检索结果-内蒙古大学图书馆

Generating High-quality Movie Tags from Social Reviews: A Learning-driven Approach

Generating High-quality Movie Tags from Social Reviews: A Le...

IEEE/ACM Int'l Conference on & Int'l Conference on Cyber, Physical and Social computing (CPSCom) Green computing and Communications (GreenCom)

作者： Zhenxiao Luo Guopin Tang Chen Wang Yipeng Zhou Xi Zheng Jessie Hui Wang Gang Liu Di Wu School of Computer Science and Engineering Sun Yat-Sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Faculty of Science and Engineering Macquarie University Sydney Australia Institute for Network Sciences and Cyberspace Beijing National Research Center for Information Science and Technology Tsinghua University Beijing China College of Computer Science and Software Engineering Shenzhen University Shenzhen China

ISBN: (纸本)9781665417631

As a kind of keywords to describe video contents, tags are extremely beneficial for viewers to locate videos when they search the site. Additionally, tags also help video platform operators to better organize and recommend videos to platform users. Previous approaches are mainly based on manually tagging or tag propagation via video content analysis, which is time-consuming and resource-consuming. Especially, given the high volume of fresh movies generated per year, it is difficult to accurately tag all the movies manually. Moreover, video content analysis is also not easy considering typical features (e.g., long duration, complicated scenarios) of commercial movies. In this paper, we propose an automatic tagging algorithm called TagRec that exploits crowdsourced user reviews to generate accurate movie tags. We observe that user reviews contain rich information about movies (e.g., quality, actors) which can be learned to generate high-quality movie tags. Inspired by the above observation, we choose to transform the movie video tagging problem into a tag recommendation problem, in which tags are recommended to different movies by extracting knowledge from crowdsourced movie reviews. We take latent topics, tag co-occurrence probability and tag semantics into account, and formulate the problem as a recommendation optimization problem. We evaluate the performance of our pro-posed TagRec algorithm with a large-scale real-world dataset. Extensive experiments demonstrate that TagRec achieves 7.1 % and 9.6% improvement compared with other state-of-the-art methods in terms of Hit Ratio and Normalized Discounted Cumulative Gain respectively.

关键词： Social computing Filtering Semantics Transforms Tagging Motion pictures Optimization

来源：评论

学校读者我要写书评

暂无评论

network Occlusion Sensitivity Analysis Identifies Regional Contributions to Brain Age Prediction

引用

Human Brain Mapping 2025年第8期46卷 e70239页

作者： Lingfei He Siyu Wang Cheng Chen Yaping Wang Qingcheng Fan Congying Chu Lingzhong Fan Junhai Xu College of Intelligence and Computing Tianjin Key Laboratory of Cognitive Computing and Application Tianjin University Tianjin China Brainnetome Center Beijing China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing China Sino-Danish Center Beijing China University of Chinese Academy of Sciences Beijing China Chinese Institute for Brain Research Beijing China CAS Center for Excellence in Brain Science and Intelligence Technology Institute of Automation Chinese Academy of Sciences Beijing China University of Health and Rehabilitation Sciences Qingdao China

Deep learning frameworks utilizing convolutional neural networks (CNNs) have frequently been used for brain age prediction and have achieved outstanding performance. Nevertheless, deep learning remains a black box as it is hard to interpret which brain parts contribute significantly to the predictions. To tackle this challenge, we first trained a lightweight, fully CNN model for brain age estimation on a large sample data set ( N = 3054, age range = [8,80 years]) and tested it on an independent data set ( N = 555, age range = [8,80 years]). We then developed an interpretable scheme combining network occlusion sensitivity analysis (NOSA) with a fine-grained human brain atlas to uncover the learned invariance of the model. Our findings show that the dorsolateral, dorsomedial frontal cortex, anterior cingulate cortex, and thalamus had the highest contributions to age prediction across the lifespan. More interestingly, we observed that different regions showed divergent patterns in their predictions for specific age groups and that the bilateral hemispheres contributed differently to the predictions. Regions in the frontal lobe were essential predictors in both the developmental and aging stages, with the thalamus remaining relatively stable and saliently correlated with other regional changes throughout the lifespan. The lateral and medial temporal brain regions gradually became involved during the aging phase. At the network level, the frontoparietal and the default mode networks show an inverted U-shape contribution from the developmental to the aging stages. The framework could identify regional contributions to the brain age prediction model, which could help increase the model interpretability when serving as an aging biomarker.

关键词： brain age brain atlas convolutional neural network interpretability MRI

来源：评论

学校读者我要写书评

暂无评论

Design and Application of a Portable Sleep Inertia Detection System Based on EEG Signals

Design and Application of a Portable Sleep Inertia Detection...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Yunzhi Cui Fuze Tian Qinglin Zhao Bin Hu Gansu Provincial Key Laboratory of Wearable Computing School of Information Science and Engineering Lanzhou University Lanzhou China CAS Center for Excellence in Brain Science and Intelligence Technology Shanghai Institutes for Biological Sciences Chinese Academy of Sciences Chinese Academy of Sciences Joint Research Center for Cognitive Neurosensor Technology of Lanzhou University & Institute of Semiconductors Ministry of Education Engineering Research Center of Open Source Software and Real-Time System (Lanzhou University) Lanzhou China

ISBN: (纸本)9781665429825

Sleep inertia is a transitional state from sleep to wakefulness, accompanied by groggy feelings and cognitive impairment. Previous research on sleep inertia mainly used expensive and cumbersome equipment, and the analysis of physiological signals relied on computers. This work introduces a sleep inertia detection system that consists of a wearable low-power electroencephalogram (EEG) acquisition module based on STM32WB55 and ADS1299, and a data processing module based on the Xilinx ® Zynq ® -7000 XC7Z020. This work recorded the EEG signals of ten subjects in the alert and sleep inertia states to extract the delta power, alpha power, beta power, EEG vigilance, and sample entropy. A linear support vector machine (SVM) was then used to classify the two states based on all subjects’ EEG signals, with an accuracy of 72.5%, and the average accuracy based on a single participant was 8S.9%. Finally, the feature extraction algorithm and SVM parameters were entered into the Zynq ® system-on-chip (SoC) development board to realize onboard processing of the algorithm. The system is capable of evaluating the severity of human sleep inertia, which has reference significance for the practical application of sleep inertia detection.

关键词： Support vector machines Sleep Conferences Feature extraction data processing Electroencephalography Physiology

来源：评论

学校读者我要写书评

暂无评论

Sea-land Segmentation of Synthetic Aperture Radar Imagery Using Deep Neural network Models

Sea-land Segmentation of Synthetic Aperture Radar Imagery Us...

引用

Progress in Electromagnetic Research Symposium (PIERS)

作者： Yinfei Zhou Gang Zheng Jingsong Yang Xiaofeng Li Bin Liu Jingliang Shao Han Jiang Ren Ren Peng Chen Lin Ren Bin Zhang Ocean College Zhejiang University Zhoushan China Ministry of Natural Resources State Key Laboratory of Satellite Ocean Environment Dynamics Second Institute of Oceanography Hangzhou China Chinese Academy of Sciences CAS Key Laboratory of Ocean Circulation and Waves Big Data Center Institute of Oceanology Qingdao China Chinese Academy of Sciences Center for Ocean Mega-Science Institute of Oceanology Qingdao China College of Marine Sciences Shanghai Ocean University Shanghai China School of Automation Engineering University of Electronic Science and Technology of China Chengdu China

ISBN: (纸本)9781665409889

The waterline is the transient boundary between land and sea, and the waterline extraction is of great significance for coastal research and monitoring. Spaceborne synthetic aperture radar (SAR) has the ability to penetrate clouds and thus can image large coastal areas with high-resolution under all-weather and all-day conditions. These advantages make SAR well complement the waterline observations by optical satellites. However, SAR images of coastal areas usually have complex patterns because of the diverse topography and geomorphology, and the SAR-imaging mechanism also results in large coherent speckle noise in the images. The above poses challenges to the waterline extraction from SAR imagery. Traditional manual annotation is highly accurate, but is laborious and time-consuming. Therefore, a more efficient method is urgent. The procedure of waterline extraction is sea-land segmentation followed by boundary-extraction where accurate segmentation is critical. From the perspective of the computer vision, sea-land segmentation can be viewed as the image semantic segmentation task. In recent years, deep neural network (DNN) models have achieved great success in the semantic segmentation task. The U-Net, named for its U-shape architecture, is one of the most popular DNN models for the task. Compared with others, the U-Net has a lighter structure that efficiently extract and fuse the multi-scale image features. In this study, we applied the U-Net and its variants to sea-land segmentation and systematically investigated the models’ performances. All models can segment the sea and land efficiently, and the results are overall consistent with the manual annotation. Our study demonstrates the application potential of DNN technology in waterline monitoring from SAR imagery.

关键词： Deep learning Image segmentation Neural networks Semantics Sea measurements Manuals Radar polarimetry

来源：评论

学校读者我要写书评

暂无评论

Extending Registration and Authentication Processes of FIDO2 External Authenticator with QR Codes

Extending Registration and Authentication Processes of FIDO2...

引用

IEEE International Conference on Trust, Security and Privacy in computing and Communications (TrustCom)

作者： Chengqian Guo Quanwei Cai Qiongxiao Wang Jingqiang Lin State Key Laboratory of Information Security Institute of Information Engineering CAS Data Assurance and Communication Security Research Center CAS School of Cyber Security University of Chinese Academy of Sciences School of Cyber Security University of Science and Technology of China

ISBN: (数字)9781665403924

ISBN: (纸本)9781665403931

FIDO2, the newest set of FIDO specifications, enables the user to leverage an external authenticator for the authentication in both mobile and desktop environments (referred to as user agent). For the secure registration and authentication, FIDO2 requires the external authenticator and user agent to establish a confidential and mutually authenticated data transport channel through either USB interfaces, Near Field Communication (NFC) or Bluetooth. However, the external authenticator and host may not be equipped with one of the above physical media simultaneously, for example, a desktop may only have USB interfaces while an external authenticator (e.g., a smartphone) may have no USB inferfaces. This affects the wide adoption of FIDO2. In this paper, we extend the registration and authentication processes of FIDO2 external authenticator with QR code, which enables the external authenticator being equipped with a camera to be used for the authentication at any user agent. During the registration process, our scheme requires the user to provide the original credential and a one-time password displayed on the authenticator, and therefore ensures the correct user will only be bound with the expected authenticator. The security of our scheme has been formally analyzed based on the Dolev-Yao style model, a widely adopted model for the analysis of web systems. We have implemented the prototype, and the performance evaluation demonstrated the efficiency of our scheme, which needs 373 ms for registration and 141 ms for authentication in our environment.

关键词： Analytical models Privacy Authentication Prototypes Near field communication Universal Serial Bus Security

来源：评论

学校读者我要写书评

暂无评论

Coupled graph neural networks for predicting the popularity of online content

arXiv

引用

arXiv 2019年

作者： Cao, Qi Shen, Huawei Gao, Jinhua Wei, Bingzheng Cheng, Xueqi CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China WeChat Tencent Inc

Predicting the popularity of online content in social network is an important problem for the practice of information dissemination, advertising, and recommendation. Previous methods mainly leverage demographics, temporal and structural patterns of early adopters for popularity prediction. These methods ignore the interaction between early adopters and potential adopters or the interactions among potential adopters over social networks. Consequently, they fail to capture the cascading effect triggered by early adopters in social networks, and thus have limited predictive power. In this paper, we consider the problem of network-aware popularity prediction, leveraging both early adopters and social networks among users for popularity *** propose a novel method, namely Coupled-GNNs, which use two coupled graph neural networks to capture the cascading effect in information diffusion. One graph neural network models the interpersonal influence, gated by the adoption state of users. The other graph neural network models the adoption state of users via interpersonal influence from their neighbors. Through such an iterative aggregation of the neighborhood, the proposed method naturally captures the cascading effect of information diffusion in social networks. Experiments conducted on both synthetic data and real-world SinaWeibo data demonstrate that our method significantly outperforms the state-of-the-art methods for popularity prediction. Copyright © 2019, The Authors. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

MDPL-net: Multi-layer Dictionary Learning network with Added Skip Dense Connections

MDPL-net: Multi-layer Dictionary Learning Network with Added...

引用

IEEE International Conference on data Mining (ICDM)

作者： Zhao Zhang Yulin Sun Zheng Zhang Yang Wang Lin Wu Meng Wang School of Computer Science and Technology Soochow University Suzhou China Ministry of Education Key Laboratory of Knowledge Engineering with Big Data School of Computer Science and Information Engineering Hefei University of Technology Hefei China Bio-Computing Research Center Harbin Institute of Technology (Shenzhen) Shenzhen China

ISBN: (数字)9781728183169

ISBN: (纸本)9781728183176

Dictionary learning (DL) is powerful for representation learning, while it fails to capture the deep hierarchical information hidden in data. In this paper, we propose a new generalized end-to-end mulita-layer representation learning architecture referred to as Multi-layer Dictionary Pair Learning network (MDPL-net) for the deep sparse and hierarchical representation of images. To enable MDPL-net to conduct accurate classification, MDPL-net clearly integrates the skip connection end-to-end network and multi-layer deep sparse dictionary learning into a unified architecture. The representation learning module has several hidden DL blocks, where each hidden DL block has a dictionary pair learning (DPL) layer, a batch-norm layer and an activation function layer, and the DL blocks are connected in a feed-forward manner. To further improve the information flow and maintain the privileged features between different DL blocks, a novel skip dense connectivity pattern is deployed between hidden DL blocks, which can obtain more stable and discriminative features. The DPL layer jointly formulates the discriminative synthesis dictionary and analysis dictionary by minimizing reconstruction error within each batch over the feature maps from front layers. Extensive results on benchmark databases demonstrate the effectiveness of MDPL-net for discriminative representation and robust image classification.

关键词： Dictionaries databases Conferences Machine learning data mining Image reconstruction Image classification

来源：评论

学校读者我要写书评

暂无评论

Deglacial and Holocene Evolution of North Pacific Upper and Intermediate Ocean Circulation

引用

Paleoceanography and Paleoclimatology 2025年第6期40卷

作者： Hu Chang Lanlan Zhang Xufeng Zheng Xun Gong Yiping Yang Rong Xiang State Key Laboratory of Tropical Oceanography Key Laboratory of Ocean and Marginal Sea Geology South China Sea Institute of Oceanology Chinese Academy of Sciences Guangzhou China School of Geographic and Oceanographic Sciences Key Laboratory for Coast and Island Development of Ministry of Education Nanjing University Nanjing China University of Chinese Academy of Sciences Beijing China State Key Laboratory of Marine Resource Utilization in South China Sea Hainan University Haikou China Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center Qilu University of Technology (Shandong Academy of Sciences) Jinan China State Key Laboratory of Physical Oceanography Jinan China

The interplay between the northward-flowing surface oceanic currents and the southward-flowing intermediate water masses regulated ocean stratification and mixing in the Northwest Pacific. Radiolaria, a biogenic proxy, was used to reconstruct changes in ocean stratification in the central Okinawa Trough (OT) paleoceanography since 20 ka. Our results revealed that changes in the shallow water (0–500 m), upper-intermediate water (500–1,000 m), and lower-intermediate water (1,000–1,600 m) are triggered by variations in the Kuroshio Current (KC), the North Pacific Intermediate Water (NPIW), and Pacific Central Water (PCW), respectively. The KC influence was stronger during the Holocene rather than during the last deglaciation, increased during the ∼11.7–8.2 ka, weakened during the 8.2–4 ka, declined at 4–2 ka, and rapidly increased after 2 ka. In contrast, the influence of PCW increased during the last deglaciation, particularly during Heinrich Stadial 1 (HS1) and Younger Dryas (YD), but remained weak throughout the Holocene. This pattern appears to be closely linked to the Atlantic/Antarctic provenance events associated with Atlantic Meridional Overturning Circulation variability. Meanwhile, the influence of NPIW was strong during the YD but gradually weakened throughout the Holocene. Notably, a rapid decline in NPIW influence occurred during the Neoglaciation (∼4–2 ka), similar to the KC, suggesting a high‒low latitude linkage driven by atmosphere‒ocean feedbacks mechanisms. This study provides robust data and a new perspective for understanding the connection to deglacial and Holocene variability in North Pacific upper and intermediate ocean circulation. Siliceous microzooplankton radiolarian records used to reconstruct the dynamic changes in the vertical structure of OT water since 20 ka Water changes at 0–500 m, 500–1,000 m, and 1,000–1,600 m in the OT under the impacts of the KC, North Pacific Intermediate Water (NPIW) and Pacific Central Water, respectively A rap

关键词： last deglacial-Holocene Kuroshio current north Pacific intermediate water Pacific central water radiolarian proxy high-low latitude linkage

来源：评论

学校读者我要写书评

暂无评论

Controlling risk of web question answering

arXiv

引用

arXiv 2019年

作者： Su, Lixin Guo, Jiafeng Fan, Yixing Lan, Yanyan Cheng, Xueqi CAS Key Lab of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

Web question answering (QA) has become an indispensable component in modern search systems, which can significantly improve users’ search experience by providing a direct answer to users’ information need. This could be achieved by applying machine reading comprehension (MRC) models over the retrieved passages to extract answers with respect to the search query. With the development of deep learning techniques, state-of-the-art MRC performances have been achieved by recent deep methods. However, existing studies on MRC seldom address the predictive uncertainty issue, i.e., how likely the prediction of an MRC model is wrong, leading to uncontrollable risks in real-world Web QA applications. In this work, we first conduct an in-depth investigation over the risk of Web QA. We then introduce a novel risk control framework, which consists of a qualify model for uncertainty estimation using the probe idea, and a decision model for selectively output. For evaluation, we introduce risk-related metrics, rather than the traditional EM and F1 in MRC, for the evaluation of risk-aware Web QA. The empirical results over both the real-world Web QA dataset and the academic MRC benchmark collection demonstrate the effectiveness of our approach. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Cyber security monitoring and data analysis at IHEP

Cyber security monitoring and data analysis at IHEP

引用

2019 International Symposium on Grids and Clouds, ISGC 2019

作者： Yan, Tian Hu, Hao An, Dehai Qi, Fazhi Jiang, Chen Computing Center Institute of High Energy Physics Chinese Academy of Sciences Beijing100049 China Key Laboratory of Network Assessment Technology CAS Institute of Information Engineering Chinese Academy of Sciences Beijing100093 China Inc. Beijing100085 China

Recently, cyber security threats becomes a noticeable challenge for academic institutes. In this paper, we present the security risk control model and the cyber security detection and monitoring system designed and deployed at institute of High Energy Physics (IHEP) in China. Security data collection and analysis plays the central role in this framework. In addition to the open-source solution like Zeek, MISP and ELK stack, we also deployed a commercial Security Operation Center (SOC) as a supplement and cross-check solution. Copyright © 2019 the author(s).

关键词： Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：