Article
Computer Science, Hardware & Architecture
Menglun Zhou, Yifeng Zheng, Songlei Wang, Zhongyun Hua, Hejiao Huang, Yansong Gao, Xiaohua Jia
Summary: With the increasing popularity of sensor-rich mobile devices, spatial crowdsourcing has become a new crowdsourcing paradigm that leverages the crowd to perform location-dependent tasks. However, location privacy of workers can be compromised in the process of location-based task assignment. To address this issue, this paper proposes PPTA, a system framework that ensures location privacy-preserving task assignment in spatial crowdsourcing with strong security guarantees. PPTA utilizes lightweight cryptography and provides tailored secure components for practical location-based task assignment processes. Experimental results on a real-world dataset demonstrate that PPTA achieves strong security while maintaining comparable efficiency to plaintext baselines.
Article
Computer Science, Information Systems
Yu Fan, Liang Liu, Xingxing Zhang, Huibin Shi, Wenbin Zhai
Summary: Due to its wide coverage and strong scalability, spatial crowdsourcing (SC) has become a research hotspot. However, accurate location provision for task assignment poses a risk to location privacy. Existing works fail to meet the different privacy requirements and do not consider multi-location tasks. In this paper, we propose the Multi-location Task Allocation Problem with personalized location privacy protection (MLTAP) and a framework called MAPP, which efficiently allocates tasks based on a filtering mechanism and ranking metrics.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Jianhao Wei, Yaping Lin, Xin Yao, Jin Zhang
Summary: This paper proposes a differential privacy-based location protection scheme for protecting the location privacy of workers and tasks in spatial crowdsourcing, while achieving efficient task allocation.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2022)
Article
Computer Science, Information Systems
Mingzhe Li, Jingrou Wu, Wei Wang, Jin Zhang
Summary: The article presents an online framework for assigning tasks to workers in a fully distributed manner while protecting location privacy. The system uses homomorphic encryption to protect the location privacy of both workers and tasks, and proposes novel wait-and-decide and proportional-backoff mechanisms to increase the number of assigned tasks efficiently and in a privacy-preserving manner.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Engineering, Civil
Junwei Zhang, Fan Yang, Zhuo Ma, Zhuzhu Wang, Ximeng Liu, Jianfeng Ma
Summary: This paper proposes a decentralized location privacy-preserving spatial crowdsourcing system for IoV, which introduces blockchain technology and encryption verification to ensure the privacy of task locations and worker locations, and utilizes multi-level protection and zero-knowledge proofs to prevent malicious activities.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Yan Zhao, Jiaxin Liu, Yunchuan Li, Dalin Zhang, Christian S. Jensen, Kai Zheng
Summary: This paper proposes a novel preference-aware group task assignment framework for spatial crowdsourcing, which includes two components: Mutual Information-based Preference Modeling and Preference-aware Group Task Assignment. The framework learns group preferences using mutual information and weights group members adaptively. It also employs tree decomposition to assign tasks to appropriate worker groups, prioritizing more interested groups.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Information Systems
Chenxi Qiu, Anna Cinzia Squicciarini, Ce Pang, Ning Wang, Ben Wu
Summary: In this paper, the authors address the issue of Vehicle-based spatial crowdsourcing Location Privacy (VLP) over road networks. They propose a location obfuscation strategy to minimize quality loss while satisfying geo-indistinguishability. They approximate VLP to a linear programming problem and employ discretization and constraint reduction techniques to improve time efficiency. Experimental results demonstrate the superiority of their approach in terms of both quality-of-service and privacy.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Article
Computer Science, Information Systems
Fagen Song, Tinghuai Ma
Summary: This paper proposes a new method for protecting location privacy, which can protect both the user's and crowdsourcing task's location privacy. Compared with other methods, this method has a higher success rate of task allocation and shorter travel distance for crowdsourcing workers. By converting coordinates to polar coordinates and performing differential privacy transformation, the utility of the sanitized dataset is improved.
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Yan Zhao, Kai Zheng, Yunchuan Li, Jinfu Xia, Bin Yang, Torben Bach Pedersen, Rui Mao, Christian S. Jensen, Xiaofang Zhou
Summary: In spatial crowdsourcing, mobile users are involved in spatio-temporal tasks that require travel to specific locations. The task assignment in spatial crowdsourcing is a challenging problem that needs to be addressed to maximize profits. This study introduces a profit-driven task assignment problem and proposes various algorithms, including an optimal algorithm based on tree decomposition and greedy algorithms based on random tuning optimization. Additionally, a heuristic algorithm based on ant colony optimization is provided to balance effectiveness and efficiency. Extensive experiments using real and synthetic data are conducted to evaluate the proposed methods.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Information Systems
Peicong He, Yang Xin, Bochuan Hou, Yixian Yang
Summary: This paper proposes a privacy-preserving hitchhiking task assignment scheme for spatial crowdsourcing (SC), called PKGS. It protects the location privacy of workers and tasks, and assigns tasks to workers with the shortest travel distance.
Article
Engineering, Multidisciplinary
Mengyao Peng, Jia Hu, Hui Lin, Xiaoding Wang, Peng Liu, Kapal Dev, Sunder Ali Khowaja, Nawab Muhammad Faseeh Qureshi
Summary: This study proposes a Spatiotemporal Prediction based Spatial Crowdsourcing strategy (SPSC) using both blockchain and artificial intelligence to address the efficiency and privacy leakage issues in spatial crowdsourcing task allocation. By considering the temporal and spatial continuity of task data, SPSC utilizes gated recurrent unit and variational autoencoder for task prediction, and adds Laplacian noises to protect the privacy of crowdsourced workers. Moreover, SPSC employs blockchain technology to classify and group tasks and workers, reducing the risk of privacy data theft.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2023)
Article
Automation & Control Systems
Xuewen Dong, Wen Zhang, Yushu Zhang, Zhichao You, Sheng Gao, Yulong Shen, Chao Wang
Summary: This article discusses the privacy protection of task locations in mobile crowdsensing and proposes a codebook-based task allocation mechanism to address the issue of task location privacy leakage. By considering the tradeoff between local privacy and system utility, the optimal task allocation scheme is derived. Experimental results show that the introduction of the selected allocation codebook (SAC) method can improve task location privacy protection by an average of 60%.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Yuan Xie, Yongheng Wang, Kenli Li, Xu Zhou, Zhao Liu, Keqin Li
Summary: With the widespread use of GPS-equipped devices, spatial crowdsourcing (SC) technology has become popular in our daily lives. SC hires mobile users as workers who physically go to the task location and perform the task. This paper addresses the satisfaction-aware task assignment (SATA) problem in SC with the goal of maximizing overall user satisfaction, which combines satisfaction towards price and cooperation quality. The paper proposes the conflict-aware greedy (CAG) algorithm and game theoretic (GT) algorithm to solve the SATA problem, with the CAG algorithm providing an efficient result with a provable approximate bound and the GT algorithm finding a convergent Nash equilibrium. Extensive experiments demonstrate the effectiveness and efficiency of the proposed approaches on real and synthetic datasets.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Zhao Liu, Kenli Li, Xu Zhou, Ningbo Zhu, Yunjun Gao, Keqin Li
Summary: This paper focuses on the task assignment problem in spatial crowdsourcing and proposes the multi-stage complex task assignment problem. Unlike previous studies, this problem considers the dependency among tasks and presents greedy and game algorithms to solve it. Experimental results show the efficiency of the proposed algorithms.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Zhibin Wu, Lijie Peng, Chuankai Xiang
Summary: In this paper, a time-prediction-based task assignment approach in spatial crowdsourcing (TP-TASC) is proposed to improve task assignment in spatial crowdsourcing. The proposed method predicts travel time using historical data, and assigns tasks to appropriate workers using a heuristic algorithm. Simulation experiments demonstrate that TP-TASC effectively minimizes waiting time and maximizes result quality.
DECISION SUPPORT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Hao Wang, Chengcheng Yang, Xiangliang Zhang, Xin Gao
Summary: This paper presents a novel disk-based LSH index that provides efficient support for both searches and updates. By utilizing write-friendly LSM trees to store LSH projections and developing a novel estimation scheme, the efficiency of search and the cost-effectiveness of disk storage and access are improved. Experimental results demonstrate that the proposed method outperforms state-of-the-art schemes on four real-world datasets.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Engineering, Civil
Yanhong Li, Wang Zhang, Yunjun Gao, Qing Li, Lihchyun Shu, Changyin Luo
Summary: This paper proposes a direction-aware augmented spatial keyword top-k query (DATkQ) that considers various factors to return the top-k objects. The paper focuses on answering the why-not question in DATkQs and introduces methods to refine the query direction and prune irrelevant search space. The efficiency of the proposed approach is demonstrated through experiments on real datasets.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Songbai Liu, Qiuzhen Lin, Ka-Chun Wong, Qing Li, Kay Chen Tan
Summary: This study compares existing optimizers for evolutionary large-scale multiobjective optimization (ELMO) on different benchmarks and finds that significant improvements are needed in both benchmarks and algorithms for ELMO. Therefore, a new test suite and optimizer framework are proposed to further advance ELMO research. The new benchmarks incorporate more realistic features challenging for existing optimizers, and the proposed optimizer, with a variable group-based learning strategy, shows distinct advantages in tackling these benchmarks.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Feng-Feng Wei, Wei-Neng Chen, Qing Li, Sang-Woon Jeon, Jun Zhang
Summary: This article defines distributed expensive constrained optimization problems (DECOPs) and proposes a distributed evolutionary constrained optimization algorithm with on-demand evaluation (DEAOE). DEAOE adaptively evolves different constraints in an asynchronous way through on-demand evaluation, improving population convergence and diversity. Experimental results demonstrate that DEAOE outperforms centralized state-of-the-art surrogate-assisted evolutionary algorithms (SAEAs) in terms of performance and efficiency.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Information Systems
Zongxi Li, Xianming Li, Haoran Xie, Fu Lee Wang, Mingming Leng, Qing Li, Xiaohui Tao
Summary: Researchers have found that emotion is not limited to one category in emotion-relevant classification tasks, and multiple emotions can exist together in a sentence. Recent studies have focused on using distribution or grayscale labels to enhance the classification model, providing additional information on the intensity of emotions and their correlations. This approach has been effective in overcoming overfitting and improving model robustness. However, it can also reduce the model's discriminative ability within similar emotion categories.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Computer Science, Information Systems
Ziang Zhou, Jieming Shi, Shengzhong Zhang, Zengfeng Huang, Qing Li
Summary: Graph neural networks (GNNs) are designed for semi-supervised node classification on graphs with limited labeled data. However, in extreme cases where very few labels are available (e.g., only 1 labeled node per class), GNNs suffer from severe performance degradation. To address this issue, the proposed Stabilized Self-Training (SST) framework effectively handles the scarcity of labeled data and improves classification accuracy.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Liangrui Ren, Guoxian Yu, Jun Wang, Lei Liu, Carlotta Domeniconi, Xiangliang Zhang
Summary: This paper proposes a method named iMClusts that uses deep autoencoders and multi-head attention to generate multiple salient embedding matrices and clusterings. It enhances the quality and diversity of clusterings by leveraging multi-facet knowledge.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Zhenguo Yang, Jiale Xiang, Jiuxiang You, Qing Li, Wenyin Liu
Summary: This paper introduces a new E-VQA dataset that includes free-form questions and answers for real-world event concepts, providing context information of events as domain knowledge. The dataset is valuable for researching and evaluating VQA methods.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Zhiqi Lei, Hai Liu, Jiaxing Yan, Yanghui Rao, Qing Li
Summary: Lifelong topic modeling (LTM) has gained attention for mining high quality topics in a stream of documents. However, the permutation of topics may lead to a semantic misalignment between the topic representations of document chunks. To address this issue, researchers propose a non-negative matrix tri-factorization (NMTF) based framework (NMTF-LTM) and a distributed parallel algorithm (PNMTF-LTM) to achieve semantic alignment and real-time processing.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Jiayuan Xie, Jiali Chen, Wenhao Fang, Yi Cai, Qing Li
Summary: Visual question generation focuses on target objects in images to generate questions for specific questioning purposes. Existing studies extract target objects corresponding to the questioning purpose based on answers. However, answers may not accurately and completely map to every target object. This study proposes a content-controlled question generation model that generates questions based on a given target object set specified from an image.
Article
Chemistry, Multidisciplinary
Runze Mao, Wenqi Fan, Qing Li
Summary: Training Graph Neural Networks (GNNs) on large-scale graphs in the deep learning era can be expensive. Graph condensation has emerged as a promising approach to reduce training cost by compressing large graphs, but its fairness in treating node subgroups during compression has not been explored.
APPLIED SCIENCES-BASEL
(2023)
Article
Acoustics
Lin Xiao, Pengyu Xu, Mingyang Song, Huafeng Liu, Liping Jing, Xiangliang Zhang
Summary: Multi-label text classification aims to tag relevant labels for documents. Annotated new documents for multi-label text classification is more difficult than in the standard multi-class case. The proposed TAPON significantly outperforms other methods for long-tailed multi-label text classification.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2023)
Article
Automation & Control Systems
Waqas W. Ahmed, Mohamed Farhat, Pai-Yen Chen, Xiangliang Zhang, Ying Wu
Summary: This article proposes and demonstrates a generative deep learning approach for shape recognition of an arbitrary object using its acoustic scattering properties. The approach utilizes deep neural networks to learn the mapping between the latent space of a 2D acoustic object and the far-field scattering amplitudes. By training an adversarial autoencoder, the neural network determines the latent space of the acoustic object, embedding important structural features and accelerating the learning process for inverse design.
ADVANCED INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Information Systems
Yuqi Bu, Liuwu Li, Jiayuan Xie, Qiong Liu, Yi Cai, Qingbao Huang, Qing Li
Summary: This article introduces a new task called scene-text oriented referring expression comprehension and proposes a scene text awareness network to address alignment and error issues. Experimental results show that the proposed method effectively comprehends scene-text oriented referring expressions and achieves excellent performance.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Computer Science, Artificial Intelligence
Jiajin Wu, Bo Yang, Runze Mao, Qing Li
Summary: Sequential recommendation systems have gained significant attention, but current models still suffer from popularity bias. To alleviate this bias, this study proposes a debiasing model that considers the dynamic user desire and conducts intervention analysis and counterfactual reasoning. The proposed model, PAUDRec, outperforms existing models while alleviating popularity bias in sequential recommendation.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
