Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks135 from public data
- SenticNet 6: Ensemble Application of Symbolic and Subsymbolic AI for Sentiment Analysis466
This work integrates logical reasoning within deep learning architectures to build a new version of SenticNet, a commonsense knowledge base for sentiment analysis, and applies it to the interesting problem of polarity detection from text.
- Word Embedding for Understanding Natural Language: A Survey241
This survey introduces the motivation and background of word embedding, the methods of text representation as preliminaries, as well as some existingword embedding approaches such as Neural Network Language Model and Sparse Coding Approach, along with their evaluation metrics.
- A Generative Model for category text generation167
Not only the proposed model is able to generate category sentences that enlarge the original dataset, but also it helps improve its generalization capability during supervised training, and is evaluated for the task of sentiment analysis.
- Learning multi-grained aspect target sequence for Chinese sentiment analysis164
This paper formalizes the problem of aspect-level sentiment analysis from a different perspective, i.e., that sentiment at aspect target level should be the main focus and proposes to explicitly model the aspect target and conduct sentiment classification directly at the aspect targets level via three granularities.
- Recent Developments in Recommender Systems: A Survey [Review Article]124
In this technical survey, the latest advancements in the field of recommender systems are comprehensively summarized, and the robustness, data bias, and fairness issues in recommender systems are analyzed.
- Disentangled Variational Auto-Encoder for semi-supervised learning97
A new model is proposed called Semi-supervised Disentangled VAE (SDVAE), which encodes the input data into disentangled representation and non-interpretable representation, then the category information is directly utilized to regularize the disentangling representation via the equality constraint.
- Learning binary codes with neural collaborative filtering for efficient recommendation systems80
A new deep learning framework is proposed for deep collaborative hashing codes on user–item ratings that adopts neural networks to better learn both user and item representations and make these close to binary codes such that the quantization loss is minimized.
- A survey on cybersecurity attacks and defenses for unmanned aerial systems73
A comprehensive review of UAS cybersecurity research, with a focus on attack and defense technologies, analyzes the UAS architecture and classifies security threats into four categories: communication network security, software security, payload security, and intelligent security.
- Multitask learning for emotion and personality traits detection63
This paper proposes a novel transferring based multitask learning framework that simultaneously predicts both personality traits and emotional behaviors and achieves the state-of-the-art performance across multiple famous personality and emotion datasets, even outperforming model-based models.
- 57
- Phonetic-enriched text representation for Chinese sentiment analysis with reinforcement learning52
Experimental results on five different Chinese sentiment analysis datasets show that the inclusion of phonetic features significantly and consistently improves the performance of textual and visual representations and outshines the state-of-the-art Chinese character level representations.
- Hformer: highly efficient vision transformer for low-dose CT denoising50
Hformer, a novel supervised learning model for low-dose computer tomography (LDCT) denoising, combines the strengths of convolutional neural networks for local feature extraction and transformer models for global feature capture.
- Informal prototyping of continuous graphical interactions by demonstration50
Informal feedback showed that Monet is a promising approach to enabling more complete tool support for early stage UI design, by showing that Monet allows continuous interactions to be eas-ily integrated with event-based, discrete interactions.
- 49
- 38
- 38
- 30
- 28
- PrimeNet: A Framework for Commonsense Knowledge Representation and Reasoning Based on Conceptual Primitives25
This work designs a new commonsense knowledge base, termed PrimeNet, organized in a three-layer structure: a small core of conceptual primitives, a bigger set of concepts that connect to such primitives, and an even larger layer of entities connecting to the concepts.
- Secure Software Engineering Education: Knowledge Area, Curriculum and Resources25
Current efforts and resources in secure software engineering education, including Common Body of Knowledge, reference curriculum, sample curriculum materials, hands-on exercises, and resources developed by industry and open source community are reviewed.
- 23
- 22
- Data Augmentation and Classification of Sea–Land Clutter for Over-the-Horizon Radar Using AC-VAEGAN21
The experiment results show that the quality of samples synthesized by AC-VAEGAN is better than those synthesizing by the state-of-the-art GAN-based methods, and the data augmentation method is able to improve the classification performance in the case of the imbalanced and scarce sea–land clutter samples.
- 21
- Graph routing between capsules20
A new capsule network with graph routing to learn both relationships, where capsules in each layer are treated as the nodes of a graph, and the approach combining bottom-up routing and top-down attention performs the best.
- Advanced Design Approaches to Emerging Software Systems: Principles, Methodologies and Tools20
This book will help to clarify the present chaotic and confusing literature of the current state of the art and knowledge in the areas of the design and engineering of the many emerging software systems.
- DOEPatch: Dynamically Optimized Ensemble Model for Adversarial Patches Generation19
The concept of energy is introduced and the adversarial patches generation process is treated as an optimization of the adversarial patches to minimize the total energy of the “person” category by adopting adversarial training and constructing a dynamically optimized ensemble model.
- Popularity prediction on vacation rental websites18
Empirical results show the effectiveness of the proposed DGRU and the encoder-decoder framework in two-correlated sequences prediction and popularity prediction, respectively.
- Bioinformatics software development: Principles and future directions17
This review outlines the key stages and considerations in the development of bioinformatics software, using clusterProfiler and CIRCexplorer2 as illustrative examples, and examines some established large-scale life sciences platforms.
- 17
- 17
- 17
- MMCR: Advancing Visual Language Model in Multimodal Multi-Turn Contextual Reasoning15
MMCR (Multimodal Multi-turn Contextual Reasoning) is presented, a novel dataset comprising a novel dataset with 310K contextual dialogues, spanning 8 domains (Humanities, Natural, Science, Education, etc.) and 40 sub-topics and showing consistent improvements on existing benchmarks.
- Teaching Global Software Engineering by Simulating a Global Project in the Classroom15
An exercise for teaching Global Software Engineering in a single classroom is described and students experienced some of the challenges of global distribution of project teams and tried to deal with them in a simulated environment.
- (Position Paper) Applying software engineering methods and tools to CSE research projects14
This paper first presents two case studies on two different CSE research projects, where common and specific problems are identified and proposes solutions that intend to apply software engineering methodologies and tools to improve the CSEResearch software development.
- Enterprise information portal14
The relationship between Enterprise Information Portal and other information systems is introduced and the future of EIP'8 technology development is expected.
- 13
- 12
- 12
- A Review of Anomaly Detection Techniques Based on Nearest Neighbor12
This paper has attempted to provide an overview of the previous work on outlier detection and anomaly detection, and to reduce the computation complexity.
- 11
- Reengineering a scientific software and lessons learned11
Through the reengineering process on SeisSol, the system is more modularized and easier to be extended, as well as the source code are more comprehensible.
- Physical Adversarial Attack on Monocular Depth Estimation via Shape-Varying Patches10
This work proposes a physics-based adversarial attack on MDE using a framework called an attack with shape-varying patches (ASP), which optimizes the content, shape, and position of patches to maximize its disruptive effectiveness on the sensor’s output.
- 10
- 9
- Towards AI-Powered AR for Enhancing Sports Playability for People with Low Vision: An Exploration of ARSports9
It is suggested that combining real-time computer vision and augmented reality to create scene-aware visual augmentations is a promising approach to enhancing sports participation for LV individuals.
- Adversarial Attack and Defense in Breast Cancer Deep Learning Systems9
A deep learning system for breast cancer pathology image recognition with better defense performance is built using the adversarial attack algorithm FGSM to reveal that breast cancer deep learning systems are vulnerable to attacks and thus misclassify Breast cancer pathology images.
- Cl2sum: abstractive summarization via contrastive prompt constructed by LLMs hallucination8
CL2Sum is proposed, a method that combines Contrastive Learning (CL) and ICL for summarization that effectively alleviates the hallucination problem of text summaries while also improving the overall quality of the generated summaries.
- 8
- 7
- Enhancing Continuous Domain Adaptation with Multi-Path Transfer Curriculum7
A novel CDA method, W-MPOT, is contributed, which rigorously addresses the domain ordering and error accumulation problems overlooked by previous studies, and constructs a transfer curriculum over the source and intermediate domains based on Wasserstein distance.
- Evolutionary public goods game on the birandom geometric graph7
A birandom geometric graph is established, in which two types of nodes, representing players and public goods respectively, are placed at random locations in the unit square, which suggests that cooperation can be significantly promoted when the group size and the average number of groups that each player participates in are relatively small.
- 7
- 6
- Research on multi-algorithm and explainable AI techniques for predictive modeling of acute spinal cord injury using multimodal data6
This study is the first to apply multiple machine learning methods and Shapley Additive Explanations (SHAP) analysis specifically to ASCI for predicting neurological recovery and demonstrated excellent performance across multiple metrics.
- 6
- ATTACK-COSM: attacking the camouflaged object segmentation model through digital world adversarial examples6
This work represents the first exploration of strategies for targeting COSM with adversarial examples in the digital world, and explores the adversarial robustness of COSM in full white-box and black-box settings.
- Data-driven active corrective control in power systems: an interpretable deep reinforcement learning approach6
A DRL model is constructed based on the Markov decision process (MDP) to effectively address active corrective control issues in a 36-bus system and a feature importance explainability method is proposed, validating that the proposed feature importance-based explainability method enhances the transparency and reliability of the DRL model for active corrective control.
- 6
- OPTAGENT: Optimizing Multi-Agent LLM Interactions Through Verbal Reinforcement Learning for Enhanced Reasoning5
A multi-agent verbal reinforcement learning algorithm that dynamically constructs and refines multi-agent collaboration structures and defines action spaces and a feedback mechanism that evaluates communication robustness and coherence throughout the debate is proposed.
- 5
- Do Deep Neural Networks Always Perform Better When Eating More Data?5
Experiments from Identically Independent Distribution and Out of Distribution are designed to provide guidance from the perspective of data, which can promote a wide range of practical applications of artificial intelligence.
- Grouped Variable Selection Using Area under the ROC with Imbalanced Data5
A methodology to select grouped variables using the area under the ROC with an adjustable prediction cut point to enhance the accuracy of classification for the minority class by maximizing the true positive rate.
- JudgeBoard: Benchmarking and Enhancing Small Language Models for Reasoning Evaluation4
This work proposes JudgeBoard, a novel evaluation pipeline that directly queries models to assess the correctness of candidate answers without requiring extra answer comparisons, and proposes MAJ (Multi-Agent Judging), a novel multi-agent evaluation framework that leverages multiple interacting SLMs with distinct reasoning profiles to approximate LLM-level judgment accuracy through collaborative deliberation.
- BeyondGender: A Multifaceted Bilingual Dataset for Practical Sexism Detection4
BeyondGender is introduced, a dataset meticulously annotated according to the latest definitions of misogyny and misandry that features innovative multifaceted labels encompassing aspects of sexism, gender, phrasing, misogyny, and misandry.
- 4
- CAMP in the Odyssey: Provably Robust Reinforcement Learning with Certified Radius Maximization3
A novel paradigm dubbed CAMP is introduced, designed to enhance DRL policies, achieving better utility without compromising provable robustness, and introduces \textit{policy imitation} as a novel technique to stabilize training.
- 3
- Reservoir Stochastic Simulation Based on Octave Convolution and Multistage Generative Adversarial Network3
A hybrid framework combined with octave convolution and multi-stage GAN (OctSinGAN) is proposed to perform reservoir simulation and results show that different simulations are similar to the corresponding TIs in terms of spatial variability, channel connectivity and spatial structures, with a relatively high simulation performance overall.
- 3
- Variable selection in strong hierarchical semiparametric models for longitudinal data3
This paper considers the variable selection problem in semiparametric additive partially linear models for longitudinal data, and proposes an iterative estimation procedure by penalizing the likelihood with a partial group minimax concave penalty and using BIC to select the tuning parameter.
- 2
- 2
- Basis Selection: Low-Rank Decomposition of Pretrained Large Language Models for Target Applications2
Deep compression results on the Llama 2-7b and -13B models show that the method significantly reduces model size while maintaining comparable accuracy to state-of-the-art low-rank compression techniques.
- Finite-Time Prescribed Performance Tracking Control for Unmanned Helicopter System Using Neural Network2
By using Lyapunov synthesis, it is strictly proven that the finite-time convergence of the closed-loop control system can be achieved and tracking errors are always within the prescribed performance bounds.
- PGN: A Perturbation Generation Network Against Deep Reinforcement Learning2
This paper explores the vulnerability of deep reinforcement learning by proposing a novel generative model for creating effective adversarial examples to attack the agent and proposes the action consistency ratio as a measure of stealthiness, and a new measurement index of effectiveness and stealthiness.
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- Attentional value-factorization-based resource allocation and performance evaluation for intelligent connected vehicles1
Extensive experimental results show that AVF can promote cooperative resource allocation among agents by evaluating which agents and what actions have been contributing most to the overall quality of communication, achieving optimal resource efficiency and accurate evaluation.
- Multi-Agent Distributed Online Collaborative Target Search Based on Clustering Expectation Detection1
- A Sea-Land Clutter Classification Framework for Over-the-Horizon-Radar Based on Weighted Loss Semi-supervised GAN1
Improved generative adversarial network WL-SSGAN can improve the performance of the fully supervised classifier with only a small number of labeled samples by utilizing a large number of unlabeled sea-land clutter samples and is superior to both the adversarial loss and the feature matching loss.
- DBA: downsampling-based adversarial attack in medical image analysis1
This paper proposes a novel adversarial attack method, which is designed to exploit the DNNs’ generic down-sampling operation to ensure the effectiveness, stealthiness, and transferability of the attack.
- 1
- 1
- Anomaly detection and recognition of video surveillance images based on deep learning1
Experimental results show that this algorithm can detect abnormal interference such as overexposure of brightness, color distortion, and video freezes in video surveillance, with an accuracy rate of 86%.
- 1
- 1
- A Brief Talk about the Modern Application of Computer Software Technology and the Development Trend of Research1
The experimental results show that the modern application of computer software technology in the future is the trend, which will get the key help and full development of the country.
- 1
- An Entropy-based Algorithm for Detecting Overlapping Communities in Hyper-networks1
An algorithm is proposed based on hypergraph entropy to detect overlapping communities, which generalizes Cho's algorithm and is applied to a collaboration network of scientists studying graph theory in mainland China and demonstrates that the approach is applicable and effective.
- Requirements Engineering for Computational Seismology Software1
This paper presents a model-based requirements engineering approach that incorporates abstractions and notations from the seismology domain, and describes two requirements patterns, which facilitate the identification and specification of requirements in seismological software development.
- 1
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Rethinking Multimodal Sentiment Analysis: A High-Accuracy, Simplified Fusion Architecture–
This work proposes a lightweight, yet effective fusion-based deep learning model tailored for utterance-level emotion classification, achieving classification accuracy of 92% across six emotion categories and demonstrates that with careful feature engineering and modular design, simpler fusion strategies can outperform or match more complex models, particularly in resource-constrained environments.
- –
- –
- –
- A Two-layer Coupled Network Disease Transmission Model Considering Multiple Messages–
A proposed SIR−UAOE1E2 information-disease two-layer coupling network propagation model takes into account individual heterogeneity and the mutual influence of nodes, introducing hesitant individuals to play a positive role in suppressing disease outbreaks.
- State-only imitation learning via generative adversarial and inverse dynamics model–
This paper uses a state-only demonstrations to guide agents learning in a reward-sparse environment and proposes a policy optimization from observation (POfO) method that is comparable to that of the method using the complete demonstrations and is significantly better than other methods of the same type.
- –
- –
- –
- BLM-17m: A Large-Scale Dataset for Black Lives Matter Topic Detection on Twitter–
A Twitter dataset that covers one of the most significant human rights contradiction in recent years which affected the whole world: the George Floyd incident is constructed and a labeled dataset for topic detection is proposed that contains about 17 million tweets.
- –
- Monitoring image anomaly detection based on lightweight network MobileNet–
The results show that this recognition method can detect abnormalities such as freeze, overexposure, blur, and colorific distortion in remote video surveillance, with an accuracy rate of 91.75%, and can reduce the amount of model parameters and facilitate model deployment.
- –
- Multi-attending Memory Network for Modeling Multi-turn Dialogue–
A multi-head attention based end-to-end approach called multi-attending memory network without additional information or knowledge is proposed, which can effectively model and reason about multi-turn history dialogue.
- –
- An uniformizing method of MR image intensity transformation–
A uniformizing method is proposed to make the distribution even while ensuring that similar intensities of MR images reflect the same tissue after processed to significantly improve labeling and segmentation accuracy as compared to conventional preprocessed methods.
- Research on Standards System Framework of Agricultural IOT–
The paper applies the system engineering and modularization concepts and puts forward the three components of agricultural IOT modularized standards system framework, aiming to provide references for the construction of Agricultural IOT standards system which accords with China's situation and international standards.
- A Remote and Unified Software Automated Deployment Platform–
The experiment shows that the deployment platform can automatically detect software deployment conflicts, adjust conflicts, make collaborative decisions of software deployment, realize the unified and intelligent deployment at last.
- Finding δ-Closure Property of BitTorrent System and Applications Based on This Property–
This paper defined δ-closure according to the complex network theory, and tried to design algorithm to attain ε-closure, which is the existing of δ -closure in BT system.
- Research on the Training Mechanism of Embedded Software Development Talents–
Practical teaching results proved that infiltrating embedded software developing ideas and knowledge to existed related computer major courses and introducing students to participate in embeddedSoftware developing course design and application project helped improve students’ embedded software design ability.
- Unified Product Data in Transmission System Development–
A UPD integration framework based on collaborative agent is presented, which shows the evolvement theory among PUPD, DUPD, SUPD and IUPD, and task-role-based access control model is proposed, and dynamic right control process is given for UPD.
- –
- Discussion on Teaching the Course of Software Project Management–
The article discusses and researches on the course's teaching content and teaching method with the author's teaching practice in order to improve teaching effect.
- Research on Performance of EJB Software Architecture–
The experimental results show that the new EJB software architecture can get good result in improving the performance and efficiency of distribute system.
- –
- Application and Research of Quality Control modes in Software Project Manayement–
The thesis begins with the two attributes (Quality relativity and Quality economics) of the Quality, looking into the importance of the quality management, and points out two control models of thequality management which could make software project management vaild and effective.
- –
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 2026-10-12. 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.