus and the Chinese literature text corpus, respectively.
- 37
- Probabilistic Non-Negative Matrix Factorization and Its Robust Extensions for Topic Modeling37
This paper follows the generative procedure of topic model and learns the topic-word distribution and topics distribution via directly approximating the word-document co-occurrence matrix with matrix decomposition technique and adopts the explicit probabilistic meaning of factors in topic models.
- 37
- Interpretation-Enabled Software Reuse Detection Based on a Multi-level Birthmark Model36
ISRD is an interpretation-enabled software reuse detection approach based on a multi-level birthmark model that contains function level, basic block level, and instruction level that is interpretable, effective, and efficient, and resilient to cross-compilation, outperforming state-of-the-art approaches.
- 36
- 36
- Identifying at-risk students based on the phased prediction model36
A phased prediction model to predict at-risk students at different stages of a semester is proposed based on students’ individual characteristics and online learning behaviors, and combined feature sets based on a time window constraint strategy and a learning time threshold constraint strategy are proposed.
- Recognizing Multidimensional Engagement of E-Learners Based on Multi-Channel Data in E-Learning Environment36
An integrated framework to identify learning engagement from three facets: affect, behavior and cognitive state, which are conveyed by learner’s facial expressions, eye movement behaviors and the overall performance during short video learning session is proposed.
- Scheduled sampling for one-shot learning via matching network36
A novel metric is proposed to evaluate the learning difficulty of support set, where this metric jointly considers the semantic diversity and similarity of visual labels and introduces a scheduled sampling strategy to train the matching network from easy to difficult.
- 35
- 35
- A prediction model of student performance based on self-attention mechanism35
A performance prediction model based on student behavior feature mode difference is proposed to improve the model's prediction accuracy and increases the model’s robustness for students with significant changes in performance.
- 35
- 34
- Distributed extreme learning machine with alternating direction method of multiplier34
A novel distributed extreme learning machine is exploited to implement theextreme learning machine algorithm in parallel for large-scale data set and a corresponding distributed algorithm is developed on the basis of alternating direction method of multipliers which shows effectiveness in distributed convex optimization.
- Reduced Kernel Extreme Learning Machine34
The experimental results show that Reduced-KELM can perform at a similar level of accuracy as K ELM and at the same time being significantly faster than KELM.
- 33
- TaxThemis: Interactive Mining and Exploration of Suspicious Tax Evasion Groups33
TaxThemis is an interactive visual analytics system to help tax officers mine and explore suspicious tax evasion groups through analyzing heterogeneous tax-related data and demonstrates the usefulness and effectiveness of TaxThemis through two case studies on real-world tax- related data and interviews with domain experts.
- Analyzing and modeling dynamics of information diffusion in microblogging social network33
A diffusion model based on cascade model framework is proposed to generate the retweeting network and results show that this model could reproduce the diffusion features of the retweeted network effectively and outperforms the most widely used independent cascade model.
- Online sequential reduced kernel extreme learning machine33
Experimental results obtained indicate that the proposed OS-RKELM showcases improved prediction accuracy and efficiency over the OS-ELM, LASVM and BSGD in many cases.
- Mining subtopics from text fragments for a web query33
This paper proposes to use surrounding text of query terms in top retrieved documents to mine subtopics and rank them, and generates a readable subtopic for each cluster based on the cluster and the language model trained from a query log.
- 32
- 32
- 31
- 31
- Relationships between interval-valued vector optimization problems and vector variational inequalities31
The weakly efficient points of the interval-valued vector optimization problems and the solutions of the weak vector variational inequalities under smooth and non-smooth LU-convexity assumptions are identified.
- 30
- 30
- A reversible sketch-based method for detecting and mitigating amplification attacks30
This work uses a Chinese Reminder Theorem based Reversible Sketch to directly collect network traffic and then monitors the abrupt changes in one-to-one mapping between request packets and response packets to identify amplification attack traffic, which enables the detection method to handle big-volume network traffic.
- 30
- 30
- 30
- Modeling and Predicting the Active Video-Viewing Time in a Large-Scale E-Learning System30
This paper collects video-viewing data from a large-scale e-learning system and uses the Cox proportional hazard function to model the leaving time, and models the extent to which factors influence it and further predicts when a user permanently leaves a course.
- 30
- Towards Entity Alignment in the Open World: An Unsupervised Approach29
This work develops an unsupervised framework that performs entity alignment in the open world that does not require labeled data and can effectively filter out unmatchable entities.
- How Learner Support Services Affect Student Engagement in Online Learning Environments29
Investigating students’ usage of two online learner support (OLS) services and their impact on student engagement indicates that the use of services and how many times the services are used are highly relevant to the pattern and level of student engagement, which suggests that theUse of OLS services can be added to the indicators that reflect student learning status.
- 29
- Feature assembly method for extracting relations in Chinese29
This study proposes a feature assembly method for relation extraction that combines six types of candidate features employed as atomic features and six constraint conditions are used to combine these features in different settings.
- 29
- A Cross-Curriculum Video Recommendation Algorithm Based on a Video-Associated Knowledge Map28
A two-stage cross-curriculum video recommendation algorithm that considers both the learners’ implicit feedback and the knowledge association between course videos, and performs better than the traditional collaborative filtering-based recommendation algorithms in terms of accuracy, recall rate, and knowledge relevance.
- Noise-Tolerant Learning for Audio-Visual Action Recognition27
A noise-tolerant learning framework that significantly improves the robustness of the action recognition model and surpasses the baselines by a clear margin and establishes a benchmark of real-world noisy correspondence in audio-visual data by relabeling the Kinetics dataset.
- Cross-modal credibility modelling for EEG-based multimodal emotion recognition27
A novel multimodal physiological signal fusion model that incorporates both intra-inter modality reconstruction and sequential pattern consistency is introduced, thereby ensuring a computable and credible EEG-based multimodal emotion recognition and indicates that credibility modelling is essential for multimodal emotion recognition.
- 27
- Representation learning over multiple knowledge graphs for knowledge graphs alignment27
This paper proposes a new automatic representation learning model over Multiple Knowledge Graphs (MGTransE) by adopting a bootstrapping method and shows that the method achieves better performance on two new multiple KGs tasks compared with state-of-the-art KG embedding models and preserves the key properties of knowledge graph embedding on traditional single KG tasks as compared to those methods learned fromsingle KG.
- Top-k multi-class SVM using multiple features27
To fuse multiple features effectively, an adaptive weight for each view is introduced and an efficient alternating optimization algorithm is exploited to learn the optimal classifiers and their corresponding weights jointly.
- A topic sentence-based instance transfer method for imbalanced sentiment classification of Chinese product reviews27
Experimental results verify that the proposed topic sentence-based instance transfer method helps a support vector machine (SVM) to outperform other classification methods, such as the J48, Naive Bayes, Random Forest and Random Committee methods, when applied to datasets produced by resampling and Smote.
- 27
- Tile Classification Based Viewport Prediction with Multi-modal Fusion Transformer26
A tile classification based viewport prediction method with Multi-modal Fusion Transformer, namely MFTR shows superior performance over state-of-the-art methods in terms of average prediction accuracy and overlap ratio, and also presents competitive computation efficiency.
- 26
- Mining learning-dependency between knowledge units from text26
This paper proposes a method of mining the learning- dependencies among the KU from text document based on the distributional asymmetry of the domain terms and the local nature of thelearning-dependency.
- Data Quality-Aware Mixed-Precision Quantization via Hybrid Reinforcement Learning25
This article proposes a novel data quality-aware mixed-precision quantization framework, dubbed DQMQ, to dynamically adapt quantization bit-widths to different data qualities, based on a bit-width decision policy that can be learned jointly with the quantization training.
- 25
- An overlay multicast protocol for live streaming and delay-guaranteed interactive media25
Simulation and experimental results show that the proposed overlay multicast protocol can simultaneously provide delay-guaranteed interactive media as well as scalable live streaming.
- Knowledge Element Extraction for Knowledge-Based Learning Resources Organization25
A knowledge element taxonomy containing 25 semantic types is built and formalized the knowledge element extraction of single semantic type as binary classification as well as constructing the multi-class classification model which can predict the semantic type of knowledge element by merge the results of binary classifiers.
- The Research of Mining Association Rules Between Personality and Behavior of Learner Under Web-Based Learning Environment25
An improved algorithm, which is based on Apriori algorithm widely used in market basket analysis, is designed and implemented, and result indicated that this algorithm for mining association rules between behavior and personality is feasible and efficient.
- 24
- 24
- 24
- Weakly Supervised Learning for Textbook Question Answering23
A Weakly Supervised learning method for TQA (WSTQ), which regards the incompletely accurate results of essential intermediate procedures for this task as supervision to develop Text Matching (TM) and Relation Detection (RD) tasks and then employs the tasks to motivate itself to learn strong text comprehension and excellent diagram semantics respectively.
- Dual-stream generative adversarial networks for distributionally robust zero-shot learning23
A dual-stream generator to synthesize visual samples and reconstruct semantic embedding simultaneously, thereby ensuring semantic consistency and outperforms the state-of-the-arts by 4.7% and 3.0% on average in two metrics over four real-world datasets.
- Topological analysis of knowledge maps23
It is discovered that knowledge maps exhibit small-world and scale-free properties like many other networks and how KUs affect the network efficiency by removing KUs based on different centrality measures varies greatly.
- The Evidence Contraction Issue in Deep Evidential Regression: Discussion and Solution22
This paper designs a Non-saturating Uncertainty Regularization term, which effectively ensures that the performance of DER is further optimized in the right direction, and shows that this proposed approach improves the performance of DER while maintaining the ability to quantify uncertainty.
- AHNG: Representation learning on attributed heterogeneous network22
A unified embedding model which represents each node in an AHN with a G aussian distribution (AHNG), which fuses multi-type nodes/links and diverse attributes through a two-layer neural network and captures the uncertainty by embedding nodes as Gaussian distributions.
- 22
- 22
- 21
- Disentangling interest and conformity for eliminating popularity bias in session-based recommendation21
This work proposes an SBR framework with Disentangling InteRest and Conformity for eliminating popularity bias in SBR, and shows that this framework outperforms other state-of-the-art debias methods consistently.
- HOFA: Twitter Bot Detection with Homophily-Oriented Augmentation and Frequency Adaptive Attention21
HBOFA is proposed, a novel graph-based Twitter bot detection framework that combats the heterophilous disguise challenge with a homophily-oriented graph augmentation module (Homo-Aug) and a frequency adaptive attention module (FaAt).
- PRIOR: deep reinforced adaptive video streaming with attention-based throughput prediction21
This work proposes PRIOR, an ABR algorithm that fuses an effective throughput prediction module and a state-of-the-art multi-agent reinforcement learning method to provide a high quality of experience (QoE).
- 21
- Adaptive Semi-Supervised Learning with Discriminative Least Squares Regression21
This paper develops a new discriminative LSR by equipping each label with an adjustment vector and distinguishes the label fitting of labeled and unlabeled training data through a probabilistic vector with an adaptive parameter, which always ensures the significant importance of labeled data and characterizes the contribution of unlabeling instance according to its uncertainty.
- Avoiding Optimal Mean ℓ2,1-Norm Maximization-Based Robust PCA for Reconstruction21
This letter equivalently reformulate the objective of conventional PCA and learn the optimal projection directions by maximizing the sum of projected difference between each pair of instances based on -norm by exploiting an efficient optimization algorithm to soften the contributions from outliers by reweighting each data point iteratively.
- Omni-word Feature and Soft Constraint for Chinese Relation Extraction21
An Omniword feature and a soft constraint method for Chinese relation extraction that makes a better use of sentence information and minimize the influences caused by Chinese word segmentation and parsing are proposed.
- 20
- 20
- 20
- Implementation of National Health Informatization in China: Survey About the Status Quo20
The nationwide survey shows that China’s health informatization is rapidly developing and current information platforms and databases technically support data exchanging between all provinces and cities.
- 20
- A Study of Learner-Oriented Negative Emotion Compensation in E-learning20
The results show that the proposed learner-oriented negative emotion compensation provides greater satisfaction for the e- learner, and is a feasible and effective method for e-learner to decrease negative emotions in e-learning.
- Understand, Refine and Summarize: Multi-View Knowledge Progressive Enhancement Learning for Fake News Video Detection19
This work introduces EvidSV, the first comprehensive benchmark supporting evidence- and attribution-aware fake news video detection, and proposes MUKE, a multi-view knowledge progressive enhancement learning framework that outperforms existing methods in both fake news detection and attribution, and generalizes effectively to previously unseen domains.
- 19
- 19
- XTQA: Span-Level Explanations for Textbook Question Answering19
In this article, spans are the combinations of sentences within a paragraph and the proposed coarse-to-fine grained explanation extracting (EE) algorithm is used to narrow down the evidence scope and extract the span-level explanations with varying lengths for answering different questions.
- 19
- 19
- 19
- 19
- A Group-Oriented Recommendation Algorithm Based on Similarities of Personal Learning Generative Networks19
The experimental results show that the group-oriented learning recommendation based on the learning generated network outperforms the group recommendation-based collaborative filtering when the amount of data is large enough.
- Detecting global and local topics via mining twitter data19
A probabilistic graphical model is proposed to extract global and local topics related to social events in a unified framework at the same time by associating each tweet with both of them simultaneously.
- Performance models and dynamic characteristics analysis for HDFS write and read operations: A systematic view19
The relationship between file size and HDFS Write/Read throughput is studied and a system identification-based approach is applied to construct performance models for HDFS W/R operations under different conditions, which can provide effective guidance and implications for the design and configuration of HDFS and Hadoop-based applications.
- Plagiarism detection for multithreaded software based on thread-aware software birthmarks19
Two dynamic birthmark based approaches are presented that can effectively detect plagiarism of multithread programs and exhibit strong resilience to various semantic-preserving code obfuscations.
- Dynamic query intent mining from a search log stream19
This work proposes a method for mining dynamic query intents from search query logs by regarding the query logs as a data stream, and shows that this method can accurately capture new intents with short response time.
- CRAWLING DEEP WEB CONTENT THROUGH QUERY FORMS19
This paper proposes the concept of Minimum Executable Pattern (MEP), and then presents a MEP generation method and a MEP-based Deep Web adaptive query method that outperforms existing methods in terms of query capability and applicability.
- 19
- 18
- 18
- 18
- 18
- 18
- Robust dictionary learning with graph regularization for unsupervised person re-identification18
A novel unsupervised method, termed as Robust Dictionary Learning with Graph Regularization (RDLGR), which can guarantee view-invariance through learning a dictionary shared by all the camera views, and introduces a cross-view graph Laplacian regularization term into the framework of dictionary learning.
- A behavioral sequence analyzing framework for grouping students in an e-learning system18
This paper aims to build a unified framework for grouping students based on the behavioral sequences and further predicting which group a newcomer will be, and proposes a shape descriptor to approximately express the geometrical information of trajectories.
- Calling Network18
The Calling Network (CN) model is proposed to provide new perspectives to analyze the dynamic execution process of a software system and presents several interesting features, such as Power-law degree distribution, Densification Power Law, and the stability of an entropy value -- Local Entropy.
- AutoEER: automatic EEG-based emotion recognition with neural architecture search17
This work proposes AutoEER (Automatic EEG-based Emotion Recognition), a framework that leverages tailored NAS to automatically discover the optimal network structure for EEG- based emotion recognition, and carefully design a customized search space specifically for EEG signals.
- 17
- A Boundary Assembling Method for Nested Biomedical Named Entity Recognition17
This paper focuses on identifying nested biomedical named entities using a boundary assembly (BA) model, which is a cascading framework consisting of three steps which is effective in handling nesting and discontinuous problems in biomedical named entity recognition tasks.
- Exploiting thread-related system calls for plagiarism detection of multithreaded programs17
This paper redesigns birthmark based software plagiarism detection algorithms to make such approach effective for multithreaded programs and shows that the new birthmarks are superior to existing birthmarks and are resilient against most state of theart obfuscation techniques.
- 17
- 16
- 16
- 16
- Mining temporal characteristics of behaviors from interval events in e-learning16
This paper discusses a method that mines big learning data by taking both the frequency and duration into account, and defines a function for evaluating the importance of events, summarizing them into big uniform events according to the semantics, and further segmenting the BUE s using a sliding window to avoid the counting bias issue.
- Deep Semisupervised Zero-Shot Learning with Maximum Mean Discrepancy16
This work proposes a novel deep semisupervised method that replaces the original labels with the corresponding textual descriptions to better capture the category semantics and overcomes the problem of distribution difference by minimizing the maximum mean discrepancy between seen and unseen instance distributions.
- 16
- 16
- 16
- A Service-oriented Approach to Integration of e-learning Information and Resource Management Systems16
- 16
- Multiple behavior information fusion based quantitative threat evaluation16
Testing in a real network environment shows that this novel model of security threat evaluation with a series of quantitative indices is reasonable and feasible in alleviating the tremendous task of data analysis and facilitating the understanding of the security evolution of the system for its administrators.
- 15
- A Feature Combination-Based Graph Convolutional Neural Network Model for Relation Extraction15
A feature combination-based graph convolutional neural network model (FC-GCN) is proposed that has the advantages of encoding structural information of a sentence, considering prior knowledge, and avoiding errors caused by parsing.
- Protein-protein interaction relation extraction based on multigranularity semantic fusion15
A multigranular semantic fusion method to support biomedical relation extraction using Transformer for embedding words of a sentence into distributed representations, which is effective to encode global semantic of a sentences.
- Current and future of technologies and services in smart e-learning15
Smart e-learning is still in the intelligence stage, showing intelligent, ubiquitous, personalized, open, and collaborative trends and demanding for a close integration of artificial intelligence and big data.
- 15
- A Controlled Attention for Nested Named Entity Recognition14
The controlled attention mechanism for recognizing named entities is proposed, where task-related cues are implanted into each sentence to indicate boundaries of possible named entities and achieves the state-of-the-art performance in all evaluation datasets.
- Toward Entity Alignment in the Open World: An Unsupervised Approach with Confidence Modeling14
An unsupervised framework UEA is offered that performs entity alignment in the open world and the concept of confidence is introduced to measure the probability of an entity pair of being true and developed a confidence-based un supervised EA framework CUEA.
- An Adaptive Security Data Collection and Composition Recognition method for security measurement over LTE/LTE-A networks14
An Adaptive Security Data Collection and Composition Recognition (ASDCCR) method is proposed to make the data collection adaptive and synthetic attack recognition become possible and verified in simulated LTE environment of NS3 to verify the usability and accuracy of the proposed methods.
- Sparse Multi-Modal Topical Coding for Image Annotation14
This paper proposes a non-probabilistic formulation of PTM for automatic image annotation, namely Sparse Multi-Modal Topical Coding, and introduces sparsity to PTM to improve the interpretability of the inferred latent representations.
- Exploring Flow Characteristics in IPv6: A Comparative Measurement Study with IPv4 for Traffic Monitoring14
There are less sharp changes caused by abnormity compared with IPv4, which shows there are less large-scale malicious activities in IPv6 currently, and a group flow model is proposed which can greatly reduce the number of flows while capturing the primary traffic features.
- Adaptive personalized recommendation based on adaptive learning14
Empirical evaluation of the proposed novel adaptive personalized recommendation demonstrates that it is possible to obtain accuracy comparable to that of the correlation-based, SVD-based and supervised learning-based approaches at a much lower computational cost.
- 14
- 13
- Deep evidential learning in diffusion convolutional recurrent neural network13
Empirical deep learning (EDL) is applied with diffusion convolutional recurrent neural network (DCRNN), and the experiment in spatiotemporal forecasting task in a real-world traffic dataset is done, and mean interval scores (MIS) is chosen as a good metric for uncertainty quantification.
- 13
- A History and Theory of Textual Event Detection and Recognition13
The hallmark of this article is that it is the first survey that combines IR, IE and GMs in a single frame and will therefore benefit researchers by acting as a reference in this field.
- 13
- 13
- 13
- 13
- DF-Miner: Domain-specific facet mining by leveraging the hyperlink structure of Wikipedia13
This paper presents an automatic approach, called domain-specific facet (DF)-Miner, to discover DFs based on the hyperlink structure within the Wikipedia article pages, and shows that DF-Miner performs better than the textual content-based approaches.
- 13
- 13
- 12
- 12
- 12
- B 2 -Sampling: Fusing Balanced and Biased Sampling for Graph Contrastive Learning12
B2-Sampling is designed as a framework to support many known GCL models and uses balanced sampling to select the most representative negative pairs regarding both the topological and embedding diversities and biased sampling to learn and correct the labels of the most error-prone negative pairs during the training.
- 12
- Knowledge forest: a novel model to organize knowledge fragments12
A novel knowledge organization model, knowledge forest, which consists of facet trees and learning dependencies is proposed, which can effectively organize knowledge fragments, and alleviate information overload and learning disorientation.
- AvatarMeeting: An Augmented Reality Remote Interaction System With Personalized Avatars12
This work presents an avatar based remote interaction system AvatarMeeting, enabling users to meet with remote peers through interactive personalized avatars just like face to face, and proposes a novel framework including a consumer-grade set-up, a complete transmission scheme and a processing pipeline.
- Dynamic structure measurement for distributed software12
The intra-component and inter-component dependencies are introduced into a Calling Network model to further represent distributed software and based on the Kieker monitoring framework, the measurement methods are proposed and implemented.
- An efficient online active learning algorithm for binary classification12
This paper proposes a new online active learning algorithm for binary classification that uses the margin-based criterion, which compares the margin of instances with a threshold to decide whether it should be queried, and proposes Iteratively Decreased Threshold (IDT), a new threshold update method for the margin -based criterion.
- 12
- 12
- 12
- 12
- An OVAL-based active vulnerability assessment system for enterprise computer networks12
The testing results from a series of experiments show that the NetScope system is well suited for vulnerability assessment of large-scale computer networks such as campus networks and enterprise networks and can also be easily integrated with other security tools based on relational databases.
- 12
- 11
- 11
- Towards Fairer and More Efficient Federated Learning via Multidimensional Personalized Edge Models11
A Customized Federated Learning system to eliminate FL heterogeneity from multiple dimensions is proposed, which tailors personalized models from the specially designed global model for each client jointly guided by an online trained model-search helper and a novel aggregation algorithm.
- Domain Knowledge Guided Deep Atrial Fibrillation Classification and Its Visual Interpretation11
A domain knowledge guided deep neural network is proposed that uses a deep residual network as a classification framework, within which key feature ~(P-wave and R-peak position) reconstruction tasks are adopted to incorporate domain knowledge in the learning process.
- 11
- A priority-based adaptive scheme for multi-view live streaming over HTTP11
This paper adopts an integrated bandwidth prediction approach to calculate the available bandwidth, solving the low bandwidth utilization problem existing in the benchmark method, and designs a unified segment request strategy to restrict the interval between each request, and guarantees the synchronization of different live video streams.
- 11
- An Emotion-oriented Music Recommendation Algorithm Fusing Rating and Trust11
Experimental results show that the presented algorithm comparing the collaborative filtering and trust-based recommendation, results in a tiny loss of accuracy with the improvement of larger coverage, thus not only obtaining a perfect tradeoff between accuracy and coverage, but also increasing the degree of listener satisfaction.
- 11
- 10
- Empower Post-hoc Graph Explanations with Information Bottleneck: A Pre-training and Fine-tuning Perspective10
A two-stage explanation strategy is considered that endows post-hoc graph explanations with the applicability to pretrained GNNs where downstream tasks are inaccessible and the capacity to explain the transferability of task-agnostic explanations.
- NerCo: A Contrastive Learning Based Two-Stage Chinese NER Method10
NerCo is presented: Named entity recognition with Contrastive learning, a novel NER framework which can better exploit labeled data and avoid the above problem.
- A Set Space Model to Capture Structural Information of a Sentence10
This paper describes a systematic and novel approach to manipulating features of a sentence using the set space model to resolve the feature sparsity problem and shows the influence of sentence structural information on information extraction.
- DFT-extractor10
This study demonstrates a novel system called DFT-Extractor, which automatically constructs domain-specific faceted taxonomies from Wikipedia in three steps: it crawls domain terms from Wikipedia by using a modified topical crawler, and exploits a classification model to extract hyponym relations with the use of motif-based features.
- 10
- Offline Definition Extraction Using Machine Learning for Knowledge-Oriented Question Answering10
This paper introduced a new framework of knowledge-based question answering (QA) system, which employs three classification models: Decision tree, Naive Bayesian and SVM, and shows that SVM has best performance in definition extraction.
- 9
- 9
- CGUA: Context-Guided and Unpaired-Assisted Weakly Supervised Person Search9
A novel C ontext-G uided and CGC algorithm to leverage context information in the clustering process and an U npaired- A ssisted M emory ( UAM) unit to distinguish unpaired and paired persons by pushing them away are proposed.
- 9
- User Sequential Behavior Classification for Click-Through Rate Prediction9
The concept of user browsing status is introduced, and a new framework named User Status Recognition Framework (USRF) is proposed, which can perform both CTR prediction and next-item recommendation and implements a simple model for CTR prediction based on USRF.
- IEEE Access Special Section Editorial: Trusted Computing9
The core of the trusted computing technologies of TCG is the Trusted Platform Module (TPM), a tamper-resistant module embedded in a platform, which is implemented by a TPM chip and designed to resist software attacks and moderate hardware attacks.
- 9
- 9
- Progressive generative adversarial networks with reliable sample identification9
It is shown by inspection that the reliable samples with smaller errors are beneficial to achieve a better generator, while the unreliable one might disturb the training procedure, and an indicator for each sample to indicate its reliability is introduced in this paper.
- 9
- A set of domain rules and a deep network for protein coreference resolution9
A rule-based and a machine learning-based method, which takes use of a recurrent neural network model, a long-short term memory network, and automatically learns global discriminative representations of all kinds of coreferences without hand-crafted features are presented.
- Opinioned Post Detection in Sina Weibo9
A weakly supervised framework named graph-based opinioned post detector (GOPD) to detect the opinioned posts in Sina Weibo and transfers cross topics is proposed.
- Deep Web adaptive crawling based on minimum executable pattern9
Experimental results show that the proposed Minimum Executable Pattern (MEP) method outperforms the state of art methods in terms of query capability and applicability, and on average, it achieves good coverage by issuing only a few hundred queries.
- Learning to Surface Deep Web Content9
Experimental results show that the proposed novel deep web crawling framework outperforms the state of art methods in crawling capability and breaks through the assumption of full-text search implied by existing methods.
- 9
- 9
- Transcriptomics and Plant Hormone Analysis Reveal the Mechanism of Branching Angle Formation in Tea Plants (Camellia sinensis)8
Analysis of transcriptome and plant hormone levels of tea plant cultivars with different branching angles offered insights into the mechanism of branching angle formation in tea plants, providing important theoretical reference for the selection and breeding of tea cultivars suitable for mechanical harvesting, thereby improving agricultural efficiency and sustainability.
- APRE: Annotation-Aware Prompt-Tuning for Relation Extraction8
This work proposes an annotation-aware prompt-tuning model for relation extraction that achieves the state-of-the-art performance on two public datasets and demonstrates the effectiveness of the model to decode semantic information in PLMs.
- 8
- Full-span named entity recognition with boundary regression8
A boundary regression model is proposed to support full-span named entity recognition, where a regression operation is adopted to refine spatial locations of entity spans in a sentence to find all possible named entities in a sentence.
- Towards characterizing bug fixes through dependency-level changes in Apache Java open source projects8
This study conducts a systematic and comprehensive study on bug fixes collected from 157 Apache open source projects, and unveils useful findings based on qualitative and quantitative analysis and also provides new insights that might benefit existing bug prediction techniques.
- 8
- A New Representation Learning Method for Individual Treatment Effect Estimation: Split Covariate Representation Network.8
A simple method to split covariates, and then a network, Split Covariate Representation Network (SCRNet), is mentioned, is mentioned, which is used to estimate ITE by splitting kinds of variables, which outperforms other state-of-arts methods on IHDP, a semi-synthetic dataset, and Jobs, a real-world dataset.
- Nonlinear feature selection on attributed networks8
This paper proposes a novel semi-supervised nonlinear framework FS-GCN based on graph convolutional networks (GCNs) to select high-quality features, which can elaborately catch the nonlinear dependency between nodal attributes and class labels and explores the possibility of totally removing the label information.
- VUC: Visualizing Daily Video Utilization to Promote Student Engagement in Online Distance Education8
This paper introduces a visualization tool called video utilization calendar (VUC), designed to visualize both the current viewing progress and the daily viewing history for all the courses in a semester for students to check their viewing progress for all videos and choose any course video to view directly.
- A Secure Remote Monitoring Framework Supporting Efficient Fine-Grained Access Control and Data Processing in IoT8
The proposed framework can protect user data privacy against unauthorized parties, with minimum performance cost compared to existing software-based solutions with such strong privacy protection, and is implemented using latest hardware-based trustworthy computing technology, such as Intel SGX.
- 8
- A Rough Set Based Approach to Find Learners' Key Personality Attributes in an E-Learning Environment8
An improved approach based on rough set theory to find the key personality attributes and evaluates the importance of these attributes is proposed and successfully used in the actual e-learning environment for a major research university in China.
- 7
- 7
- 7
- All-content text recognition method for financial ticket images7
A financial ticket character recognition framework (FTCRF) is proposed, which can improve the accuracy of multi-character mixed character recognition and make the detection and recognition of financial ticket surface information more efficient.
- Research on fast text recognition method for financial ticket image7
A simple yet efficient network named the Financial Ticket Faster Detection network (FTFDNet) based on a Faster R-CNN is proposed, and a comparison with the best ticket recognition model from the ICDAR2019 invoice competition is performed.
- Cue prompt adapting model for relation extraction7
A cue prompt adapting (CPA) model for relation extraction (RE) that encodes contextual features and semantic dependencies by implanting task-relevant cues in a sentence and a new transformer architecture is proposed to adapt pre-trained language models (PLMs) to perceive named entities in a relation instance.
- 7
- 7
- 7
- FVT: A Fragmented Video Tutor for "Dubbing" Software Development Tutorials7
This study proposed a novel method and its supporting tool — Fragmented Video Tutor (FVT), aiming at augmenting the strengths of both forms of tutorials while offsetting the weaknesses inherent to use each form of tutorials by itself.
- LearnerExp: Exploring and Explaining the Time Management of Online Learning Activity7
In this demo, this work introduces the LearnExp, an interactive visual analytic system designed to explore the temporal patterns of learning activities and explain the relationships between academic performance and these patterns.
- 7
- 7
- An efficient multi-feature SVM solver for complex event detection7
A self-adaptive multi-feature learning framework with efficient Support Vector Machine (SVM) solver for complex event detection in this paper that is able to utilize multiple features reasonably with an adaptively weighted linear combination manner, which is simple yet effective.
- A Bandwidth Variation Pattern-Differentiated Rate Adaptation for HTTP Adaptive Streaming Over an LTE Cellular Network7
The bandwidth variation pattern-differentiated rate adaptation (BVPDRA) algorithm for LTE cellular networks is proposed, which achieves a higher average bitrate, and lower rebuffering ratio with a reduced bitrate switching frequency.
- 7
- Modeling User Psychological Experience and Case Study in Online E-learning7
A quantitative approach of modeling user psychological experience in the context of intelligent e-learning is presented and the holistic user experience is quantified through the fusion of analytic hierarchy process (AHP) and Delphi methods.
- Optimal Resource Provisioning Approach based on Cost Modeling for Spark Applications in Public Clouds7
An approach based on time and monetary cost modeling is presented for cloud resource provisioning optimization under two typical constrained scenarios, which may save a significant amount of time and money, compared to randomly selected settings.
- Workload modeling for virtual machine-hosted application7
A workload model for VM-hosted application based on grey system theory, which identifies that the VM memory is a key bottleneck resource of the system, and a bimodal relationship exists between the workload and VM memory consumption.
- In-depth measurement and analysis on densification power law of software execution7
It is shown that when software systems are modeled as a series of dynamic Call Graphs during their executions, they always obey DPL with very close correlation, and the universality of DPL makes it an appropriate metric for software execution process.
- 7
- 7
- 7
- 7
- Lifetime characteristics measurement of a P2P streaming system: Focusing on snapshots of the overlay7
- 7
- Semantics-Based Answers Selection in Question Answering System7
A novel answer selection scheme for of NL-WASWebAnswer, designed for e-learning users, using semantic types of the user’s question and the semantic templates to implement semantics-based answers selection.
- SAUI: Scale-Aware Unseen Imagineer for Zero-Shot Object Detection6
The scale-varying challenge in ZSD is revealed and a Scale-Aware Unseen Imagineer (SAUI) is proposed to lead the way of a novel scale-aware ZSD paradigm to generate unseen features scale by scale.
- RDLNet: A Novel and Accurate Real-world Document Localization Method6
A novel Real-world Document Localization Network (RDLNet) for locating targeted documents in the wild is proposed and the results reveal that the RDLNet remarkably outperforms contemporary methods, demonstrating its superiority in terms of both accuracy and practicability.
- GraTO6
GraTO is proposed, a framework based on neural architecture search to automatically search for GNNs architecture that demonstrates that GraTo outperforms baselines in the over-smoothing metrics and achieves competitive performance in accuracy.
- 6
- A Tax Evasion Detection Method Based on Positive and Unlabeled Learning with Network Embedding Features6
This paper proposes a novel tax evasion detection method based on PU learning with Network Embedding features (PUNE), which effectively detects tax evasion based on basic features and transaction network features that are extracted by a network embedding algorithm.
- 6
- BioNorm: deep learning-based event normalization for the curation of reaction databases6
BioNorm is proposed, a novel method of normalizing bio-events extracted from scientific literature to entries in the bio-molecular reaction database, e.g. IntAct, which is the first attempt of event normalization in the biomedical text mining.
- Using Face Recognition to Detect “Ghost Writer” Cheating in Examination6
The SIFR framework is proposed, a three layers architecture based on face recognition technique and micro-service principle, to detect the ghostwriter who takes the exam for others and a prototype system is implemented based on open source projects and public cloud services.
- ATTENet: Detecting and Explaining Suspicious Tax Evasion Groups6
ATTENet is a novel visual analytic system for detecting and explaining suspicious affiliated-transaction-based tax evasion (ATTE) groups that combines basic features and structure features of each group in the network with network embedding method structure2Vec.
- 6
- Improving Chinese Sentiment Analysis via Segmentation-Based Representation Using Parallel CNN6
A novel sentiment-analysis model, which is capable of sentiment analysis task in Chinese, and which concatenate the representation of each segment to obtain the final representation of the text which does not only contain the semantic and syntactic features but also retains the essential sequential information.
- LTE-EMU: A High Fidelity LTE Cellar Network Testbed for Mobile Video Streaming6
It is proved that the fidelity of LTE-EMU is significantly higher than an LTE simulation environment through the fidelity model presented in the paper, which is a hybrid solution by combining methods of simulation and emulation.
- 6
- 6
- MOTIF-RE: Motif-Based Hypernym/Hyponym Relation Extraction from Wikipedia Links6
It is found that the proportions of hyperlinks with different semantic type vary markedly in different network motifs, which proves that MOTIF-RE, an algorithm of extracting hypernym/hyponym relation from Wikipedia hyperlinks has fairly good domain adaptation ability.
- Emotion Chat: A Web Chatroom with Emotion Regulation for E-Learners6
The result of the initial experiment shows that EmotionChat can recommend valuable emotion regulation policies for e-learners, and this work aims to compensate for lack of emotion communication between teachers and students in e-learning systems.
- Ordinal Regression with Sparse Bayesian6
A probabilistic framework for ordinal prediction is proposed, which can be used in modeling ordinal regression, and a sparse Bayesian treatment is given, in which an automatic relevance determination prior over weights is used.
- 6
- 6
- 6
- 6
- 6
- TED: related party transaction guided tax evasion detection on heterogeneous graph5
A novel graph neural network model is proposed to extract the comprehensive information of heterogeneous graphs and uses heterogeneous and complex related party transaction groups to filter low-level noise information in tax scenarios.
- Long- and Short-Term Preference Model Based on Graph Embedding for Sequential Recommendation5
A Long- and Short-Term Preference Network (LSPN) based on graph embedding for sequential recommendation is proposed, which constructs the network to obtain the user long- and short-term preferences separately, combing them through the fuzzy gate mechanism to provide the user final preference.
- 5
- 5
- Improving generalization ability of instance transfer-based imbalanced sentiment classification of turn-level interactive Chinese texts5
The experimental results show that the proposed approach outperforms classic algorithms by alleviating the imbalanced problem in interactive texts effectively and a classification model that is trained on immigrated datasets produced by employing the proposed method achieves the best ability of generalization.
- Recognizing physical contexts of mobile video learners via smartphone sensors5
This paper tests a classic learning scenario regarding mobile video viewing and validates the proposed recognition method by comprehensively taking the recognizing accuracy, effectiveness and the energy consumption into consideration.
- Security Collaborative Optimization Strategy for Perception Layer in Cognitive IoT5
An autonomic idea of cluster user collaboration to incorporate the concept of autonomic discipline to fine-tune single user processes and improve overall security performance in multilayer security elements is adopted.
- Dynamic cohesion measurement for distributed system5
Two new cohesion metrics are proposed to describe the correlation at component level, by extending the cohesion metric of single-server software system by differentiating methods of components deployed on different nodes.
- 5
- Directed test suite augmentation via exploiting program dependency5
An efficient approach called Directed Test Suite Augmentation (DTSA) is proposed to automatically generate test cases that can reach the changed statements, produce different program states after executing the changed states, and propagate the different states to the output of the program.
- How to Avoid Herd Behavior: A Stochastic Multi-Choice Scheduling Algorithm and Parameters Analysis in Grid Scheduling5
Experiments address that the multi-choice strategy can decrease herd behavior in large-scale sharing environment, at the same time providing increased scheduling performance and causing less scheduling burden than greedy algorithms.
- CBC: Caching for cloud-based VOD systems5
A systemic caching scheme is proposed, which seamlessly integrates a caching algorithm and a cache deployment algorithm together to maximize the resources utilization of cloud-based VOD system under resource-constrained situation and reduces the reject ratio of user requests.
- A hybrid approach for privacy-preserving processing of knn queries in mobile database systems5
The idea is to compute approximate query results based on cloaked location information and then refine query results by applying homomorphic encryption, and it is demonstrated that this approach can be used for efficient and privacy-preserving processing of KNN queries.
- Projection vector machine5
Different from random weight idea used in ELM, in PVM, the input weights of SLFN are obtained by singular decomposition, and the hidden neurons are ranked by their contribution level and only those that are important are selected to make PVM to obtain better generalization ability and more compact structure than ELM.
- 5
- 5
- 5
- 5
- An approach to acquire semantic relationships between terms5
The experimental result shows that the proposed method can delete most noisy pairs of terms and improve accuracy and efficiency of NL-WAS, and is complementary to statistically based approaches that find semantic relationships between terms.
- 5
- Simultaneous High-Resolution Blind Image Reconstruction and Perturbation Defense4
An efficient image high-resolution reconstruction and perturbation defense model, named DGMR, is proposed based on deep Gaussian mixture learning, which is superior to the existing state-of-the-art systems.
- 4
- 4
- 4
- 4
- Rethinking Sentiment Analysis under Uncertainty4
This work proposes a new SA paradigm with the consideration of uncertainty and introduces uncertainty quantification to obtain confidence intervals for predictions, which enables the risk assessment ability of the model and can improve the credibility of SA-aids decision-making.
- Towards Explanation for Unsupervised Graph-Level Representation Learning4
The Information Bottleneck principle is advanced and the connection between graph representations and explanatory subgraphs on the label space is theoretically analyzed, which reveals that the expressiveness and robustness of representations benefit the fidelity of explanatory subgraphs.
- 4
- 4
- Memory transformation networks for weakly supervised visual classification4
A Memory Transformation Network (MTN) is proposed to convert external knowledge, by involving embedded and concentrated memories, so as to leverage it feasibly for video classification with weak supervision, and outperforms the state-of-the-arts on visual classification with limited labeled exemplars.
- An Empirical Study of Architectural Changes in Code Commits4
A light-weight framework to identify changes in architectures from code commits automatically, which identifies architectural changes without heavy architecture recovery techniques and can be integrated into prevalent continuous integration systems to monitor architectural changes.
- Enhancing QoE for Mobile Users by Environment-Aware HTTP Adaptive Streaming4
This paper constructs and validates a video QoE model based on environment perception and then designs a model-driven, environment-aware HAS rate adaptation algorithm, which can provide appropriate differentiated rate adaptation for different environments, resulting in a higherQoE.
- 4
- 4
- Mining Suspicious Tax Evasion Groups in a Corporate Governance Network4
A colored and weighted network-based model for characterizing economic behaviors, controller interlock and other relationships, and IATs between corporations, and a heterogeneous information network-corporate governance network are presented.
- How Unlabeled Web Videos Help Complex Event Detection?4
A new robust dictionary learning framework for complex event detection, able to handle both labeled and easy-to-get unlabeled web videos by sharing the same dictionary is proposed, which shows strong robustness against the substantial noisy and outlier videos from open source.
- 4
- On Interval-Valued Pseudolinear Functions and Interval-Valued Pseudolinear Optimization Problems4
A class of interval-valued pseudolinear programs is considered, and the solution set of the interval- values of the Pseudolinear optimization problem is characterized.
- Analysis of user’s behavior and resource characteristics for private trackers4
An in-depth measurement and analysis of the PT system, in terms of resources and users’ behavior characteristics, indicates that the sequence of torrent activities followed a power-law distribution while the sequence of torrent’s age followed a linear distribution.
- A Dynamic Nonparametric Model for Characterizing the Topical Communities in Social Streams4
This work proposes a fully nonparametric Dynamic Topical Community Model (nDTCM), where infinite latent community variables coupled with infinite latent topic variables in each epoch, and the temporal dependencies between variables across epochs are modeled via the rich-gets-richer scheme.
- A Random Walk Approach to Selectional Preferences Based on Preference Ranking and Propagation4
Experimental results show that this unsupervised random walk approach outperforms several state-of-the-art methods on the pseudo-disambiguation task, and it better correlates with human plausibility judgements.
- 4
- 4
- 4
- Mining preorder relation between knowledge units from text4
The experimental results show the method can efficiently extract the preorder relation, and reduce the computational complexity caused by the quadratic problem of link mining.
- 4
- 4
- Using Cross Entropy to Detect and Classify Network Anomalous Traffic4
Experimental results and comparisons with the Shannon entropy measurement on Netflow traffic in a router show that under relatively weaker attacks, the true positive rate, average precision and accuracy of the cross entropy measurement in attack classification rise by 13%,15%, and 13%,respectively.
- 4
- Cost-Sensitive Supported Vector Learning to Rank Imbalanced Data Set4
The experimental results show that the performance of cost-sensitive supported vector learning to rank is better than Ranking SVM on the document retrieval data set.
- A boosting approach for intrusion detection4
Boosting classification algorithm is introduced into the area of intrusion detection to learn attack signatures and the Principle Component Analysis (PCA) approach is employed, an effective data reduction approach, to extract the key attribute set from the original high-dimensional network traffic data.
- 4
- 4
- 4
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- Exploring open information via event network3
An event network framework is proposed to address this challenge of discovering information from a large amount of data in an open domain and techniques such as social network or complex network can be applied to analyze the event network for exploring open information.
- A Segmentation Matrix Method for Chinese Segmentation Ambiguity Analysis3
A Segmentation Matrix (SM) method is presented, which encodes each sentence as a matrix, then maps string operation into set operations, and shows that SM is useful for CSA analysis.
- 3
- 3
- 3
- Faceted fusion of RDF data3
FF aggregates distributed RDF data on a specific topic according to different facets, based on two topological properties of a topic-specific RDF Graph (TRG), and outperforms three other approaches based on data distance, structure distance and structure/attribute respectively.
- 3
- 3
- Software structure evaluation based on the interaction and encapsulation of methods3
An approach to evaluate software quality based on the interaction and encapsulation of methods, which concurs with the intuition that minor upgrades, aimed at fixing bugs and adding minor functionality quickly, reduce software quality, whereas major upgrades, which normally clean up messy code and re-construct the software, improve software quality.
- 3
- Visualization of Knowledge Map: A Focus and Context Approach3
Experimental results show that the visualization approach highly speeds up the initialization time of KM, simultaneously displays both the local focus and global context, and effectively facilitates learner’s navigate learning.
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- Taxonomy Building and Machine Learning Based Automatic Classification for Knowledge-Oriented Chinese Questions3
A taxonomy for knowledge-oriented question is built after the study of previous work and analysis of 6776 Chinese knowledge- oriented questions collected from different realistic sources, and the machine learning based classification based on this taxonomy is investigated.
- 3
- 3
- Hopf Bifurcation in Differential Algebraic Equations and Applications to Circuit Simulation3
This paper solves an eigenvalue problem for the Jacobi-matrix D 1 f(x 0,λ) at an equilibrium point x 0 and uses the fundamental theorem of Hopf to solve this problem.
- Efficient Bi-objective SQL Optimization for Enclaved Cloud Databases with Differentially Private Padding2
TPOA incorporates two novel ideas: divide-and-conquer to separately handle parameters according to their types in optimization for dimensionality reduction; on-demand-optimization to progressively build a set of necessary Pareto-optimal plans instead of seeking a complete set for saving resources.
- 2
- 2
- 2
- Unsupervised Hierarchical Feature Selection on Networked Data2
A novel unsupervised feature selection framework (HNFS) that can simultaneously learn the implicit hierarchical structure among the nodes and embed the hierarchical structure into the feature selection process is proposed.
- Transmission Latency based Network Friendly Tree for Peer-to-Peer Streaming2
In this paper, network friendly tree is provided as overlay multicasting tree construction strategy for P2P streaming system, and the transmission performance improves significantly at minor additional cost.
- Cross-Graph Representation Learning for Unsupervised Graph Alignment2
An unsupervised cross-graph representation learning framework to jointly learn the node representations of different graphs in a unified deep model based on both attribute and structure reconstruction and introduces a discriminator to better align the learned representations for different graphs via adversarial training.
- Deploying external bandwidth guaranteed media server clusters for real-time live streaming in media cloud2
The evaluation results show that the proposed scheme guarantees the external bandwidth requirement of a real-time live streaming application, at the same time, greatly reduces the intra-bandwidth utilization of a media cloud with different DCN structures.
- TaxVis: a Visual System for Detecting Tax Evasion Group2
The demo presents TaxVis, a visual detection system for tax auditor that supports tax evasion group detection based on a two-phase detection approach that can analyze the suspicious groups automatically without artificial extraction of tax evasion patterns.
- Heterogeneous Information Network Hashing for Fast Nearest Neighbor Search2
This paper proposes a novel heterogeneous information network hashing algorithm based on collective matrix factorization that can lead to faster NNS and requires lower memory usage than several state-of-the-art network embedding methods while showing comparable performance in typical learning tasks on HINs, including link prediction and cross-type node similarity search.
- 2
- 2
- 2
- 2
- Modeling Users' Adoption Behaviors with Social Selection and Influence2
A probabilistic Latent Factors with Diffusion Model (LFDM) is proposed which explicitly considers both social selection and influence by projecting cascading processes into latent factor spaces and is validated on three kinds of real world data sets.
- 2
- A self-adaptive trust management scheme for wireless sensor networks2
This paper proposes a self-adaptive trust management scheme, which can adaptively adjust the trust evaluation process to cope with application requirement changes of trust management, and it is more suitable for large-scale sensor networks.
- Faceted Exploring for Domain Knowledge over Linked Open Data2
KFM is demonstrated, which can aggregate the distributed RDF data of a topic according to the facets of this topic and provides a new way for users to obtain and explore domain knowledge in the LOD cloud.
- Knowledge Unit Relation Recognition Based on Markov Logic Networks2
A knowledge unit relation recognition framework based on Markov Logic Networks, which combines probabilistic graphical models and first- order logic by attaching a weight to each first-order formula is proposed.
- MSC: a multi-version shared caching for multi-bitrate VoD services2
The sharing probability among different versions of the same video is leveraged and a multi-version shared caching (MSC) method is proposed to maximize the benefit of caching proxy to improve caching efficiency.
- 2
- PLIDMiner: A Quality Based Approach for Researcher’s Homepage Discovery2
Based on the observation that informative researchers’ homepages are organized in two ways, integrated and scattered, an effective discovering model is proposed, PLIDMiner, with F1 scores over 0.9 on labeled data, which can be applied to verify homepages’ quality.
- Ranking the Difficulty Level of the Knowledge Units Based on Learning Dependency2
The authors provide algorithms for ranking the knowledge units with objective and subjective difficulty scores and propose three hypotheses of factors influencing difficulty based on the correlation between the difficulty level of knowledge units and the learning dependency.
- 2
- 2
- An Emotion Regulation Model in an E-Learning Environment2
A model of emotion regulation is proposed that provides an optimal mood to learners in an e-Learning environment, and uses the emotion regulation strategy scale to determine the learner’s strategy of emotional regulation.
- Incremental Projection Vector Machine: A One-Stage Learning Algorithm for High-Dimension Large-Sample Dataset2
A novel learning algorithm for high-dimension large-scale data, by combining a new incremental dimension reduction with feed-forward neural network training simultaneously, called Incremental Projection Vector Machine (IPVM), which showed that the proposed one-stage algorithm IPVM was faster than two-stage learning approach, and performed better than conventional supervised algorithms.
- Multi-channel live streaming in service overlay network2
A tree-aware queue scheduling algorithm is proposed to reduce the overlay processing delay in view of the entire overlay network, and achieves a differentiated control among different sessions, provides load balancing among overlay nodes, and improves the delay performance on SON.
- 2
- 2
- 2
- A Multi-Tree Construction Algorithm for Multi-Channel Live Media Delivery on Overlay Service Network2
- 2
- 2
- A method for detecting code security vulnerability based on variables tracking with validated-tree2
The concept of validated tree is developed to track variables referenced by database operations in scripts, and by checking whether these variables are influenced by outside inputs, the database operations are proved to be secure or not.
- A Differentiated Service Supported Bandwidth Allocation Algorithm for Multiple Points Communications2
- Network intrusion detection method based on RS-MSVM2
The testing results show that the RS-MSVM method outperforms other methods mentioned in this paper on six aspects: detection accuracy, number of support vectors, false positive rate, false negative rate, training time and testing time.
- An Approach to Acquire Semantic Relationships Between Words from Web Document2
This work applies not only strict unambiguous linguistic rules but also loose ambiguous linguistic rules to extract relationships and proposed efficient approach to refine the outputs of these rules.
- 2
- 1
- 1
- Uncertainty-aware large language model response length perception1
This paper thoroughly analyze the challenges involved in the LLM response length prediction task and proposes a new framework that treats it as an uncertainty-aware regression problem, and benchmark four uncertainty quantification methods, including both Frequentist and Bayesian approaches, and finds that evidential deep learning (EDL) is the most effective and efficient for this task.
- 1
- 1
- Teaching MLP More Graph Information: A Three-stage Multitask Knowledge Distillation Framework1
This work is the first work to include hidden layer distillation for student MLP on graphs and to combine graph Positional Encoding with MLP, and is the first work to include hidden layer distillation for student MLP on graphs.
- 1
- 1
- Disentangled Generation with Information Bottleneck for Few-Shot Learning1
A novel Information Bottleneck (IB) based Disentangled Generation Framework for FSL, termed as DisGenIB, that can simultaneously guarantee the discrimination and diversity of generated samples and can effectively utilize priors to further facilitate disentanglement.
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- Toward Cost-Effective Mobile Video Streaming through Environment-Aware Watching State Prediction1
This paper takes into account the environment-aware watching state and proposes a cost-effective mobile video streaming scheme to reduce power consumption and mobile data usage and introduces a downloading algorithm based on the user equipment running mode in the LTE system and the VLC player.
- 1
- 1
- 1
- A Fast Exact Algorithm for Deployment of Sensor Nodes for Internet of Things1
The experiments show that the new algorithm can reduce the running time by at least an order of magnitude, compared to a state-of-the-art algorithm for listing all the minimal dominating sets.
- 1
- Attributing hacks with survival trend filtering1
An algorithm for estimating the provenance of hacks on websites that uses hazard regression with a time-varying additive hazard function parameterized in a generalized linear form to attribute individual attacks to joint causes and vulnerabilities as well as estimate the evolution of these vulnerabilities over time.
- 1
- Nonparametric models for characterizing the topical communities in social network1
This work proposes two nonparametric statistic models where infinite latent community variables coupled with infinite latent topic variables are coupled via a rich-gets-richer scheme and develops an effective online posterior inference algorithm for the models, which is concordant with the online nature of social streams.
- Reducing test cases with causality partitions1
A heuristic approach based on program causality is developed to selectively explore program paths in symbolic execution and can significantly reduce the number of explored paths and the symbolic execution time to improve the performance of test case generation.
- PERFORMANCE ANALYSIS OF GRID ARCHITECTURE VIA QUEUEING THEORY1
Two new grid performance metrics are proposed to measure performance of grid systems and some scientific principles guiding the grid architecture design and optimization are obtained.
- 1
- 1
- 1
- 1
- Automatic Chinese Topic Term Spelling Correction in Online Pinyin Input1
This paper proposes a novel Chinese spelling correction model directly targeting at the original keyboard input, and integrates this model to an online Chinese input method, to improve the spelling suggestion feature.
- 1
- A Method of Building Learner Model in Personalized e-Learning1
The work discusses and evaluates how to find the key personality attributes and their weight, and how to mine learner interest from learner behavior, and shows the modeling method is effective in personalized e-learning.
- 1
- 1
- 1
- 1
- A mini-bandwidth guaranteed and differentiated service supporting bandwidth allocation algorithm for Multimedia Group Communication1
The proposed algorithm not only guarantees each media stream's basic bandwidth, but also provides differentiated service according to users' different priorities, which allows users to control the bandwidth of each video stream independently under various quality of service requirements and resource constraints.
- A Classified Method Based on Support Vector Machine for Grid Computing Intrusion Detection1
A novel ID method based on Support Vector Machine (SVM) is proposed to solve the classification problem for the large amount of raw intrusion event dataset of the grid computing environment and a new radial basic function (RBF) based on heterogeneous value difference metric (HVDM) of heterogeneous datasets is developed.
- –
- Early detection of gastric cancer via multiplex blood assay targeting CfDNA methylation signatures–
This assay demonstrated the ability to distinguish early-stage GC patients from normal controls and specifically differentiated GC patients from those with other tumors, achieving an AUCs of 0.932 and 0.927.
- –
- –
- –
- The Innovative Development of Artificial Intelligence and STEM Education-Cognition and Practice–
It is essential to understand the principles and objectives of talent development, emphasizing the integration of theoretical research and practical engineering, and the mechanisms of the integrated development of AI and science education, as well as the scientific and technological issues involved.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- A Boundary Regressing Model for Nested Named Entity Recognition.–
The regression operation is introduced to locate named entities in a sentence and predict the class probability by sharing parameters for locating and predicting, which can take full advantage of annotated data and enable more potent nonlinear function approximators to enhance model discriminability.
- –
- –
- A Behavior-Item Based Hybrid Intention-Aware Frame for Sequence Recommendation–
A new Behavior-Item based Hybrid Intent-aware Framework (BIHIF) is proposed, in which the user’s main intent is extracted based on user behaviors and interactive items, respectively, and the two intent vectors are combined and extracted by the full connection layer to obtain the user's real intent.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Well-Balanced Successive Simple-9 for Inverted Lists Compression–
This paper proposes a successive storage structure and processing metric to compress two successive Simple-9 encoded sequence of integers in a single data processing procedure, thus the name SSimple-9 (SSimple9), and studies the compression and decompression efficiency speedup of this algorithm.
- –
- –
- –
- –
- –
- –
- Next Generation of Web-based Information Management System for National Volunteer Service–
This paper presents a survey about the existing web-based information management system for volunteer services and highlights the new features and functions of the next generation of the volunteer service system.
- Analysis of academic thought febrile diseases of ZHANG Bi–
His academic thought about febrile diseases is of innovation based on the inheritance of ZHANG Zhong-jing's academic thought and studies on his academic thought are of great significance for understanding development of Chinese medicine academic thought of febRIle disease in the Jin and Yuan dynasties.
- –
- –
- –
- –
- –
- –
- –
- –
- Application-Level Multicast Routing Algorithms for Supporting Multipoint Interaction Synchronization–
An effective low-delay multicast routing algorithm that supports multipoint interaction synchronization is proposed and both simulation experiments and practical applications demonstrate that the algorithm is correct and effective.
- –
- –
- Design of analog multiplier for PFC–
An analog multiplier based on BCD process is introduced, and principle and construction of the circuit are analyzed, and the result shows that output is a half sine wave which has same frequency and phase as input, its amplitude is up to 1.2V.
- –
- –
- –
- An Attribute Reduction Algorithm to Find Learner?s Key Characteristics Based on the Discernible Function–
The attribute reduction method is applied to the network English learning platform of Xi′an Jiaotong University and the result shows the dimensions of key personality characteristics are one-quarter of the original ones.
- –
- A MAC Protocol for Ad Hoc Networks Based on Geographic Information–
A new protocol called GM-MAC is presented that introduces geographic information into MAC based on multiple channels that agrees the shortest distance between two nodes has a better priority and provides better QoS.
- –
- –
- –
- Neighbor Words Selection Algorithm on Imbalanced Data–
A novel feature selection algorithm, named PN (Positive-Negative feature selection), is proposed for term extraction and the experimental results show that the performance of PN algorithm outperforms those of CC and IG.
- A Method to Find Learner’s Key Characteristic in Wed-Based Learning–
The concept of key personality characteristic and an improved algorithm basing on rough set theory to find the key attributes are put forward and it is shown the dimensions of personality characteristic can be decreased to below one-quarter.
- –
- –
- –
- –
- Performability Analysis of Grid Architecture Via Queueing Networks–
On the basis of a coarse-grain classification of grid applications, a novel grid architecture taxonomy, interaction-intensive and computation-intensive architecture is presented and scientific principles guiding the grid architecture design are obtained.
- Unexpected Exceptions Handling Based on Chinese Question Answering in Collaborative Design–
Two methods for mining similar records are proposed, and the performance comparison is shown that the proposed knowledge acquisition based approach is more efficient, extendable and precise.
- –
- –
- –
- New approach to WLAN security with synchronized pseudo random–
SPRNG has already been simulated and tested in experiment, it shows that SPRNG has stronger security than WEP because it reveals little information for attackers, and the key problem of SPRNG, synchronization loss problem, is presented.
- Technical Issues of Sharing and Integration of OPAC and E-Learning Resources–
A prototype of sharing search system named Fsearch was developed, with which users can search for both OPAC and e-learning resources simultaneously and shows that Fsearch system can effectively improve the efficiency of acquiring knowledge for users and bring forward a preferable application prospect.
- –
- Workstation cluster as a parallel hardware environment for high performance computing–
Concepts for remote procedure calls specially suited for heterogeneous workstation clusters, based on concepts such as dynamical load balancing, fault tolerance and a classification of input- and output parameters the code generator TAPAS is developed.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.